is speech intelligibility meaning

• Home →
• Blog →
• Late Talker Blog Series
• → What is Speech Intelligibility?

What is Speech Intelligibility

Intelligibility of speech is the percentage of speech that a listener can understand. If you can only understand half of what a child is saying then their speech intelligibility rating would be 50%.

​Speech Intelligibility changes with a child’s age. Speech development begins with babbling and then speech matures until older children can say all the sounds in their primary language/s and everyone can understand them.

​Speech Intelligibility is Expected to Improve with a Child’s Age

Younger children are expected to be harder to understand then older children. Speech Sound Developmental Checklists and Speech sound charts can help parents, teachers and carers to see if the level of speech intelligibility is at expected levels for a child’s age.

​The graph below shows that an 18 month old child will have lower speech intelligibility levels than a 3 year old child. This does not mean that a 3 year old will not still make some speech sound errors. It means that they are only using a few speech sound errors and that most people will be able to understand what they are saying. Lynch, Brookshire & Fox (1980), p. 102, cited in Bowen (1998)

​Why does Speech Intelligibility Change with Speech Developmental Stages?

There are 3 main areas to consider when looking at speech intelligibility

1 articulation, toddlers do not have the fine motor control of their tongues and lips to be able to say all the sounds correctly and so their speech is characterised by speech sound developmental errors..

​These are called articulation errors. For example, sounds like the “r” sound can be difficult for toddlers. The “r” sound requires a fine curling of the tongue and young children often will substitute the “r” sound for an easier sound (e.g., w). You will often hear a 2 year old say “wed” for “red”.

2 Motor speech co-ordination 

​Young children also cannot move their tongue and lips fast enough to keep their speech clear as they try to say longer words and sentences. This is called motor speech co-ordination. It is a bit like doing buttons up. Young children may be able to do easy buttons but might take a long time. As their fine motor co-ordination increase their ability to dress becomes easier and faster.

3 Phonological Processes

​toddlers use more phonological processes that reduces speech intelligibly than older children. as toddlers do not have the oral motor skills to say words 100% correctly, they use speech sound patterns that simplify words..

​One example of a phonological process is called “cluster reductions”. Speech sound clusters like “sp, sk, dr, bl” are very difficult for young children to say. Most young children reduce the cluster to one sound (e.g., “dep” for “step”, “back” for “black”). Young children use lots of phonological processes reducing speech intelligibility. As children get older they use less and less and so speech intelligibility improves.

​As speech develops children use less articulation errors, their rate & co-ordination of speech improves and they use less phonological errors.

​See the Speech Sound Developmental Checklist to see the ages speech sounds and phonological processes are typically present.

​Speech Intelligibility as a measure of Severity of Speech Delays and Speech Disorders.

Children with speech delays and speech disorders will often have lower speech intelligibility percentages than same age children.

​If a 3 year old child has speech delays they may be still making speech sound errors like a 2 year old would be making. This can reduce their speech intelligibility for their age.

​Speech disorders such as dyspraxia of speech (CAS) is characterised by significant reductions in speech intelligibility. Speech intelligibility may be one of the criteria used to determine how functional a child’s speech is in their community. If most people they interact with them cannot understand them, then it would be considered a significant speech impairment.

You may also be interested in...

Baby speech development.

Toddler Speech Development

Preschool Speech Development

Primary School Speech Development

High School Speech Development

Sandra McMahon

Leave a reply:.

Subscribe to Get the Latest News!

March 10, 2024

Speech Intelligibility in Children

There are many factors that impact intelligibility, or how easy it is to understand someone who is speaking. A young child learning how to talk will be much more unintelligible than an older child who is mastering their speech. Speech intelligibility in children increases as a child ages, with most children being quite intelligible by 7 years of age.

Scroll to the bottom of this post to download a Speech Intelligibility Norms PDF!

What is Speech Intelligibility?

Speech intelligibility is how clear or understandable speech is to others. In other words, when someone is speaking to you, how much of what they’re saying can you understand?

Intelligibility in adults is often dependent on many factors – including background noise, the complexity of the conversation, differing accents or dialects, and more. Adults are not always 100% understandable or clear when speaking to others, and something like loud noise at a party can make it more difficult to understand your communication partner.

Intelligibility in children improves with age and speech and language development. Younger children are more unintelligible (difficult to understand) as they have not mastered all speech sounds and rules of language. Children are often more intelligible to parents and those whom they communicate with frequently versus strangers. Additionally, children with speech and language delays, hearing loss, genetic syndromes, or other disorders are likely to have lower speech intelligibility than their peers.

Speech Intelligibility by Age

In a 2021 study on speech intelligibility in children, Hustad et al. sampled the speech of 538 typically developing children, had a group of 1,076 adults transcribe the speech samples, and used this information to calculate single-word and multi-word intelligibility.

Results from the study found that the average child should be at least 50% intelligible by 4 years old (meaning you should be able to understand about half of what a 4-year-old is saying when they speak to you) and 90% intelligible by 7 years old. Other speech intelligibility norms from this study include the following:

• 25% intelligible by 3 years old
• 50% intelligible by 4 years old
• 75% intelligible by 5 years old
• 90% intelligible by 7 years old

Important Note: These percentages are based on unfamiliar listeners with unknown contexts . An unfamiliar listener is someone who does not communicate with that person on a regular basis. Parents can expect their children to be more intelligible than these reported numbers since they are familiar with their children’s speech. Children will also be more intelligible, or understandable if the listener understands the context – the situation in which the conversation is taking place.

Keeping these caveats in mind, the following intelligibility percentages are likely to be HIGHER for parents and those who understand the context (the environment or situation that helps clarify the meaning) of what the child is saying.

As previously mentioned, a child’s intelligibility should improve as they get older. While 2- and 3-year-olds are still fairly difficult to understand, older school-age children 7 years old and older should be 90-100% understandable.

Here’s a closer look at intelligibility with unfamiliar listeners by age:

3 Years Old

A 3-year-old’s speech is approximately 25% intelligible to unfamiliar listeners. Intelligibility improves the fastest between 2 1/2 to 3 1/2 years of age. If you have a 2-year-old who is very difficult for others to understand, you can anticipate a major growth over the next year or two with them being much easier to understand by the time they are 4 years old.

A 3-year-old’s intelligibility ranges between the following:

• Single words: 24% – 48% intelligible
• Sentences: 12% – 41% intelligible

As you can see, there is quite a range in how understandable 3-year-olds should be, especially when they are speaking in sentences. This lower intelligibility is still age-appropriate and will continue to improve into their 4th year.

4 Years Old

A 4-year-old’s speech is approximately 50% intelligible to unfamiliar listeners. As a quick rule of thumb, a 4-year-old is about 50% intelligible to unfamiliar listeners. A family member or someone who communicates with a 4-year-old very often should expect that child to be over 50% intelligible.

A 4-year-old’s intelligibility ranges between the following:

• Single words: 44% – 67% intelligible
• Sentences: 41% – 69% intelligible

5 Years Old

A 5-year-old’s speech is approximately 75% intelligible to unfamiliar listeners. When a child turns 5 years old, they should be about 75% intelligible to an unfamiliar listener. Remember, that it is likely they will be more intelligible to parents, siblings, and other close individuals, so if a parent or caregiver is not understanding over half of what their child is saying, there may be cause for concern.

A 5-year-old’s intelligibility ranges between the following:

• Single words: 59% – 79% intelligible
• Sentences: 66% – 85% intelligible

6 Years Old

A 6-year-old’s speech is approximately 85% intelligible to unfamiliar listeners. Six-year-olds are around 80% intelligible when speaking to others. From the ages of 6 and up, children are fairly understandable when they speak to others, but they still have some growth to make in the next few years to get them closer to being 100% intelligible.

A 6-year-old’s intelligibility ranges between the following:

• Single words: 68% – 86% intelligible
• Sentences: 79% – 92% intelligible

7 Years Old

A 7-year-old’s speech is approximately 90% intelligible to unfamiliar listeners. By 7 years of age, we should expect children to be around 90% intelligible. This means most words and sentences a 7-year-old says should be understandable with a very small portion of their speech unintelligible.

A 7-year-old’s intelligibility ranges between the following:

• Single words: 73% – 89% intelligible
• Sentences: 87% – 95% intelligible

8 Years Old

An 8-year-old’s speech is approximately 94% intelligible to unfamiliar listeners. Intelligibility continues to improve after 7 years, albeit at a much slower pace. As you can see in the next 3 ages, intelligibility only increases by a percentage or two each year.

An 8-year-old’s intelligibility ranges between the following:

• Single words: 76% -91% intelligible
• Sentences: 91% – 97% intelligible

9 Years Old

A 9-year-old’s speech is approximately 96% intelligible to unfamiliar listeners. There is little change in intelligibility from 8 years of age to 9 years of age, but you can expect a slight improvement in how much of a child’s speech unfamiliar listeners understand.

A 9-year-old’s intelligibility ranges between the following:

• Single words: 77% – 91% intelligible
• Sentences: 94% – 98% intelligible

10 Years Old

A 10-year-old’s speech is approximately 97% intelligible to unfamiliar listeners. By the time a child celebrates their 10th birthday, you can expect them to be highly intelligible. Most teenagers and adults are never 100% intelligible, so you can expect to understand a 10-year-old very well.

A 10-year-old’s intelligibility ranges between the following:

• Single words: 78% – 93% intelligible
• Sentences: 95% – 98% intelligible

Speech Intelligibility Norms Chart

This speech intelligibility norms chart is a great download to quickly reference the sentence-level intelligibility of English-speaking children. All norms are reference-based and updated with the most recent information available.

Click on the image below to download the speech intelligibility norms PDF.

Is Your Child Difficult to Understand?

These intelligibility norms are a guideline for teachers, parents, and speech-language pathologists to use to assess a child’s speech development. Many other considerations should be taken when determining if a child’s speech and communication development is on track.

Remember! These intelligibility norms indicate sentence-level intelligibility of English-speaking children with unfamiliar listeners in unknown contexts. If a familiar listener (such as a close family member or friend) is having difficulty understanding a child at the percentages listed above, it may be beneficial to reach out to a licensed speech therapist in your area.

If you have concerns about your child’s speech, please contact us or speak with a local speech-language pathologist for further information and recommendations.

More Speech Norms

Interested in learning more about speech norms and speech pathology? Check out these posts below!

• Speech Sound Development
• What is a Speech Pathologist, and What Do They Do

Hustad, K.C., Mahr, T.J., Natzke, P., & Rathouz, P.J. (2021). Speech development between 30 and 119 months in typical children I: Intelligibility growth curves for single-word and multiword productions. Journal of Speech, Language, and Hearing Research. https://doi.org/10.1044/2021_JSLHR-21-00142

Share This Article

To the newsletter, the newsletter.

Subscribe to The Speech Guide newsletter and never miss a new post, handy resource, or freebie.

You have successfully joined our subscriber list.

Leave A Comment Cancel reply

Save my name, email, and website in this browser for the next time I comment.

Related Posts

Speech Sound Development Chart

© Copyright 2022

• Get new issue alerts Get alerts

Secondary Logo

Journal logo.

Colleague's E-mail is Invalid

Your message has been successfully sent to your colleague.

Save my selection

The Speech Intelligibility Index

What is it and what's it good for.

Editor(s): Mueller, Gus

Page Ten Editor

If you're not quite sure what's the difference between the AI and the SII and how you use each of them, let this month's Page Ten explain the two measures and discuss their value in a daily clinical practice.

1 I guess I haven't been keeping up very well, so, for starters, let me ask you, what exactly is the Speech Intelligibility Index?

The Speech Intelligibility Index, or SII, is a measure, ranging between 0.0 and 1.0 “that is highly correlated with the intelligibility of speech.” (ANSI, S3.5, 1997, p. 1). 1 You're probably not the only one who isn't familiar with the term. Although drafts of the standard were around in the mid-1990s, it wasn't until the revision of the ANSI S3.5 standard in 1997 that the term SII formally replaced the more familiar term AI (for Articulation Index). The term SII is just starting to find its way into clinical settings.

2 So, are you saying this is just a new name for the Articulation Index?

We can talk about the details later, but for now let me just say yes, there are many similarities between the old AI and the SII. The SII, like the AI, is a quantification of the proportion of speech information that is both audible and usable for a listener. Basically, an SII of 0 implies that none of the speech information, in a given setting, is available (audible and/or usable) to improve speech understanding. An SII of 1.0 implies that all the speech information in a given setting is both audible and usable for a listener.

Generally, there is a monotonic relationship between the SII and speech understanding. That is, as the SII increases, speech understanding generally increases. The method for calculating the SII is described in the ANSI S3.5 (1997) standard titled “American National Standard Methods for Calculation of the Speech Intelligibility Index.” 1

3 You're talking as if the SII and speech understanding are similar, but not exactly the same. Is that right?

You are correct. They are not the same, although this is a common misconception. For example, having an SII of 0.5 in a certain environment does not mean you would understand 50% of speech. It simply means that about 50% of speech cues are audible and usable in a given setting. It turns out that for most conversational speech stimuli an SII of 0.5 would correspond to close to 100% intelligibility for individuals with normal hearing. Some researchers have suggested that we use the term “Audibility Index” to remind us that the AI (or SII) is not a direct measure of intelligibility . 2

4 So are you saying that we can predict a person's speech understanding based on his SII score?

Yes, the SII (and AI) can be used to predict speech recognition scores by means of an empirically derived transfer function . These transfer functions are based on the specific speech materials being used during testing. Examples of transfer functions for several types of speech materials are shown in Figure 1 . These functions show that a single SII value can correspond to multiple speech-recognition scores. The actual speech score depends on the speech material used during testing, as well as the proficiency of the talker and listener. 3

5 So it still sounds as if the SII is just the AI in a new package. Am I right?

Well, you're persistent. Here are some of the details: The 1997 document is a revision of the 1969 ANSI S3.5 document titled “American National Standard Methods for Calculation of the Articulation Index.” 7 In this article, I'll use the acronym SII to refer specifically to the 1997 version of the ANSI S3.5 standard and the term AI to refer to the 1969 version of the S3.5 document and other more simplified calculation methods.

The SII and AI share a common ancestry being based, in large part, on early work conducted by Harvey Fletcher and colleagues at American Telephone and Telegraph's (AT&T's) Western Electric Research Laboratories (later to become Bell Telephone Laboratories) in the early and mid-20th century. 8 These researchers were interested in developing a method to predict the impact of changes in telephone circuits (e.g., changes in frequency response, distortion, noise) on speech understanding. The method of calculating the AI, as described in the 1969 standard, was summarized by French and Steinberg, colleagues of Fletcher, in 1947. 9 Prior to the development of the AI, quantifying the impact of changes in a telephone circuitry on speech understanding was done via behavioral assessment. This involved substantial speech testing with multiple talkers and listeners and was quite time consuming and expensive.

6 So I was right, wasn't I? The SII and the AI are the same thing.

No, they are similar in that they are based on the same underlying theory that speech intelligibility is directly related to the proportion of audible speech information. However, there are some important differences between the two methods.

7 Are you going to tell me what the differences are?

The primary differences are listed in the foreword of the 1997 standard, but a major difference is that the 1997 standard provides a more general framework for making the calculations than the 1969 version. This framework was designed to allow flexibility in defining the basic input variables (e.g., speech and noise levels, auditory threshold) needed for the calculation. The general framework also allows for flexibility in determining the reference point for your measurements (e.g., free-field or eardrum).

Additional differences include corrections for upward spread of masking and high presentation levels (maybe we can talk more about this later) and the inclusion of useful data such as various frequency importance functions (FIFs). The most notable difference, however, was the name change from AI to SII.

8 Okay, there do seem to be some differences. Can you briefly explain how the SII is calculated?

Let's start with an overview. To calculate the SII, or the AI, we need certain basic information. Specifically, we need frequency-specific information about speech levels, noise levels, auditory threshold, and the importance of speech. In its simplest form, the SII is calculated by determining the proportion of speech information that is audible across a specific number of frequency bands.

To do this you compare the level of speech peaks to either (1) auditory threshold (although you have to account for bandwidth differences between the pure tones used to measure threshold and bands of speech) or (2) the RMS level of the noise (if present), also in frequency-specific bands. The proportion of audible speech, in a frequency region, is then multiplied by the relative importance of that frequency region. Finally, the resulting values are summed across the total number of frequency bands used to make your measures.

9 That seems to make sense, but I think I'm going to need a few more details. But please, no formulas!

Sorry, but I was just getting ready to toss out a formula when you interrupted. I'll try to make it as painless as possible. The general formula for calculating the SII is:

In this formula, the n refers to the number of individual frequency bands used for the computation. The current SII standard is flexible in that you can choose how specific you want your frequency measures (ranging from 6 [octave bandwidth] to 21 [critical bandwidth] bands). Generally speaking, the more frequency-specific your measures, the more accurate your computations.

The I i refers to the importance of a given frequency band ( i ) to speech understanding. The values for I i , also known as the frequency importance function (FIF), are based on specific speech stimuli, and when summed across all bands are equal to approximately 1.0. Again the 1997 S3.5 standard allows flexibility in using the most appropriate FIF for your situation. Figure 2 provides examples of the 1/3-octave band FIFs for two test materials. This figure highlights the substantial differences in terms of band importance that can exist between speech materials.

Finally, the values for A i , or band audibility, which range from 0 to 1, indicate the proportion of speech cues that are audible in a given frequency band. The determination of the A i variable is based simply on the level of the speech, in a given frequency band, relative to the level of noise in that same band. For determining A i a dynamic range of speech of 30 dB is assumed (in both the 1969 and 1997 standards). Using the basic formula for calculating A i we simply subtract the spectrum level of noise from the spectrum level of the speech (in dB) in a given band, add 15 dB (the assumed speech peaks), and divide by 30. Resulting values greater than 1 or less than 0 are set to 1 and 0, respectively. This value essentially provides the proportion of the 30-dB dynamic range of speech that is audible to the listener.

The value A i is then multiplied by the FIF ( I i ) to determine the contribution that frequency region will provide to speech recognition. By summing these values across the various frequency bands a single numerical index (the SII) is obtained.

10 Wait a minute. You mention speech spectrum level and noise spectrum level. What if you are testing in quiet? How does threshold fit in here?

When calculating the SII, we assume that both noise and thresholds function in the same way to limit audibility. The 1997 standard has a conversion factor that is used to convert thresholds (in dB HL) to a hypothetical “internal noise” that would give rise to the measured threshold in quiet.

11 I'm not sure I'm ready to pull out my calculator, but thanks anyway. You've used the term “basic” or “general” formula several times. Is there more to this calculation?

Yes, as I mentioned earlier, the 1997 version of the S3.5 standard includes correction factors in the calculation of the SII that are designed to account for upward spread of masking effects and the negative effects of high presentation levels. These factors may reduce the SII value in a given band and consequently reduce the overall SII for that specific situation.

These factors are important to take into account in certain situations. Although unlikely to play a role in everyday situations, when a lower-frequency, narrowband masking noise is present, upward spread of masking may significantly limit the utility of speech information in higher frequency regions. In contrast, high presentation levels are something experienced by persons without hearing loss and persons wearing hearing aids on a daily basis and can have a significant impact on speech understanding.

12 Okay, I see how the SII can be calculated and even how it could be useful, at least for AT&T. But how is it used in audiology?

Although the scope of the SII standard is limited to individuals without hearing loss, the SII (and its predecessor, the AI) has been an incredibly useful tool for researchers in audiology.

One of the most common complaints of persons with hearing loss is difficulty understanding speech, particularly in noise. The source of this difficulty has been hotly debated for many years. Researchers have tried to determine how much of the difficulty is due to reduced audibility, resulting from hearing loss, and how much is due to factors other than audibility. The SII allows us to quantify the impact hearing loss has on the audibility of speech, both in quiet and in noise. This allows us to “predict” speech understanding in specific test settings and compare the predicted and measured performance of persons with hearing loss.

Using this method, researchers have investigated the impact of factors other than audibility, such as high presentation levels, age, degree and configuration of hearing loss, and cognitive function, on speech understanding. 10,12–15 The results of these studies, and many others like them, have contributed to our basic understanding of how hearing loss interacts with other factors to affect communication function.

13 But what about for people in routine audiologic practice? Are there relatively simple ways to calculate the SII? And how can I use the information in the clinic?

Good questions. Let's take them one at a time. Several investigators have developed relatively simple graphic, paper and pencil, methods of calculating an AI measure. A nice review of these simplified procedures is provided by Amlani et al. 16 These graphic tools, often referred to as “count the dot” methods, use an audiogram format to display both auditory threshold and the dynamic range of speech in dB HL. Figure 3 shows an example of several methods. 2

You can see that the density of dots in the speech spectrum varies with frequency. This variation in density corresponds to the relative importance of speech information in each frequency region (the FIF in the SII calculation). In the Mueller and Killion method, for example, there are a total of 100 dots, so to calculate the AI you simply count the number of dots that are above threshold and divide by 100. Thus, if 30 dots are above threshold, then the AI for that individual, listening in quiet, is 0.3.

14 Wait a minute! Do you mean I can do the math you described earlier or I can just count the dots and I'll get the same answer?

It's not likely that you would get the same answer , although you may be pretty close in some situations. Recall that the ANSI S3.5 method provides a general framework and allows room for varying many of the input parameters used in the calculation, such as speech level, noise level, and the type of speech materials. The input parameters, other than auditory threshold, are fixed in these simplified methods. For example, the Mueller and Killion method assumes the speech is presented in quiet, at a conversational level, and has a frequency importance function representative of nonsense syllables. If all these assumptions are incorporated in the ANSI S3.5 (1997) method, then the two methods will produce similar results. However, for situations that deviate from these assumptions, differences between methods could be quite large.

15 Okay, I can count the dots, but how might I use this information in the clinic?

Well, methods for calculating audible speech information can be useful clinically in several ways. Mueller and Hall listed six clinical applications of AI type measures. 17 The following is a summary of their suggestions:

Perhaps the most obvious uses are in determining candidacy and in patient counseling and education. For example, individuals with very high unaided AI measures are unlikely to show large aided benefit, at least for conversational speech inputs. At the same time, it is unclear how “high” an unaided AI is “high enough.” Self-assessment questionnaires, however, could also help identify borderline candidates. In terms of counseling, a visual image of audibility (or lack of) provides a powerful tool for patients and significant others to drive home the point that hearing loss can impact speech audibility and speech understanding.

AI measures may also be helpful in making circuit selections or for comparisons across hearing aids. For example, the AI can help quantify the potential improvement in speech audibility as you change hearing aid circuits or instruments. It is important to remember, however, that increasing the AI does not always mean speech understanding will increase, particularly for persons with substantial hearing loss (more about that later).

The AI also provides an objective measure of hearing aid benefit. Although inadequate in isolation, documenting improved audibility in conjunction with other measures of hearing aid benefit (such as subjective assessments) is important in these days of managed healthcare.

Finally, Mueller and Hall suggest that AI measures could be useful in educating referral sources. Information about speech-understanding abilities provided to referral sources typically comes from the diagnostic battery and consists of word-recognition scores based on monosyllabic word testing done at high levels. However, reporting an unaided AI measure based on conversational and soft speech is another method that may be useful when describing hearing handicap for speech to referral sources.

16 Doesn't my probe-microphone system provide unaided and aided AI information too?

Yes, some systems do calculate a version of the AI, but to my knowledge none actually implement the 1997 version of the ANSI S3.5 standard. In fact, in many cases it is unclear what input parameters, or what “method” the software is basing its calculations on. It's very possible that your probe-mic AI calculations would differ from the SII or from your favorite pencil-and-paper AI calculations.

17 Does that mean the values they provide are useless?

No, the information can be very useful in making decisions regarding hearing aid adjustments. For example, your patient has returned for a follow-up session and reports that, despite your perfect match to your prescriptive target, she is continuing to have substantial difficulties understanding speech at work. Based on your AI measures you determine that your client's aided AI value is still substantially below 1.0 (e.g., ∼0.55). (It is important to note that this will actually be the case for many individuals with moderate hearing loss, even after perfectly matching some prescriptive targets.) The client is not complaining of tolerance issues but rather of speech-understanding difficulties. This may well be a case where gain for soft to moderate sounds, particularly in the high-frequency regions, could be increased to improve audibility and, potentially, speech understanding.

That said, the usefulness of the AI values provided by your probe-microphone system will also depend on your reasons for obtaining them in the first place. In most cases, we are simply looking to verify a change in the AI between the unaided and aided conditions. Because we are comparing data obtained using the same calculation method and are not trying to actually predict speech recognition, this relative comparison is quite valid. If, however, your needs are more precise, then a separate implementation of the ANSI S3.5 1997 SII calculation would be more appropriate.

18 Let me get this straight. It sounds as if all we need to do to maximize speech understanding is crank up the hearing aid gain until we get an AI/SII of 1.0. Is it really that simple?

Unfortunately, no. Several factors make this approach unreasonable. For one, people with cochlear pathology usually have LDLs at or near normal levels, resulting in very narrow dynamic ranges, usually in the high frequencies. This limits the maximum amount of gain, particularly for linear devices, that can be applied to soft sounds without making louder sounds uncomfortably loud.

Recall that speech understanding actually decreases at high presentation levels, particularly in noise. Several researchers have reported that maximizing the AI does not necessarily improve speech understanding compared with other prescriptive procedures (e.g., the NAL) and in some cases may even result in poorer performance. 18,19

Finally, research suggests that the gain-frequency response that provides optimal speech understanding may not always be preferred by listeners in terms of optimal sound quality. 20 Clearly, in most cases, simply turning up hearing aid gain to achieve AIs of 1.0 is not appropriate. An exception might be in cases of mild hearing loss when WDRC, in conjunction with appropriate frequency shaping, is used.

19 You mentioned earlier that there were programs available that helped in the calculation of the SII?

Yes, a useful web site for those interested in the standard is http://www.sii.to . The site was created and is maintained by members of the Acoustical Society of America (ASA) Working Group S3–79. This working group, which is in charge of reviewing the ANSI S3.5–1997 standard, has provided access to several computer software programs to aid in calculating the SII. It is important to note, though, that the programs are not a part of the standard and are provided by the developers on an “as is” basis. The site also contains errata to the 1997 standard and describes typographical errors in the existing standard.

20 I know this is my last question, so tell me one more time. When I'm seeing my patients in the weeks to come, when would I use the SII and when would I use the old AI?

Clinically, you are most likely to obtain AI (rather than SII) type measures using a “count the dots” method or from your probe-mic system. The information provided by either of these methods is quite adequate for clinical uses. In addition, calculating the complete SII (i.e., incorporating all correction factors) would be tedious without the use of computer programs and, to my knowledge, these programs are not yet incorporated in current audiologic equipment. I would, however, expect to see this in the near future.

Section Description

The Articulation Index (AI) has been with us for many years. No, it has nothing to do with speech production—it's about hearing. For simplicity, think of it as an “audibility index.”

For many years the AI rarely found its way outside of research laboratories. This probably was because the calculations were more complicated than the busy clinician cared to tackle. Then, in 1988, Chas Pavlovic introduced a clinically friendly version of the AI—equal weighting for 500, 1000, 2000, and 4000 Hz. The math simply required adding together four key numbers and then dividing by 120. Not bad.

But soon, things got even simpler as the Mueller-Killion count-the-dot audiogram was introduced. Now, the ability to count to 100 was all that was required. This was soon followed by the Humes 33-bigger-dot version, which was followed by the Pavlovic 100-square version, which was followed by the Lundeen 100-dot modification of the Pavlovic squares. Soon, manufacturers of probe-mic equipment included automated AI calculations, which encouraged even greater use of audibility considerations in the selection, fitting, and adjustment of hearing aids.

But then, in 1997, just when we all were getting comfortable with the AI, a new calculation method was introduced—the Speech Intelligibility Index or SII. This has led to questions such as: Did the SII replace the AI? How does the SII differ from the AI? Is it still okay to use the AI?

To help us make sense of all this, we've brought in a guest author who has used both the AI and the SII, in the clinic and in the research lab. Benjamin W.Y. Hornsby , PhD, is a research assistant professor at Vanderbilt University. Although Ben has logged many years as a clinician, these days he spends most of his time in the Dan Maddox Hearing Aid Research Lab. You're probably familiar with many of his recent publications. To assure you that Dr. Hornsby is not a single-minded kind of guy, prior to his interest in audiology he worked as a welder, rock climbing instructor, sign language interpreter, and junior high school science teacher.

After reading Ben's excellent review I think you'll have a better understanding concerning the similarities and differences of the SII and the AI, and how these measures fit into your daily practice. And yes, “articulation” is about hearing, not speech production.

• + Favorites
• View in Gallery

Readers Of this Article Also Read

Building blocks: speech intelligibility index: no magic number, but a..., marketrak viii patients report improved quality of life with hearing aid usage, new</em> count-the-dots method', 'killion mead c.; mueller, h. gustav', 'the hearing journal', 'january 2010', '63', '1' , 'p 10,12-14,16-17');" onmouseout="javascript:tooltip_mouseout()" class="ejp-uc__article-title-link">twenty years later: a new count-the-dots method, speech-in-noise tests: how and why to include them in your basic test battery, the frequency pattern test: a guide.

Look up a word, learn it forever.

Speech intelligibility, /spitʃ ɪnˈtɛlədʒəˌbɪlədi/.

• noun the intelligibility of speech (usually measured in the presence of noise or distortion) see more see less type of: intelligibility the quality of language that is comprehensible

Sign up now (it’s free!)

Whether you’re a teacher or a learner, vocabulary.com can put you or your class on the path to systematic vocabulary improvement..

SPEECH THERAPY

• May 20, 2020

Speech Intelligibility: How clear is your child’s speech?

Updated: Nov 25, 2020

Have you ever found it difficult to understand your child’s speech? How they articulate sounds impacts their overall clarity in speech. A Speech Language Pathologist (SLP) can help you understand if your child’s speech intelligibility is at an appropriate level for their age.

What is Speech Intelligibility?

Speech intelligibility refers to how well someone can be understood when they’re speaking. The factors that determine someone’s clarity include knowing what sounds to make and how to pronounce them, the rhythm (also known as prosody ), the volume, staying on topic, formulating syntax, and more. There’s a lot that goes into our conversations without us even realizing it!

How is it Measured?

There are a variety of methods SLPs use to measure a client’s speech intelligibility. Some use their trained experience to listen and estimate the percentage of speech that can be understood. A more formalized method involves:

1) recording the client saying a list of pre-selected words/sentences

2) providing a second person with a list of similar sounding words

3) the second person will listen to the recordings and select from the list which word they think was being said.

This helps eliminate the bias that family members can have since they hear the client speak every day and might understand them better than a stranger would.

When Should My Child Be Fully Intelligible?

Ideally, as children age, their speech intelligibility should be increasing. According to data presented at the 2003 American Speech-Language-Hearing Association convention, the typical norms you want to look for in a child are:

26 - 50% intelligible by age 2

75% intelligible by age 3

90% intelligible by age 4

By age 5, a child following the typical development norms should be 100% intelligible. Errors in pronunciation can still occur, but this just means that a stranger should have no problem understanding what the child is trying to say.

What Causes Difficulties with Speech Intelligibility?

Of course, things like background noise or hearing difficulties on the listener’s end can affect how well someone is understood. However, there are some conditions that can impact someone’s ability to communicate clearly and be understood. Here are some of the most common ones:

Childhood Apraxia of Speech / Developmental Apraxia of Speech

The brain has issues with planning movement of the lips, jaw and tongue

Muscle paralysis or weakness causing slurred speech or other issues with intelligibility

Speech Sound Disorders

Replacing certain sounds, distorting them, or omitting them altogether; can be caused by hearing loss, dental abnormalities, cleft palate, etc.

How Can It Improve?

An SLP will assess a client’s situation and needs in order to determine what type of therapy is needed and how it can improve their situation. Typically, without any other underlying cause for low speech intelligibility, speech therapy itself can help the individual learn how to improve their speech.

When there is an underlying cause to the reduced speech clarity or a delay in improvement, other supportive measures can be put in place, such as:

Incorporating sign language, hand gestures or facial cues in speech

Learning to control pace and breathing

Introducing AAC devices

Working on volume and resonance

Developing speaker-listener communication strategies -- providing feedback or asking questions, looking for facial cues, maintaining eye contact, etc.

If you’re having any trouble understanding your child and want to schedule an assessment with an SLP, we are still providing teletherapy during these times of social distancing -- feel free to contact us with any questions!

Dolgin, E. (2019, May 11). Dysarthria. Retrieved from https://www.apraxia-kids.org/apraxia_kids_library/dysarthria/

Keintz, C., Hustad, K., Garcia, J., & Klasner, E. (n.d.). Speech Intelligibility: Clinical Treatment Approaches for Children and Adults.

McCleod, S., & Bleile, K. (2003). Neurological and developmental foundations of speech acquisition . American Speech-Language-Hearing Association Convention . Retrieved from http://www.speech-language-therapy.com/pdf/docs/ASHA03McLeodBleile.pdf

Scanlon, K. (n.d.). Is Your Child Intelligible? Retrieved from https://www.scanlonspeech.com/2012/07/05/is-your-child-intelligible/

Andalusia Speech Therapy has two Toronto speech therapy clinics and offers speech teletherapy to anywhere in the world. Contact us more for information.

• Speech delay
• Late Talkers

Recent Posts

Does my child have a stutter?

Lockdown and Language – What are the Effects on Children's Development?

Ages 1-5 Speech Milestones – Language Development Norms

Is Your Child Intelligible?

Is it challenging to understand what your child is saying?

Highly unintelligible speech can significantly impact functional communication, social participation and educational achievement.

Definition of Intelligibility

Intelligibility can be defined as “the extent to which an acoustic signal, generated by a speaker, can be correctly recovered by a listener” (Kent, Weismer, Kent, & Rosenbek, 1989).

In other words, intelligibility is how well the speaker’s speech is understood by the listener. Two parties are involved in the measurement of intelligibility.

Intelligibility is typically measured by parents via surveys and questionnaires. For instance, The Intelligibility in Context is a quick parent report that many speech language pathologists use to obtain information on the child’s perceived intelligibility. It asks parents to rate the degree, on a 5-point scale, to which their children’s speech is understood by 7 different communication partners (parents, immediate family, extended family, friends, acquaintances, teachers, and strangers).

However, parents are familiar listeners and usually over-estimate their children’s intelligibility because they are keen on listening to and understanding their children. This is necessary to acknowledge because a parent’s rating of their child’s intelligibility may not match that of a stranger. People need to be understood by all listeners – familiar and unfamiliar.

When Are Children 100% Intelligible?

Intelligibility norms tend to vary. The following information has been adapted from ASHA’s comprehensive summary  – Neurological and developmental foundations of speech acquisition.  It may help you in deciding whether or not to seek the assistance of an experienced and certified speech language pathologist:

By 2 years old = 26-50% of children are intelligible (Weiss, 1982)

By 2 years 6 months old = 51-70% of children are intelligible (Weiss, 1982)

By 3 years old = 71-80% of children are intelligible (Weiss, 1982)

Additional research reveals that by 3 years old  = 73% (or, 50-80%) of children are intelligible when judged by three unfamiliar listeners. (Vihman, 1988)

It should also be noted that according to Vihman,  “…children who used more complex sentences were more difficult to understand (1988).”

By 4 years old 93% (or, 73-100%) of children are intelligible in conversational speech with unfamiliar listeners (Gordon-Brannan, 1993 cited in Gordon-Brannan, 1994).

If the above information is too hard to remember, here’s a good GENERAL rule of thumb:

• Most 2 year olds are at least 50% intelligible
• Most 3 year olds are at least 75% intelligible
• *Most 4 year olds are approximately 100% intelligible

*This does not mean that they are saying every sound correctly. A speaker can still have articulation errors or other speech subsystems errors and still be intelligible.

There are 5 speech subsystems:

• respiration (e.g. volume – loudness vs. softness of speech)
• phonation (e.g. breathy or hoarse vocal quality and vocal fry)
• articulation (e.g. the production of clear and distinct sounds in speech)
• resonance (e.g. hypernasality/hyponasality, and cul-de-sac resonance)
• prosody (e.g. the patterns of stress and intonation in speech – monotone).

Source – https://www.asha.org/practice-portal/clinical-topics/dysarthria-in-adults/

According to an online CEU course I recently completed titled, Assessments and Differential Diagnosis of Speech Sound Disorders, Dr. Lynn Williams, Ph.D., talks about a critical age hypotheses for remediating unintelligible speech. According to research by Bishop and Adams (1990), she explains that

“Unintelligible speech must be resolved by age five or six in order to significantly reduce academic problems…” She goes on to say that unintelligible speech in the early years may affect literacy development.

Since most unintelligible children do not seek speech therapy treatment until approximately 4 years of age, speech therapy must be efficient and effective at remediating the speech disorder.

What Causes Difficulty with Intelligibility?

Some parents often wonder why their child is unintelligible.

The following can negatively impact intelligibility:

• Speech sound disorders or active phonological processes
• Imprecise articulation
• Hearing loss
• Fluctuating fluency and stuttering
• Fast or slow rate of speech
• Dysarthrias related to certain medical conditions (e.g. Down Syndrome and Cerebral Palsy)

According to Bishop (2010), 10% of children have developmental speech sound disorders.

Children with decreased intelligibility secondary to speech sound disorders typically benefit from traditional articulation therapy that aims to correct the incorrect speech sounds.

What Can You Do to Help Increase Your Child’s Intelligibility?

• Have your child’s speech formally evaluated by a speech language pathologist
• Have your child’s hearing tested by an audiologist
• Model clear speech and over-articulate your words

Hopefully, this post has been helpful!

The Mega Bundle for Articulation Therapy is a must-have for your speech therapy closet!

Check it out below:

• Enhance Your Speech Therapy Sessions with ‘Picky Kitty’ – A Must-Have Game for Every Speech Therapist
• Same and Different: Why Speech Therapists and Educators Should Target this Basic Concept
• New Beginnings for Scanlon Speech Therapy
• Articulation Speech Therapy: Favorite Activities, Games, and Books to Target the TH Sounds
• Articulation Speech Therapy: Favorite Activities, Games and Books to Target the L Sound

Similar Articles

Join my email list to access my resource library.

Sign Up For Our E-Blast To Receive Information on our Books, Speech Therapy materials and our latest freebies!

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Karger Author's Choice

Speech Characteristics and Intelligibility in Adults with Mild and Moderate Intellectual Disabilities

Marjolein c. coppens-hofman.

a Department of Medical Psychology and NCEBP, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

Hayo Terband

b Utrecht Institute of Linguistics OTS, Utrecht University, Utrecht, The Netherlands

c Centre for Language and Cognition (CLCG), University of Groningen, Groningen, The Netherlands

Ad F.M. Snik

d Department of Otorhinolaryngology and Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

Ben A.M. Maassen

Adults with intellectual disabilities (ID) often show reduced speech intelligibility, which affects their social interaction skills. This study aims to establish the main predictors of this reduced intelligibility in order to ultimately optimise management.

Spontaneous speech and picture naming tasks were recorded in 36 adults with mild or moderate ID. Twenty-five naïve listeners rated the intelligibility of the spontaneous speech samples. Performance on the picture-naming task was analysed by means of a phonological error analysis based on expert transcriptions.

The transcription analyses showed that the phonemic and syllabic inventories of the speakers were complete. However, multiple errors at the phonemic and syllabic level were found. The frequencies of specific types of errors were related to intelligibility and quality ratings.

Conclusions

The development of the phonemic and syllabic repertoire appears to be completed in adults with mild-to-moderate ID. The charted speech difficulties can be interpreted to indicate speech motor control and planning difficulties. These findings may aid the development of diagnostic tests and speech therapies aimed at improving speech intelligibility in this specific group.

Introduction

Communication is an important aspect of quality of life, and speech is the primary means of human communication. In adults with intellectual disabilities (ID), speech communication is often troubled by disordered speech production and/or impaired hearing [ 1 ], resulting in miscommunication and consequently impairing social interactions, possibly behavioural problems and isolation.

Speech intelligibility can be defined as how clearly a person speaks so that his or her speech is comprehensible to a listener [ 2 ]. Reduced speech intelligibility leads to misunderstanding, frustration, and loss of interest by communication partners. As a result, communication decreases or remains at a low level. Thus, to improve the quality of life of adults with ID, it is essential to enable them to make themselves understood. Impairment of speech production is among the most commonly reported difficulties in children, adolescents and adults with ID [ 3 ]. The deficiencies are not resolved when growing up and speech intelligibility remains a problem throughout life [ 4 ]. Yet, still little is known about the predictors of the reduced speech intelligibility in adults with ID.

Speech Intelligibility

Speech intelligibility is usually approached as word or utterance recognition in natural communication situations [ 5 ]. The intelligibility of a spoken message is influenced by a number of factors. Above all, intelligibility is a joint product of the speaker and the listener. Listener factors are familiarity with the speaker and with the topic of conversation [ 6 , 7 ]. Familiarity with the speaker is especially helpful when speech is disordered or poorly intelligible due to speech errors [ 6 , 7 ]. In addition, intelligibility varies with the nature of the spoken material (e.g., linguistic structure and length of utterance) and the context of communication (e.g., the quality of the acoustic transmission of the speech signal, the availability of visual cues from the speaker, and contextual support for the message to be transmitted). A gold standard for speech intelligibility measurement is not available, but it can be argued that a standardised speech intelligibility assessment can be obtained by making audio-recordings of spontaneous speech samples in an interview situation where the proband is asked to tell about neutral, everyday situations. These audio-recordings are presented to naïve listeners, who either produce transcriptions or give intelligibility judgments (for an overview of methods of standardisation, see [ 7 ]).

Speech Difficulties in Adults with ID

Speech characteristics have not been studied widely in adults with ID in general, but Roberts et al. [ 1 ] documented the following characteristics of the speech of adults with Down syndrome: consonant cluster reduction; final consonant deletion; unstressed syllable omission, mostly at the start of a multisyllabic word; and consonant substitution (e.g., a fricative sound /s/ becomes a plosive /t/). Errors in the speech of persons with Down syndrome have been characterised to be inconsistent [ 1 , 8 , 9 ], and the production errors and articulation difficulties observed mainly occur in the phonemes that are typically acquired in the final stages of normal speech-language development [ 10 , 11 , 12 ]. Van Borsel attributed the errors to a developmental delay [ 10 ]. Unfortunately, no data are available for adults with ID of mixed aetiology or other groups of persons with a specific ID.

The present study set out to examine the common difficulties in the speech of adults with ID of mixed aetiology, thereby focusing on the main predictors of the reduced intelligibility of their speech. Objectified measures that form a strong predictor of intelligibility may be used to evaluate treatment outcomes, and the observed speech irregularities may become therapeutic targets. To this end, a quantitative analysis of standardised speech samples elicited in a picture naming task was made resulting in inventories of segmental (phonemic) and syllabic error characteristics. In addition, intelligibility of spontaneous speech samples was judged by naïve listeners such that the relevance of the quantified speech characteristics in determining reduced speech intelligibility could be assessed. An additional factor to consider as a potential cause of reduced intelligibility is poor auditory feedback due to deficient auditory processing in combination with chronic hearing difficulties [ 1 ]. In the current study, we took hearing difficulties into account on a functional level by measuring the auditory discrimination abilities of all our participants, and by analysing any relation with speech difficulties. Finally, severity of ID and aetiology were included in the analysis.

This study was approved by the Ethics Committee of the Radboud University Nijmegen Medical Centre. Potential participants and their legal representatives received written and oral information on the study, after which oral and written consent was obtained from the legal representatives, parents, caregivers, and participants.

Participants

Inclusion was based on the level of ID, i.e. IQ 40–70 (DSM- IV – Mild and Moderate). The 36 participants that entered the study (19 men and 17 women aged 18–40 years) had all been identified as poorly intelligible by their caregivers and relatives. Their speech problems had not been assessed before by diagnostic tests and the cause of the reduced intelligibility was unclear. As to ID aetiologies, 11 of the 36 participants had Down syndrome, 1 participant had Fragile-X syndrome, and another Turner syndrome, 3 participants had sustained brain damage in the first year of life, 2 had a chromosomal deficiency (microcephaly), 1 had suffered trauma, and another 2 had suffered hypoxia during birth, while during the study 2 participants were diagnosed with a specific genetic microdeletion [ 13 ]; in 12 cases, the cause of the ID was unknown. Sixteen of the participants were classified as having a mild ID (IQ 55-70) and 20 as having a moderate ID (IQ 40-55). For the analysis of a possible effect of ID aetiology, the aetiologies with a small ( n < 10) number of cases were aggregated, thus creating 4 groups (unknown [ n = 12]; Down syndrome [ n = 11]; other syndromes [Fragile X, Turner syndrome, microcephaly, genetic microdeletion; n = 6]; acquired ID [oxygen deficiency, brain damage, electrographic status epilepticus of sleep, trauma; n = 7]).

Data Collection

Recordings of spontaneous speech production and a picture-naming task were obtained from all participants. To determine the participants' vocabulary level and test their understanding of the instructions, all participants took the Peabody Vocabulary Test [ 14 ] 1 week prior to the individual recording sessions. Participants were thus also able to familiarise themselves with the researcher (first author), the recording equipment, and setting.

In testing situations, people with ID often speak in subdued tones or use explosive voice modulations – vocal characteristics which can severely limit the quality of the speech signal in both live-voice and tape-recorded presentations. Special care was therefore given to optimise the speech signals in our samples. Some of the participants indeed tended to speak quietly, while in others vocal intensity varied from utterance to utterance, all prosodic variations that could confound the goals of the assessment to some extent. In these cases, feedback was offered prior to the actual recording to reinforce or modify the participant's speech efforts. Moreover, all speech recordings were made in a quiet, familiar room in the participant's own care or residential facility by the same researcher (first author), who was seated at the same table opposite the participant, allowing eye contact with the participant. Also, a silent observer was present during each recording session, which mostly involved a parent, the primary caregiver or a person close to the participant. In no case was the participant asked to come to an unknown place or a clinical setting.

The first recording contained 3–10 min of spontaneous speech in response to open questions about the participants' hobby's and daily pastime. As it is appropriate to use a closed-set evaluation for people with ID using items within their linguistic competence, a second recording was made using Logo-Art [ 15 ]. This Dutch picture-naming test consists of 128 easily recognisable pictures that represent words of everyday life and was developed to assess articulation skills in children aged 4–8 years. The target words included all vowels, diphthongs, consonants and consonant clusters used in the Dutch language in all word positions (initial, medial, and final).

A professional solid-state recorder (Marantz PMD620) was used to obtain digital speech samples. As several participants found the external microphone threatening or highly distracting, we placed the internal microphone at approximately 40 cm distance from the speaker's mouth. All recordings were made in the same way using a bit rate of 705 kbps and a frequency of 44.1 kHz. The duration of the recordings varied depending on the participant's movements, in-between conversation, pace of responding and speech rate. No participant was rushed at any time. The Logo-Art recordings took 20–40 min.

Data Processing

Spontaneous Speech

Of the 36 participants, 34 produced a sufficient amount of spontaneous speech for an intelligibility judgment. Two participants were excluded from the assessment as they only used one-word utterances, consistently repeating this one word.

Two relevant, continuous segments were selected from each of the 34 speech samples by use of the PRAAT software programme [ 16 ] based on auditory and visual cues in the recorded speech, yielding a total of 68 segments to be judged. The segments each contained a total of 2 min of uninterrupted sentences or ongoing speech. These were assessed for intelligibility by 25 speech and language pathology students in their final year of training. All students resided in the same socio-geographical environment as the participants. Before the speech assessments, the students listened to 3 random speech samples of non-participants (similarly aged adults with mild or moderate ID) to familiarise themselves with the speech recordings, the rating method and overall setting. All speech samples were presented through loudspeakers at normal loudness in a quiet room. The students were asked to individually rate speech intelligibility on a 5-point scale, with 1 denoting essentially unintelligible , 3 intelligible at times , and 5 essentially intelligible . The mean scores of the 25 listeners were taken to indicate the intelligibility of a participant's speech.

Picture-Naming Task and Transcription

Picture-naming performance on the Logo-Art, in terms of the words produced in response to the presented images, was transcribed in broad phonetic alphabet (IPA) and keyed into the Logical International Phonetics Program (LIPP) transcription system [ 17 ]. This computer-based transcription system allows transcribed utterances to be analysed with respect to their phonetic characteristics, e.g. by providing an inventory of phonemes and syllable structures, and compared to target utterances, yielding variables such as percentage of consonants correct (PCC) and number of cluster reductions (for details, see below).

The participants recognised and correctly named more than 85% of the Logo-Art pictures. Utterances produced while the participant was chewing or had fingers or objects in or over the mouth, as well as yells, grunts, and coughs were excluded. All recordings were independently transcribed by 2 professional transcribers (both speech language pathologists) who did not know the participants and were not informed about ID nature and cause. Transcription reliability was assessed by comparing multiple transcriptions of the same utterances. Mean inter-rater segment-by-segment correspondence in all transcriptions was 94% and higher.

Analysis of Logo-Art Transcripts

Comparison of produced and target utterances was conducted at the segmental and syllable-structure level. Analyses at the segmental level concerned the identity of the segments and yielded 2 types of variables: proportions of consonants correct (PCC; both overall and separated out for different developmental complexity levels [ 18 ]), and proportions of substitutions (overall and in relation to syllable position). The substitutions were further classified as typical or atypical phonological processes [ 18 ]. Analyses at the syllabic level concerned the structure of the syllables only, not the identity of the phonemes within the syllable. The syllable structures that were evaluated were: V (vowel); CV (consonant-vowel); VC; VCC; CVC; CCVC; CVCC; CCVCC; CCCVC; and CCCVCC. For each of these structures, the percentage correct was computed. The resulting variables are detailed in Table ​ Table1 1 .

Overview of the variables determined by the segmental and syllable structure comparison of target word and produced utterance

Statistical Analysis

Data analysis was layered. First, we explored possible influences of the different between-subject factors, hearing loss, severity of ID, and ID aetiology group, on the intelligibility scores as well as the general phonological measures using a series of analyses of variance.

The second step consisted of a more in-depth analysis of the specific types of errors and phonological processes, including an assessment of the different stages in phonological acquisition (according to the Phonological Analysis of Dutch [ 18 ]).

Finally, multiple linear regression analyses were carried out to identify which (subset of) the general phonological measures could reliably be used as a predictor of speech intelligibility. The regression analysis featured the stepwise method with backward elimination with a removal criterion of 0.1 based on the F statistic.

Significance level was set at p < 0.05, while p values <0.10 were qualified as trends. Homogeneity of variance was tested with the Levene test of homogeneity, and the Mauchly test of sphericity was applied.

General Factors

First, the influence of the different between-subject factors on the main speech parameters was investigated using a series of univariate analyses of variance with the intelligibility scores, as well as the phonological measures (Table ​ (Table1) 1 ) as dependent variables, and hearing loss, severity of ID, and ID aetiology group as fixed factors. The analyses yielded a significant effect of severity of ID on intelligibility ( F (1, 33) = 8.60, p < 0.01), as well as trend effects for this factor on the proportions syllable structures correct (PSSC; F (1, 33) = 3.14, p = 0.09) and consonant deletion (PCDel; F (1, 33) = 3.20, p = 0.09). No further effects or interactions were observed. Therefore, the data were collapsed over the factors hearing loss, and ID aetiology.

In a subsequent multivariate analysis of variance on the differences in intelligibility scores and phonological measures, with severity of ID only, this latter factor was significant ( F (12, 21) = 3.19, p < 0.01). Mean intelligibility scores were higher in the participants with mild ID (mean 3.32; SD 0.77; range 2.15-4.61) as compared to the moderate group (mean 2.44, SD 0.69; range 1.32-3.60). The outcomes on the general phonological measures are presented in Figures ​ Figures1 1 and ​ and2. 2 . Univariate analysis showed a trend of a higher proportion of correct consonants in syllable-initial position (PCCI) in the mild ID as compared to the moderate group ( F (1, 33) = 3.510, p = 0.07; Fig. ​ Fig.1). 1 ). Furthermore, the participants with moderate ID made significantly more substitutions of single consonants in syllable initial position (PSubCI; F (1, 33) = 5.097, p < 0.05) and deletions of single initial consonants (PIC1Del; F (1, 33) = 5.012, p < 0.05) than their peers with mild ID (Fig. ​ (Fig.2 2 ).

Mean scores on the 3 measures of phonological accuracy: proportion consonants correct in syllable-initial position (PCCI), proportion syllable structures correct (PSSC), proportion consonant clusters correct in initial position (PClusCI). Error bars indicate 95% CI.

Mean scores on the 7 phonological error measures: proportion substitutions of single consonants in initial position (PSubCI), proportion substitutions of single consonants in syllable-final position (PSubCF), proportion abnormal substitution processes (PAbnProc; h-sation, abnormal stopping, backing, and denasalisation), proportion normal substitution processes (PNormProc; fronting, stopping of fricatives, nasalisation, and gliding), proportion consonant deletions (PCDel), proportion deletion of consonants in syllable-initial position (PIC1Del), and the proportion reduction of consonant clusters in syllable-initial position containing 2 and 3 consonants (PIC2Red, and PIC3Red). Error bars indicate 95% CI.

The main error types concern the pronunciation of single and clustered initial consonants as reflected by the high frequencies of single consonant deletions and cluster reductions in syllable-initial position. In order to gain more insight into the processes underlying the production difficulties, we conducted further phonological analyses. Our first query was whether the error patterns were similar to the phonological delay observed in otherwise unimpaired children. To this end, we compared the phonemic inventory of our sample with the patterns described for typically developing children during speech acquisition.

Phonemic Inventory

Beers developed a system to analyse phonological development, called the Phonological Analysis of Dutch [ 18 ]. She found that the completeness of the phonemic inventory can be assessed in 5 levels of complexity. An incomplete inventory typically results in a pattern in which the higher levels are produced at a lower percentage accuracy. The speakers with mild ID tended to produce higher proportions consonants correct than the speakers with moderate ID (Table ​ (Table2). 2 ). The mean percentages correct for all complexity levels were above 75%, indicating that, overall, the phonological repertoire is complete and that no systematic pattern can be discerned that reflects a phonological delay. Those participants who produced lower percentages correct did not show a declining tendency at the higher complexity levels; thus, these participants make errors irrespective of phonological complexity.

Mean proportions and standard deviations (SD) of correctly produced consonants according to developmental levels of complexity [ 17 ]

L1CI–L5CI, level of complexity 1 – 5 of consonants in syllable initial position.

Typical and Atypical Speech Processes

Young children and (young) adults with speech difficulties may produce errors that affect entire classes of sounds rather than individual sounds. At a particular age, these so-called phonological processes (a detailed description is provided in Table ​ Table1) 1 ) are a normal, natural part of their speech development and therefore denoted as typical processes. A speech profile consisting of processes that are typical for younger children can be interpreted as speech delay. Thus, consonant deletion in syllable-final position, fronting, consonant-cluster reduction and stopping are considered typical speech processes.

In contrast, atypical speech processes comprise classes of errors that are not normal during any stage of speech development and are therefore taken to indicate speech pathology. The atypical processes we analysed are: h-sation, abnormal stopping, backing, and denasalisation; all in syllable-initial position. The results are presented in Table ​ Table3 3 (see also Fig. ​ Fig.2). 2 ). Our analyses revealed 2 dominant atypical speech processes, irrespective of severity of ID. We first found a high proportion of h-sations and, secondly, a high proportion of denasalisations.

Means and standard deviations (SD) of proportions of atypical phonological processes

To summarise, the number of errors in the speech production of our sample was high. For most participants, the phoneme repertoire appears to be well-completed. However, on average, the participants showed high error rates across complexity levels, and many typical as well as atypical errors in their speech.

Predictors of Intelligibility

But which of the deviations in the speech of our participants with ID account for its reduced intelligibility? The clinical relevance, of course, is that since phonemic quality contributes to intelligibility, the observed irregularities may become therapeutic targets. Furthermore, phonological error measures that form a strong predictor of intelligibility may be used to evaluate treatment outcomes. Multiple regression analyses (method: stepwise with backward elimination) were conducted with intelligibility – the students' ratings of spontaneous speech – as the dependent variable and the phonological error measures of the Logo-Art transcription analysis as the independent variables.

For the mild ID group, the model with PCCI, PIC3Red, PNormProc and PClusCI showed up in the regression analysis ( R 2 = 0.790, adj R 2 = 0.706, standard error = 0.42, p = 0.002). These 4 measures together accounted for 79% of the variance in intelligibility in this group. For the moderate ID group the model with PCCI, PIC2Red, and PClusCI showed that these 3 measures together were responsible for 69% of the variance ( R 2 = 0.693, adj R 2 = 0.631, standard error = 0.42, p = 0.000).

The participants with mild and moderate ID of the present study showed distinct difficulties in their speech production that affect both the quality and intelligibility of their verbal output. Their speech is characterised by an overall high error rate and the occurrence of both typical and atypical phonological processes; the percentages of correct productions at all levels of phonemic complexity showed a non-declining pattern, indicating deviant speech development rather than an incomplete phonemic inventory. We found no significant difference between our speech parameters based on hearing loss, or ID aetiology. The severity of ID (mild vs. moderate), however, did show a strong association with speech intelligibility and was also related to particular error frequencies.

The phonological processes cannot be explained by weakness or paralysis of the speech muscles or other sensorimotor deficits alone. Kumin [ 3 ] documented that symptoms of ‘childhood verbal apraxia’ could be found in children with Down syndrome. Among the symptoms of apraxia that were identified [ 3 ], we observed: inconsistency in phoneme (sound) productions, difficulties in combining and sequencing phonemes (like in consonant clusters), and speech-rhythm problems; that is, as has been reported for Down syndrome, speech patterns in adults with ID show many errors not seen in typical development.

To our knowledge, we are the first to analyse prompted and unprompted speech of adults with various cognitive impairments. The participants were selected on the basis of their reduced speech intelligibility and their wish to improve their verbal communication skills but had never been tested for or diagnosed with any speech disorders. Our results showed the participants to have a complete phonemic and syllabic repertoire and also showed distinctive processes in their speech production. It is worth mentioning that our speech samples were recorded in a quiet room, with a microphone positioned close to the speaker, causing the effects of low speaking volume and other voice characteristics on intelligibility to be relatively small. These are likely to be far more detrimental when communication takes place in a busy and noisy environment with poor acoustics (such as echo resulting from safety and hygiene regulations) that characterises many residential and work facilities for people with ID.

Another issue to consider in this context is that short-term and long-term verbal memory are both highly involved in speech production and that the 2 systems are impaired in people with ID [ 19 , 20 , 21 ]. From a psycholinguistic perspective, a distinction can be made between representations stored in memory and the transformation and execution processes that utilise these representations. Word production and phonological planning skills must first be acquired and then become automated through practice. Short-term memory plays an important role in these ongoing processes of planning, sequencing and coordination of speech movements. Long-term memory is essential for the storage of word forms; deficits in long-term memory therefore can account for the persistence of the speech difficulties into adulthood.

Based on our results, we propose that the deviant speech characteristics in this population should be targeted in tailored therapeutic management schemes, and that they can serve as measures to evaluate treatment results. Speech skills can improve through dedicated training under the guidance of speech language pathologists specialised in working with people with cognitive deficits. Continuously providing feedback in understandable terms and offering and demonstrating the right learning processes in well-structured exercises are essential elements in optimising the speech output in people with impaired cognitive functions.

Naïve ratings of the intelligibility of the spontaneous speech of adults with mild to moderate ID related significantly to the phonemic and syllabic variables derived from expert transcriptions of their verbal output obtained with a picture-naming task. The main speech difficulties in this group of impaired speakers appear to be related to underlying speech motor planning difficulties. There are early indications that tailored speech therapy can remediate some of the speech deficits in adults with ID, fostering their communicative skills by reducing the severity of their speech difficulties.

SPEECH INTELLIGIBILITY

the degree that a person's speech can be understood by other people. Have a look on articulation index .

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest Posts

What Happens At An ADHD Assessment

A Quick Look at the History Behind Hypnosis

A Brief History of Brainwashing: The Science of Thought Control

A Deep Dive into the Social Psychology of Leadership

Counseling Approaches to Client Care: Theories to Apply in Practice

The Future Of Education: Can You Earn A Psychology Degree Online?

Insomnia & Mental Illness: What is the Correlation?

Stop Guessing: Here Are 3 Steps to Data-Driven Psychological Decisions

Getting Help with Grief: Understanding Therapy & How It Can Help

Exploring the Psychology of Risk and Reward

Understanding ADHD in Women: Symptoms, Treatment & Support

Meeting the Milestones: A Guide to Piaget's Child Developmental Stages

Popular psychology terms, medical model, hypermnesia, affirmation, brainwashing, backup reinforcer, message-learning approach, affiliative behavior, behavioral modeling, approach motivation.

Speech Intelligibility

Speech is our primary method of communication. It is therefore important that uttered speech is received intelligibly. The intelligibility of speech depends (in part) on the acoustical properties of the enclosure in which the speech is transmitted from speaker to listener. Another important factor determining the speech intelligibility is the background noise level. Although there have been many attempts to objectively quantify the speech intelligiblity, the most widely used parameter is no doubt the Speech Transmission Index (STI) and its derivatives. The STI is based on the relation between perceived speech intelligibility and the intensity modulations in the talker's voice, as described by Houtgast, Steeneken and Plomp. 1) The STI method is described in the IEC 60268-16 standard.

When a sound source in a room is producing noise that is intensity modulated by a low frequency sinusoidal modulation of 100% depth, the modulation at the receiver position will be reduced due to room reflections and background noise. The Modulation Transfer Function (MTF) describes to what extent the modulation m is transferred from source to receiver, as a function of the modulation frequency F, which ranges from 0.63 to 12.5 Hz. Hence, the MTF depends on the system properties and the background noise.

With the introduction of Dirac 6 and the Echo Speech Source , speech intelligibility measurements can be performed very quickly and easily. The Echo delivers a calibrated signal that is used by Dirac to calculate different speech intelligibility parameters. The Echo / Dirac combination also performs well with high levels of background noise. It is therefore no longer necessary to work with external equalizers or to set the output level. Quite often, the difficulty of emitting a signal with the correct speech spectrum and level, meant that shortcuts were taken which led to questionable results.

1) T. Houtgast, H.J.M. Steeneken and R. Plomp, 'Predicting Speech Intelligibility in Rooms from the Modulation Transfer Function. I. General Room Acoustics,' Acustica 46, 60 - 72 (1980).

You can download our speech intelligibility technical note as a pdf.

Modulated noise versus impulse responses.

The work of Houtgast and Steeneken is based on a STI method using modulated noise as a test signal. Although Schroeder 2) has shown that the MTF can also be calculated from the impulse response, handheld (STIPA) meters use modulated noise as a test signal, whereas most PC based simulation and measurement software uses the impulse response method. This is because movement of the handheld device during a measurement would make impulse response measurements unsuitable. In general it can be said that the modulated noise method is somewhat more resilient to non-linear and time-varying systems, whereas the impulse response method is faster and provides more information.

2) M.R. Schroeder, 'Modulation Transfer Functions: Definition and Measurement,' Acustica 49, 179 - 182 (1981).

STI and STIPA

The STIPA parameter features prominently in the latest edition of IEC 60268-16. It is described as the preferred parameter for almost all measurement situations. However, the STIPA was originally conceived much like the RASTI as a way to estimate the STI within a reasonable time. Traditionally the STIPA is measured with modulated noise. A full STI measurement with modulated noise would take at least 15 minutes. However, by using the impulse response approach the STI can be measured just as fast as the STIPA. The STI having many more modulation frequencies in each octave band, has far greater diagnostic power than the STIPA.

A short history of the STI method

1971 : The first mention of the STI (Speech Transmission Index, a measure of speech intelligibility) in an article in Acustica by Tammo Houtgast and Herman Steeneken. 1981 : Manfred Schroeder writes an article in Acustica in which he showed that the modulation transfer function (MTF) can be derived from an impulse response. 1985 : B&K introduces the Rapid Speech Transmission Index Meter 3361. 1988 : The STI method and specifically the RASTI parameter are described in the IEC 60268-16 standard. Subsequent revisions of the standard have added i.a. gender specific test signals, redundancy factors and level dependent masking. 2003 : The STIPA parameter is added to the revision 3 of the IEC 60268-16 standard. 2011 : IEC 60268-16 Ed 4.0 is adopted.

Echo stimuli

MLS versus sweep The Echo Speech Source contains a male speech signal to help adjust the volume of PA systems. For STI measurements it contains pink MLS signals. Normally this would present a serious problem because the slightest difference in clock rate between source and receiver would make it impossible to properly extract the impulse response. Dirac however has been able to handle asynchronous MLS signals since version 4. The advantage of MLS as opposed to an e-sweep signal is that the MLS is far less intrusive. Also, because in sweeps all the energy is always concentrated at a single frequency, it is more difficult for amplifiers and speakers to handle this type of signal.

Intermittent stimulus New in Dirac 6 and also used in the Echo is the intermittent stimulus. With the standard impulse response technique it is difficult to measure a high quality impulse response and at the same time retrieve an accurate (background) noise level from this measurement. The new intermittent stimulus consists of an MLS sequence followed by an equally long period of silence. The full stimulus (MLS + silence) is measured in one pass, and Dirac extracts the impulse response and the background noise into two separate channels of a .wav file. This stimulus allows you to increase the output level and perform pre-averaging to improve the INR of the impulse response in speech intelligibility measurements, while still retaining accurate noise values.

The Echo signals One of the signals in the Echo is a speech fragment that can be used to set the volume of a PA system to a ‘normal’ level. The speech signal has a standard level of 60 dB(A), and cannot be used for speech intelligibility measurements. When the background noise level is relatively low and/or the reverberation time is relatively long (SNR * RT > 120 dBs), a simple continuous MLS stimulus can be used. This signal is available at 60 dB(A) and at a raised level of 72 dB(A). Note that the MLS sequences are coded such that Dirac can always determine the output level, and correct the STI calculations appropriately. For scenarios where the background noise has a significant impact (SNR * RT < 120 dBs), an intermittent MLS signal is available both at 60 dB(A) and 72 dB(A). The signals generated by the Echo can also be injected directly into a PA system using the BNC output connector. The electrical output always operates at the same level. You can also play the Echo signals via the PA system from a CD or MP3 player. For this purpose we have made the Echo signal available as a separate download .

ISO 3382-3 open plan offices

For speech intelligibility measurements in open plan offices, the ISO 3382-3 suggests you perform 4 measurements at each position. With Dirac 6 you only need a single (system- and level-calibrated) measurement per position. Using the new intermittent stimulus, the speech signal levels, the impulse response and the background noise levels can be acquired in a single pass. Plots of the STI versus the source-receiver distance can be created with a few mouse clicks. You perform the minimum amount of measurements and DIrac will give you L P,A,S , L P,A,S,4 , L P,A,B , D 2,S , STI, r D and r P .

The ISO 3382-3 standard prescribes the use of an omnidirectional sound source such as the OmniPower 4292-L . Also, the STI that is used is a little diferent from the STI defined in IEC 60268-16, in that the auditory masking and hearing threshold corrections are not used. Dirac contains separate STI parameters for ISO 3382-3 and IEC 60268-16. In some cases it may be useful to use a directional source for ISO 3382-3 measurements. The Echo speech source can now be used for this purpose as explained in this blog post .

The video below will show you how open plan office measurements are prepared, performed and analyzed.

The modulation transfer function (MTF) displays the modulation depth as a function of the modulation frequency. It is an intermediate result obtained during the calculation of the STI and related parameters. However, it is also an important diagnostic tool to investigate the causes of poor speech intelligibility. An MTF that is constant over the modulation frequencies indicates that the speech intelligibility is mainly determined by background noise. A continuously decreasing MTF indicates the influence of reverberation and an MTF that decreases first and then increases again indicates the presence of an echo.

Traditional modulated noise based speech intelligibility measurements contain a limited number of modulation frequencies. This means that many problems can remain hidden in the coarsely sampled modulation spectrum. Impulse responses contain the full spectrum of modulation frequencies, and Dirac 6 now has the ability to show them in continuous MTF graphs.

American Speech-Language-Hearing Association

• Certification
• Publications
• Continuing Education
• Practice Management
• Audiologists
• Speech-Language Pathologists
• Academic & Faculty
• Audiology & SLP Assistants

Speech Sound Disorders-Articulation and Phonology

View All Portal Topics

See the Speech Sound Disorders Evidence Map for summaries of the available research on this topic.

The scope of this page is speech sound disorders with no known cause—historically called articulation and phonological disorders —in preschool and school-age children (ages 3–21).

Information about speech sound problems related to motor/neurological disorders, structural abnormalities, and sensory/perceptual disorders (e.g., hearing loss) is not addressed in this page.

See ASHA's Practice Portal pages on Childhood Apraxia of Speech and Cleft Lip and Palate for information about speech sound problems associated with these two disorders. A Practice Portal page on dysarthria in children will be developed in the future.

Speech Sound Disorders

Speech sound disorders is an umbrella term referring to any difficulty or combination of difficulties with perception, motor production, or phonological representation of speech sounds and speech segments—including phonotactic rules governing permissible speech sound sequences in a language.

Speech sound disorders can be organic or functional in nature. Organic speech sound disorders result from an underlying motor/neurological, structural, or sensory/perceptual cause. Functional speech sound disorders are idiopathic—they have no known cause. See figure below.

Organic Speech Sound Disorders

Organic speech sound disorders include those resulting from motor/neurological disorders (e.g., childhood apraxia of speech and dysarthria), structural abnormalities (e.g., cleft lip/palate and other structural deficits or anomalies), and sensory/perceptual disorders (e.g., hearing loss).

Functional Speech Sound Disorders

Functional speech sound disorders include those related to the motor production of speech sounds and those related to the linguistic aspects of speech production. Historically, these disorders are referred to as articulation disorders and phonological disorders , respectively. Articulation disorders focus on errors (e.g., distortions and substitutions) in production of individual speech sounds. Phonological disorders focus on predictable, rule-based errors (e.g., fronting, stopping, and final consonant deletion) that affect more than one sound. It is often difficult to cleanly differentiate between articulation and phonological disorders; therefore, many researchers and clinicians prefer to use the broader term, "speech sound disorder," when referring to speech errors of unknown cause. See Bernthal, Bankson, and Flipsen (2017) and Peña-Brooks and Hegde (2015) for relevant discussions.

This Practice Portal page focuses on functional speech sound disorders. The broad term, "speech sound disorder(s)," is used throughout; articulation error types and phonological error patterns within this diagnostic category are described as needed for clarity.

Procedures and approaches detailed in this page may also be appropriate for assessing and treating organic speech sound disorders. See Speech Characteristics: Selected Populations [PDF] for a brief summary of selected populations and characteristic speech problems.

Incidence and Prevalence

The incidence of speech sound disorders refers to the number of new cases identified in a specified period. The prevalence of speech sound disorders refers to the number of children who are living with speech problems in a given time period.

Estimated prevalence rates of speech sound disorders vary greatly due to the inconsistent classifications of the disorders and the variance of ages studied. The following data reflect the variability:

• Overall, 2.3% to 24.6% of school-aged children were estimated to have speech delay or speech sound disorders (Black, Vahratian, & Hoffman, 2015; Law, Boyle, Harris, Harkness, & Nye, 2000; Shriberg, Tomblin, & McSweeny, 1999; Wren, Miller, Peters, Emond, & Roulstone, 2016).
• A 2012 survey from the National Center for Health Statistics estimated that, among children with a communication disorder, 48.1% of 3- to 10-year old children and 24.4% of 11- to 17-year old children had speech sound problems only. Parents reported that 67.6% of children with speech problems received speech intervention services (Black et al., 2015).
• Residual or persistent speech errors were estimated to occur in 1% to 2% of older children and adults (Flipsen, 2015).
• Reports estimated that speech sound disorders are more prevalent in boys than in girls, with a ratio ranging from 1.5:1.0 to 1.8:1.0 (Shriberg et al., 1999; Wren et al., 2016).
• Prevalence rates were estimated to be 5.3% in African American children and 3.8% in White children (Shriberg et al., 1999).
• Reports estimated that 11% to 40% of children with speech sound disorders had concomitant language impairment (Eadie et al., 2015; Shriberg et al., 1999).
• Poor speech sound production skills in kindergarten children have been associated with lower literacy outcomes (Overby, Trainin, Smit, Bernthal, & Nelson, 2012). Estimates reported a greater likelihood of reading disorders (relative risk: 2.5) in children with a preschool history of speech sound disorders (Peterson, Pennington, Shriberg, & Boada, 2009).

Signs and Symptoms

Signs and symptoms of functional speech sound disorders include the following:

• omissions/deletions —certain sounds are omitted or deleted (e.g., "cu" for "cup" and "poon" for "spoon")
• substitutions —one or more sounds are substituted, which may result in loss of phonemic contrast (e.g., "thing" for "sing" and "wabbit" for "rabbit")
• additions —one or more extra sounds are added or inserted into a word (e.g., "buhlack" for "black")
• distortions —sounds are altered or changed (e.g., a lateral "s")
• syllable-level errors —weak syllables are deleted (e.g., "tephone" for "telephone")

Signs and symptoms may occur as independent articulation errors or as phonological rule-based error patterns (see ASHA's resource on selected phonological processes [patterns] for examples). In addition to these common rule-based error patterns, idiosyncratic error patterns can also occur. For example, a child might substitute many sounds with a favorite or default sound, resulting in a considerable number of homonyms (e.g., shore, sore, chore, and tore might all be pronounced as door ; Grunwell, 1987; Williams, 2003a).

Influence of Accent

An accent is the unique way that speech is pronounced by a group of people speaking the same language and is a natural part of spoken language. Accents may be regional; for example, someone from New York may sound different than someone from South Carolina. Foreign accents occur when a set of phonetic traits of one language are carried over when a person learns a new language. The first language acquired by a bilingual or multilingual individual can influence the pronunciation of speech sounds and the acquisition of phonotactic rules in subsequently acquired languages. No accent is "better" than another. Accents, like dialects, are not speech or language disorders but, rather, only reflect differences. See ASHA's Practice Portal pages on Multilingual Service Delivery in Audiology and Speech-Language Pathology and Cultural Responsiveness .

Influence of Dialect

Not all sound substitutions and omissions are speech errors. Instead, they may be related to a feature of a speaker's dialect (a rule-governed language system that reflects the regional and social background of its speakers). Dialectal variations of a language may cross all linguistic parameters, including phonology, morphology, syntax, semantics, and pragmatics. An example of a dialectal variation in phonology occurs with speakers of African American English (AAE) when a "d" sound is used for a "th" sound (e.g., "dis" for "this"). This variation is not evidence of a speech sound disorder but, rather, one of the phonological features of AAE.

Speech-language pathologists (SLPs) must distinguish between dialectal differences and communicative disorders and must

• recognize all dialects as being rule-governed linguistic systems;
• understand the rules and linguistic features of dialects represented by their clientele; and
• be familiar with nondiscriminatory testing and dynamic assessment procedures, such as identifying potential sources of test bias, administering and scoring standardized tests using alternative methods, and analyzing test results in light of existing information regarding dialect use (see, e.g., McLeod, Verdon, & The International Expert Panel on Multilingual Children's Speech, 2017).

See ASHA's Practice Portal pages on Multilingual Service Delivery in Audiology and Speech-Language Pathology and Cultural Responsiveness .

The cause of functional speech sound disorders is not known; however, some risk factors have been investigated.

Frequently reported risk factors include the following:

• Gender —the incidence of speech sound disorders is higher in males than in females (e.g., Everhart, 1960; Morley, 1952; Shriberg et al., 1999).
• Pre- and perinatal problems —factors such as maternal stress or infections during pregnancy, complications during delivery, preterm delivery, and low birthweight were found to be associated with delay in speech sound acquisition and with speech sound disorders (e.g., Byers Brown, Bendersky, & Chapman, 1986; Fox, Dodd, & Howard, 2002).
• Family history —children who have family members (parents or siblings) with speech and/or language difficulties were more likely to have a speech disorder (e.g., Campbell et al., 2003; Felsenfeld, McGue, & Broen, 1995; Fox et al., 2002; Shriberg & Kwiatkowski, 1994).
• Persistent otitis media with effusion —persistent otitis media with effusion (often associated with hearing loss) has been associated with impaired speech development (Fox et al., 2002; Silva, Chalmers, & Stewart, 1986; Teele, Klein, Chase, Menyuk, & Rosner, 1990).

Roles and Responsibilities

Speech-language pathologists (SLPs) play a central role in the screening, assessment, diagnosis, and treatment of persons with speech sound disorders. The professional roles and activities in speech-language pathology include clinical/educational services (diagnosis, assessment, planning, and treatment); prevention and advocacy; and education, administration, and research. See ASHA's Scope of Practice in Speech-Language Pathology (ASHA, 2016).

Appropriate roles for SLPs include the following:

• Providing prevention information to individuals and groups known to be at risk for speech sound disorders, as well as to individuals working with those at risk
• Educating other professionals on the needs of persons with speech sound disorders and the role of SLPs in diagnosing and managing speech sound disorders
• Screening individuals who present with speech sound difficulties and determining the need for further assessment and/or referral for other services
• Recognizing that students with speech sound disorders have heightened risks for later language and literacy problems
• Conducting a culturally and linguistically relevant comprehensive assessment of speech, language, and communication
• Taking into consideration the rules of a spoken accent or dialect, typical dual-language acquisition from birth, and sequential second-language acquisition to distinguish difference from disorder
• Diagnosing the presence or absence of a speech sound disorder
• Referring to and collaborating with other professionals to rule out other conditions, determine etiology, and facilitate access to comprehensive services
• Making decisions about the management of speech sound disorders
• Making decisions about eligibility for services, based on the presence of a speech sound disorder
• Developing treatment plans, providing intervention and support services, documenting progress, and determining appropriate service delivery approaches and dismissal criteria
• Counseling persons with speech sound disorders and their families/caregivers regarding communication-related issues and providing education aimed at preventing further complications related to speech sound disorders
• Serving as an integral member of an interdisciplinary team working with individuals with speech sound disorders and their families/caregivers (see ASHA's resource on interprofessional education/interprofessional practice [IPE/IPP] )
• Consulting and collaborating with professionals, family members, caregivers, and others to facilitate program development and to provide supervision, evaluation, and/or expert testimony (see ASHA's resource on person- and family-centered care )
• Remaining informed of research in the area of speech sound disorders, helping advance the knowledge base related to the nature and treatment of these disorders, and using evidence-based research to guide intervention
• Advocating for individuals with speech sound disorders and their families at the local, state, and national levels

As indicated in the Code of Ethics (ASHA, 2023), SLPs who serve this population should be specifically educated and appropriately trained to do so.

See the Assessment section of the Speech Sound Disorders Evidence Map for pertinent scientific evidence, expert opinion, and client/caregiver perspective.

Screening is conducted whenever a speech sound disorder is suspected or as part of a comprehensive speech and language evaluation for a child with communication concerns. The purpose of the screening is to identify individuals who require further speech-language assessment and/or referral for other professional services.

Screening typically includes

• screening of individual speech sounds in single words and in connected speech (using formal and or informal screening measures);
• screening of oral motor functioning (e.g., strength and range of motion of oral musculature);
• orofacial examination to assess facial symmetry and identify possible structural bases for speech sound disorders (e.g., submucous cleft palate, malocclusion, ankyloglossia); and
• informal assessment of language comprehension and production.

See ASHA's resource on assessment tools, techniques, and data sources .

Screening may result in

• recommendation to monitor speech and rescreen;
• referral for multi-tiered systems of support such as response to intervention (RTI) ;
• referral for a comprehensive speech sound assessment;
• recommendation for a comprehensive language assessment, if language delay or disorder is suspected;
• referral to an audiologist for a hearing evaluation, if hearing loss is suspected; and
• referral for medical or other professional services, as appropriate.

Comprehensive Assessment

The acquisition of speech sounds is a developmental process, and children often demonstrate "typical" errors and phonological patterns during this acquisition period. Developmentally appropriate errors and patterns are taken into consideration during assessment for speech sound disorders in order to differentiate typical errors from those that are unusual or not age appropriate.

The comprehensive assessment protocol for speech sound disorders may include an evaluation of spoken and written language skills, if indicated. See ASHA's Practice Portal pages on Spoken Language Disorders and Written Language Disorders .

Assessment is accomplished using a variety of measures and activities, including both standardized and nonstandardized measures, as well as formal and informal assessment tools. See ASHA's resource on assessment tools, techniques, and data sources .

SLPs select assessments that are culturally and linguistically sensitive, taking into consideration current research and best practice in assessing speech sound disorders in the languages and/or dialect used by the individual (see, e.g., McLeod et al., 2017). Standard scores cannot be reported for assessments that are not normed on a group that is representative of the individual being assessed.

SLPs take into account cultural and linguistic speech differences across communities, including

• phonemic and allophonic variations of the language(s) and/or dialect(s) used in the community and how those variations affect determination of a disorder or a difference and
• differences among speech sound disorders, accents, dialects, and patterns of transfer from one language to another. See phonemic inventories and cultural and linguistic information across languages .

Consistent with the World Health Organization's (WHO) International Classification of Functioning, Disability and Health (ICF) framework (ASHA, 2016a; WHO, 2001), a comprehensive assessment is conducted to identify and describe

• impairments in body structure and function, including underlying strengths and weaknesses in speech sound production and verbal/nonverbal communication;
• co-morbid deficits or conditions, such as developmental disabilities, medical conditions, or syndromes;
• limitations in activity and participation, including functional communication, interpersonal interactions with family and peers, and learning;
• contextual (environmental and personal) factors that serve as barriers to or facilitators of successful communication and life participation; and
• the impact of communication impairments on quality of life of the child and family.

See ASHA's Person-Centered Focus on Function: Speech Sound Disorder [PDF] for an example of assessment data consistent with ICF.

Assessment may result in

• diagnosis of a speech sound disorder;
• description of the characteristics and severity of the disorder;
• recommendations for intervention targets;
• identification of factors that might contribute to the speech sound disorder;
• diagnosis of a spoken language (listening and speaking) disorder;
• identification of written language (reading and writing) problems;
• recommendation to monitor reading and writing progress in students with identified speech sound disorders by SLPs and other professionals in the school setting;
• referral for multi-tiered systems of support such as response to intervention (RTI) to support speech and language development; and
• referral to other professionals as needed.

Case History

The case history typically includes gathering information about

• the family's concerns about the child's speech;
• history of middle ear infections;
• family history of speech and language difficulties (including reading and writing);
• languages used in the home;
• primary language spoken by the child;
• the family's and other communication partners' perceptions of intelligibility; and
• the teacher's perception of the child's intelligibility and participation in the school setting and how the child's speech compares with that of peers in the classroom.

See ASHA's Practice Portal page on Cultural Responsiveness for guidance on taking a case history with all clients.

Oral Mechanism Examination

The oral mechanism examination evaluates the structure and function of the speech mechanism to assess whether the system is adequate for speech production. This examination typically includes assessment of

• dental occlusion and specific tooth deviations;
• structure of hard and soft palate (clefts, fistulas, bifid uvula); and
• function (strength and range of motion) of the lips, jaw, tongue, and velum.

Hearing Screening

A hearing screening is conducted during the comprehensive speech sound assessment, if one was not completed during the screening.

Hearing screening typically includes

• otoscopic inspection of the ear canal and tympanic membrane;
• pure-tone audiometry; and
• immittance testing to assess middle ear function.

Speech Sound Assessment

The speech sound assessment uses both standardized assessment instruments and other sampling procedures to evaluate production in single words and connected speech.

Single-word testing provides identifiable units of production and allows most consonants in the language to be elicited in a number of phonetic contexts; however, it may or may not accurately reflect production of the same sounds in connected speech.

Connected speech sampling provides information about production of sounds in connected speech using a variety of talking tasks (e.g., storytelling or retelling, describing pictures, normal conversation about a topic of interest) and with a variety of communication partners (e.g., peers, siblings, parents, and clinician).

Assessment of speech includes evaluation of the following:

• Accurate productions
• sounds in various word positions (e.g., initial, within word, and final word position) and in different phonetic contexts;
• sound combinations such as vowel combinations, consonant clusters, and blends; and
• syllable shapes —simple CV to complex CCVCC.
• Speech sound errors
• consistent sound errors;
• error types (e.g., deletions, omissions, substitutions, distortions, additions); and
• error distribution (e.g., position of sound in word).
• Error patterns (i.e., phonological patterns)—systematic sound changes or simplifications that affect a class of sounds (e.g., fricatives), sound combinations (e.g., consonant clusters), or syllable structures (e.g., complex syllables or multisyllabic words).

See Age of Acquisition of English Consonants (Crowe & McLeaod, 2020) [PDF] and ASHA's resource on selected phonological processes (patterns) .

Severity Assessment

Severity is a qualitative judgment made by the clinician indicating the impact of the child's speech sound disorder on functional communication. It is typically defined along a continuum from mild to severe or profound. There is no clear consensus regarding the best way to determine severity of a speech sound disorder—rating scales and quantitative measures have been used.

A numerical scale or continuum of disability is often used because it is time-efficient. Prezas and Hodson (2010) use a continuum of severity from mild (omissions are rare; few substitutions) to profound (extensive omissions and many substitutions; extremely limited phonemic and phonotactic repertoires). Distortions and assimilations occur in varying degrees at all levels of the continuum.

A quantitative approach (Shriberg & Kwiatkowski, 1982a, 1982b) uses the percentage of consonants correct (PCC) to determine severity on a continuum from mild to severe.

To determine PCC, collect and phonetically transcribe a speech sample. Then count the total number of consonants in the sample and the total number of correct consonants. Use the following formula:

PCC = (correct consonants/total consonants) × 100

A PCC of 85–100 is considered mild, whereas a PCC of less than 50 is considered severe. This approach has been modified to include a total of 10 such indices, including percent vowels correct (PVC; Shriberg, Austin, Lewis, McSweeny, & Wilson, 1997).

Intelligibility Assessment

Intelligibility is a perceptual judgment that is based on how much of the child's spontaneous speech the listener understands. Intelligibility can vary along a continuum ranging from intelligible (message is completely understood) to unintelligible (message is not understood; Bernthal et al., 2017). Intelligibility is frequently used when judging the severity of the child's speech problem (Kent, Miolo, & Bloedel, 1994; Shriberg & Kwiatkowski, 1982b) and can be used to determine the need for intervention.

Intelligibility can vary depending on a number of factors, including

• the number, type, and frequency of speech sound errors (when present);
• the speaker's rate, inflection, stress patterns, pauses, voice quality, loudness, and fluency;
• linguistic factors (e.g., word choice and grammar);
• complexity of utterance (e.g., single words vs. conversational or connected speech);
• the listener's familiarity with the speaker's speech pattern;
• communication environment (e.g., familiar vs. unfamiliar communication partners, one-on-one vs. group conversation);
• communication cues for listener (e.g., nonverbal cues from the speaker, including gestures and facial expressions); and
• signal-to-noise ratio (i.e., amount of background noise).

Rating scales and other estimates that are based on perceptual judgments are commonly used to assess intelligibility. For example, rating scales sometimes use numerical ratings like 1 for totally intelligible and 10 for unintelligible, or they use descriptors like not at all, seldom, sometimes, most of the time, or always to indicated how well speech is understood (Ertmer, 2010).

A number of quantitative measures also have been proposed, including calculating the percentage of words understood in conversational speech (e.g., Flipsen, 2006; Shriberg & Kwiatkowski, 1980). See also Kent et al. (1994) for a comprehensive review of procedures for assessing intelligibility.

Coplan and Gleason (1988) developed a standardized intelligibility screener using parent estimates of how intelligible their child sounded to others. On the basis of the data, expected intelligibility cutoff values for typically developing children were as follows:

22 months—50%

37 months—75%

47 months—100%

See the Resources section for resources related to assessing intelligibility and life participation in monolingual children who speak English and in monolingual children who speak languages other than English.

Stimulability Testing

Stimulability is the child's ability to accurately imitate a misarticulated sound when the clinician provides a model. There are few standardized procedures for testing stimulability (Glaspey & Stoel-Gammon, 2007; Powell & Miccio, 1996), although some test batteries include stimulability subtests.

Stimulability testing helps determine

• how well the child imitates the sound in one or more contexts (e.g., isolation, syllable, word, phrase);
• the level of cueing necessary to achieve the best production (e.g., auditory model; auditory and visual model; auditory, visual, and verbal model; tactile cues);
• whether the sound is likely to be acquired without intervention; and
• which targets are appropriate for therapy (Tyler & Tolbert, 2002).

Speech Perception Testing

Speech perception is the ability to perceive differences between speech sounds. In children with speech sound disorders, speech perception is the child's ability to perceive the difference between the standard production of a sound and his or her own error production—or to perceive the contrast between two phonetically similar sounds (e.g., r/w, s/ʃ, f/θ).

Speech perception abilities can be tested using the following paradigms:

• Auditory Discrimination —syllable pairs containing a single phoneme contrast are presented, and the child is instructed to say "same" if the paired items sound the same and "different" if they sound different.
• Picture Identification —the child is shown two to four pictures representing words with minimal phonetic differences. The clinician says one of these words, and the child is asked to point to the correct picture.
• Speech production–perception task —using sounds that the child is suspected of having difficulty perceiving, picture targets containing these sounds are used as visual cues. The child is asked to judge whether the speaker says the item correctly (e.g., picture of a ship is shown; speaker says, "ship" or "sip"; Locke, 1980).
• Mispronunciation detection task —using computer-presented picture stimuli and recorded stimulus names (either correct or with a single phoneme error), the child is asked to detect mispronunciations by pointing to a green tick for "correct" or a red cross for "incorrect" (McNeill & Hesketh, 2010).
• Lexical decision/judgment task —using target pictures and single-word recordings, this task assesses the child's ability to identify words that are pronounced correctly or incorrectly. A picture of the target word (e.g., "lake") is shown, along with a recorded word—either "lake" or a word with a contrasting phoneme (e.g., "wake"). The child points to the picture of the target word if it was pronounced correctly or to an "X" if it was pronounced incorrectly (Rvachew, Nowak, & Cloutier, 2004).

Considerations For Assessing Young Children and/or Children Who Are Reluctant or Have Less Intelligible Speech

Young children might not be able to follow directions for standardized tests, might have limited expressive vocabulary, and might produce words that are unintelligible. Other children, regardless of age, may produce less intelligible speech or be reluctant to speak in an assessment setting.

Strategies for collecting an adequate speech sample with these populations include

• obtaining a speech sample during the assessment session using play activities;
• using pictures or toys to elicit a range of consonant sounds;
• involving parents/caregivers in the session to encourage talking;
• asking parents/caregivers to supplement data from the assessment session by recording the child's speech at home during spontaneous conversation; and
• asking parents/caregivers to keep a log of the child's intended words and how these words are pronounced.

Sometimes, the speech sound disorder is so severe that the child's intended message cannot be understood. However, even when a child's speech is unintelligible, it is usually possible to obtain information about his or her speech sound production.

For example:

• A single-word articulation test provides opportunities for production of identifiable units of sound, and these productions can usually be transcribed.
• It may be possible to understand and transcribe a spontaneous speech sample by (a) using a structured situation to provide context when obtaining the sample and (b) annotating the recorded sample by repeating the child's utterances, when possible, to facilitate later transcription.

Considerations For Assessing Bilingual/Multilingual Populations

Assessment of a bilingual individual requires an understanding of both linguistic systems because the sound system of one language can influence the sound system of another language. The assessment process must identify whether differences are truly related to a speech sound disorder or are normal variations of speech caused by the first language.

When assessing a bilingual or multilingual individual, clinicians typically

• gather information, including
• language history and language use to determine which language(s) should be assessed,
• phonemic inventory, phonological structure, and syllable structure of the non-English language, and
• dialect of the individual;
• assess phonological skills in both languages in single words as well as in connected speech;
• account for dialectal differences, when present; and
• identify and assess the child's
• common substitution patterns (those seen in typically developing children),
• uncommon substitution patterns (those often seen in individuals with a speech sound disorder), and
• cross-linguistic effects (the phonological system of one's native language influencing the production of sounds in English, resulting in an accent—that is, phonetic traits from a person's original language (L1) that are carried over to a second language (L2; Fabiano-Smith & Goldstein, 2010).

See phonemic inventories and cultural and linguistic information across languages and ASHA's Practice Portal page on Multilingual Service Delivery in Audiology and Speech-Language Pathology . See the Resources section for information related to assessing intelligibility and life participation in monolingual children who speak English and in monolingual children who speak languages other than English.

Phonological Processing Assessment

Phonological processing is the use of the sounds of one's language (i.e., phonemes) to process spoken and written language (Wagner & Torgesen, 1987). The broad category of phonological processing includes phonological awareness , phonological working memory , and phonological retrieval .

All three components of phonological processing (see definitions below) are important for speech production and for the development of spoken and written language skills. Therefore, it is important to assess phonological processing skills and to monitor the spoken and written language development of children with phonological processing difficulties.

• Phonological Awareness is the awareness of the sound structure of a language and the ability to consciously analyze and manipulate this structure via a range of tasks, such as speech sound segmentation and blending at the word, onset-rime, syllable, and phonemic levels.
• Phonological Working Memory involves storing phoneme information in a temporary, short-term memory store (Wagner & Torgesen, 1987). This phonemic information is then readily available for manipulation during phonological awareness tasks. Nonword repetition (e.g., repeat "/pæɡ/") is one example of a phonological working memory task.
• Phonological Retrieval is the ability to retrieve phonological information from long-term memory. It is typically assessed using rapid naming tasks (e.g., rapid naming of objects, colors, letters, or numbers). This ability to retrieve the phonological information of one's language is integral to phonological awareness.

Language Assessments

Language testing is included in a comprehensive speech sound assessment because of the high incidence of co-occurring language problems in children with speech sound disorders (Shriberg & Austin, 1998).

Spoken Language Assessment (Listening and Speaking)

Typically, the assessment of spoken language begins with a screening of expressive and receptive skills; a full battery is performed if indicated by screening results. See ASHA's Practice Portal page on Spoken Language Disorders for more details.

Written Language Assessment (Reading and Writing)

Difficulties with the speech processing system (e.g., listening, discriminating speech sounds, remembering speech sounds, producing speech sounds) can lead to speech production and phonological awareness difficulties. These difficulties can have a negative impact on the development of reading and writing skills (Anthony et al., 2011; Catts, McIlraith, Bridges, & Nielsen, 2017; Leitão & Fletcher, 2004; Lewis et al., 2011).

For typically developing children, speech production and phonological awareness develop in a mutually supportive way (Carroll, Snowling, Stevenson, & Hulme, 2003; National Institute for Literacy, 2009). As children playfully engage in sound play, they eventually learn to segment words into separate sounds and to "map" sounds onto printed letters.

The understanding that sounds are represented by symbolic code (e.g., letters and letter combinations) is essential for reading and spelling. When reading, children have to be able to segment a written word into individual sounds, based on their knowledge of the code and then blend those sounds together to form a word. When spelling, children have to be able to segment a spoken word into individual sounds and then choose the correct code to represent these sounds (National Institute of Child Health and Human Development, 2000; Pascoe, Stackhouse, & Wells, 2006).

Components of the written language assessment include the following, depending on the child's age and expected stage of written language development:

• Print Awareness —recognizing that books have a front and back, recognizing that the direction of words is from left to right, and recognizing where words on the page start and stop.
• Alphabet Knowledge —including naming/printing alphabet letters from A to Z.
• Sound–Symbol Correspondence —knowing that letters have sounds and knowing the sounds for corresponding letters and letter combinations.
• Reading Decoding —using sound–symbol knowledge to segment and blend sounds in grade-level words.
• Spelling —using sound–symbol knowledge to spell grade-level words.
• Reading Fluency —reading smoothly without frequent or significant pausing.
• Reading Comprehension —understanding grade-level text, including the ability to make inferences.

See ASHA's Practice Portal page on Written Language Disorders for more details.

See the Treatment section of the Speech Sound Disorders Evidence Map for pertinent scientific evidence, expert opinion, and client/caregiver perspective.

The broad term "speech sound disorder(s)" is used in this Portal page to refer to functional speech sound disorders, including those related to the motor production of speech sounds (articulation) and those related to the linguistic aspects of speech production (phonological).

It is often difficult to cleanly differentiate between articulation and phonological errors or to differentially diagnose these two separate disorders. Nevertheless, we often talk about articulation error types and phonological error types within the broad diagnostic category of speech sound disorder(s). A single child might show both error types, and those specific errors might need different treatment approaches.

Historically, treatments that focus on motor production of speech sounds are called articulation approaches; treatments that focus on the linguistic aspects of speech production are called phonological/language-based approaches.

Articulation approaches target each sound deviation and are often selected by the clinician when the child's errors are assumed to be motor based; the aim is correct production of the target sound(s).

Phonological/language-based approaches target a group of sounds with similar error patterns, although the actual treatment of exemplars of the error pattern may target individual sounds. Phonological approaches are often selected in an effort to help the child internalize phonological rules and generalize these rules to other sounds within the pattern (e.g., final consonant deletion, cluster reduction).

Articulation and phonological/language-based approaches might both be used in therapy with the same individual at different times or for different reasons.

Both approaches for the treatment of speech sound disorders typically involve the following sequence of steps:

• Establishment —eliciting target sounds and stabilizing production on a voluntary level.
• Generalization —facilitating carry-over of sound productions at increasingly challenging levels (e.g., syllables, words, phrases/sentences, conversational speaking).
• Maintenance —stabilizing target sound production and making it more automatic; encouraging self-monitoring of speech and self-correction of errors.

Target Selection

Approaches for selecting initial therapy targets for children with articulation and/or phonological disorders include the following:

• Developmental —target sounds are selected on the basis of order of acquisition in typically developing children.
• Complexity —focuses on more complex, linguistically marked phonological elements not in the child's phonological system to encourage cascading, generalized learning of sounds (Gierut, 2007; Storkel, 2018).
• Dynamic systems —focuses on teaching and stabilizing simple target phonemes that do not introduce new feature contrasts in the child's phonological system to assist in the acquisition of target sounds and more complex targets and features (Rvachew & Bernhardt, 2010).
• Systemic —focuses on the function of the sound in the child's phonological organization to achieve maximum phonological reorganization with the least amount of intervention. Target selection is based on a distance metric. Targets can be maximally distinct from the child's error in terms of place, voice, and manner and can also be maximally different in terms of manner, place of production, and voicing (Williams, 2003b). See Place, Manner and Voicing Chart for English Consonants (Roth & Worthington, 2018) .
• Client-specific —selects targets based on factors such as relevance to the child and his or her family (e.g., sound is in child's name), stimulability, and/or visibility when produced (e.g., /f/ vs. /k/).
• Degree of deviance and impact on intelligibility —selects targets on the basis of errors (e.g., errors of omission; error patterns such as initial consonant deletion) that most effect intelligibility.

See ASHA's Person-Centered Focus on Function: Speech Sound Disorder [PDF] for an example of goal setting consistent with ICF.

Treatment Strategies

In addition to selecting appropriate targets for therapy, SLPs select treatment strategies based on the number of intervention goals to be addressed in each session and the manner in which these goals are implemented. A particular strategy may not be appropriate for all children, and strategies may change throughout the course of intervention as the child's needs change.

"Target attack" strategies include the following:

• Vertical —intense practice on one or two targets until the child reaches a specific criterion level (usually conversational level) before proceeding to the next target or targets (see, e.g., Fey, 1986).
• Horizontal —less intense practice on a few targets; multiple targets are addressed individually or interactively in the same session, thus providing exposure to more aspects of the sounds system (see, e.g., Fey, 1986).
• Cyclical —incorporating elements of both horizontal and vertical structures; the child is provided with practice on a given target or targets for some predetermined period of time before moving on to another target or targets for a predetermined period of time. Practice then cycles through all targets again (see, e.g., Hodson, 2010).

Treatment Options

The following are brief descriptions of both general and specific treatments for children with speech sound disorders. These approaches can be used to treat speech sound problems in a variety of populations. See Speech Characteristics: Selected Populations [PDF] for a brief summary of selected populations and characteristic speech problems.

Treatment selection will depend on a number of factors, including the child's age, the type of speech sound errors, the severity of the disorder, and the degree to which the disorder affects overall intelligibility (Williams, McLeod, & McCauley, 2010). This list is not exhaustive, and inclusion does not imply an endorsement from ASHA.

Contextual Utilization Approaches

Contextual utilization approaches recognize that speech sounds are produced in syllable-based contexts in connected speech and that some (phonemic/phonetic) contexts can facilitate correct production of a particular sound.

Contextual utilization approaches may be helpful for children who use a sound inconsistently and need a method to facilitate consistent production of that sound in other contexts. Instruction for a particular sound is initiated in the syllable context(s) where the sound can be produced correctly (McDonald, 1974). The syllable is used as the building block for practice at more complex levels.

For example, production of a "t" may be facilitated in the context of a high front vowel, as in "tea" (Bernthal et al., 2017). Facilitative contexts or "likely best bets" for production can be identified for voiced, velar, alveolar, and nasal consonants. For example, a "best bet" for nasal consonants is before a low vowel, as in "mad" (Bleile, 2002).

Phonological Contrast Approaches

Phonological contrast approaches are frequently used to address phonological error patterns. They focus on improving phonemic contrasts in the child's speech by emphasizing sound contrasts necessary to differentiate one word from another. Contrast approaches use contrasting word pairs as targets instead of individual sounds.

There are four different contrastive approaches— minimal oppositions, maximal oppositions , treatment of the empty set, and multiple oppositions.

• Minimal Oppositions (also known as "minimal pairs" therapy)—uses pairs of words that differ by only one phoneme or single feature signaling a change in meaning. Minimal pairs are used to help establish contrasts not present in the child's phonological system (e.g., "door" vs. "sore," "pot" vs. "spot," "key" vs. "tea"; Blache, Parsons, & Humphreys, 1981; Weiner, 1981).
• Maximal Oppositions —uses pairs of words containing a contrastive sound that is maximally distinct and varies on multiple dimensions (e.g., voice, place, and manner) to teach an unknown sound. For example, "mall" and "call" are maximal pairs because /m/ and /k/ vary on more than one dimension—/m/ is a bilabial voiced nasal, whereas /k/ is a velar voiceless stop (Gierut, 1989, 1990, 1992). See Place, Manner and Voicing Chart for English Consonants (Roth & Worthington, 2018) .
• Treatment of the Empty Set —similar to the maximal oppositions approach but uses pairs of words containing two maximally opposing sounds (e.g., /r/ and /d/) that are unknown to the child (e.g., "row" vs. "doe" or "ray" vs. "day"; Gierut, 1992).
• Multiple Oppositions —a variation of the minimal oppositions approach but uses pairs of words contrasting a child's error sound with three or four strategically selected sounds that reflect both maximal classification and maximal distinction (e.g., "door," "four," "chore," and "store," to reduce backing of /d/ to /g/; Williams, 2000a, 2000b).

Complexity Approach

The complexity approach is a speech production approach based on data supporting the view that the use of more complex linguistic stimuli helps promote generalization to untreated but related targets.

The complexity approach grew primarily from the maximal oppositions approach. However, it differs from the maximal oppositions approach in a number of ways. Rather than selecting targets on the basis of features such as voice, place, and manner, the complexity of targets is determined in other ways. These include hierarchies of complexity (e.g., clusters, fricatives, and affricates are more complex than other sound classes) and stimulability (i.e., sounds with the lowest levels of stimulability are most complex). In addition, although the maximal oppositions approach trains targets in contrasting word pairs, the complexity approach does not. See Baker and Williams (2010) and Peña-Brooks and Hegde (2015) for detailed descriptions of the complexity approach.

Core Vocabulary Approach

A core vocabulary approach focuses on whole-word production and is used for children with inconsistent speech sound production who may be resistant to more traditional therapy approaches.

Words selected for practice are those used frequently in the child's functional communication. A list of frequently used words is developed (e.g., based on observation, parent report, and/or teacher report), and a number of words from this list are selected each week for treatment. The child is taught his or her "best" word production, and the words are practiced until consistently produced (Dodd, Holm, Crosbie, & McIntosh, 2006).

Cycles Approach

The cycles approach targets phonological pattern errors and is designed for children with highly unintelligible speech who have extensive omissions, some substitutions, and a restricted use of consonants.

Treatment is scheduled in cycles ranging from 5 to 16 weeks. During each cycle, one or more phonological patterns are targeted. After each cycle has been completed, another cycle begins, targeting one or more different phonological patterns. Recycling of phonological patterns continues until the targeted patterns are present in the child's spontaneous speech (Hodson, 2010; Prezas & Hodson, 2010).

The goal is to approximate the gradual typical phonological development process. There is no predetermined level of mastery of phonemes or phoneme patterns within each cycle; cycles are used to stimulate the emergence of a specific sound or pattern—not to produce mastery of it.

Distinctive Feature Therapy

Distinctive feature therapy focuses on elements of phonemes that are lacking in a child's repertoire (e.g., frication, nasality, voicing, and place of articulation) and is typically used for children who primarily substitute one sound for another. See Place, Manner and Voicing Chart for English Consonants (Roth & Worthington, 2018) .

Distinctive feature therapy uses targets (e.g., minimal pairs) that compare the phonetic elements/features of the target sound with those of its substitution or some other sound contrast. Patterns of features can be identified and targeted; producing one target sound often generalizes to other sounds that share the targeted feature (Blache & Parsons, 1980; Blache et al., 1981; Elbert & McReynolds, 1978; McReynolds & Bennett, 1972; Ruder & Bunce, 1981).

Metaphon Therapy

Metaphon therapy is designed to teach metaphonological awareness —that is, the awareness of the phonological structure of language. This approach assumes that children with phonological disorders have failed to acquire the rules of the phonological system.

The focus is on sound properties that need to be contrasted. For example, for problems with voicing, the concept of "noisy" (voiced) versus "quiet" (voiceless) is taught. Targets typically include processes that affect intelligibility, can be imitated, or are not seen in typically developing children of the same age (Dean, Howell, Waters, & Reid, 1995; Howell & Dean, 1994).

Naturalistic Speech Intelligibility Intervention

Naturalist speech intelligibility intervention addresses the targeted sound in naturalistic activities that provide the child with frequent opportunities for the sound to occur. For example, using a McDonald's menu, signs at the grocery store, or favorite books, the child can be asked questions about words that contain the targeted sound(s). The child's error productions are recast without the use of imitative prompts or direct motor training. This approach is used with children who are able to use the recasts effectively (Camarata, 2010).

Nonspeech Oral–Motor Therapy

Nonspeech oral–motor therapy involves the use of oral-motor training prior to teaching sounds or as a supplement to speech sound instruction. The rationale behind this approach is that (a) immature or deficient oral-motor control or strength may be causing poor articulation and (b) it is necessary to teach control of the articulators before working on correct production of sounds. Consult systematic reviews of this treatment to help guide clinical decision making (see, e.g., Lee & Gibbon, 2015 [PDF]; McCauley, Strand, Lof, Schooling, & Frymark, 2009 ). See also the Treatment section of the Speech Sound Disorders Evidence Map filtered for Oral–Motor Exercises .

Speech Sound Perception Training

Speech sound perception training is used to help a child acquire a stable perceptual representation for the target phoneme or phonological structure. The goal is to ensure that the child is attending to the appropriate acoustic cues and weighting them according to a language-specific strategy (i.e., one that ensures reliable perception of the target in a variety of listening contexts).

Recommended procedures include (a) auditory bombardment in which many and varied target exemplars are presented to the child, sometimes in a meaningful context such as a story and often with amplification, and (b) identification tasks in which the child identifies correct and incorrect versions of the target (e.g., "rat" is a correct exemplar of the word corresponding to a rodent, whereas "wat" is not).

Tasks typically progress from the child judging speech produced by others to the child judging the accuracy of his or her own speech. Speech sound perception training is often used before and/or in conjunction with speech production training approaches. See Rvachew, 1994; Rvachew et al., 2004; Rvachew, Rafaat, & Martin, 1999; Wolfe, Presley, & Mesaris, 2003.

Traditionally, the speech stimuli used in these tasks are presented via live voice by the SLP. More recently, computer technology has been used—an advantage of this approach is that it allows for the presentation of more varied stimuli representing, for example, multiple voices and a range of error types.

Treatment Techniques and Technologies

Techniques used in therapy to increase awareness of the target sound and/or provide feedback about placement and movement of the articulators include the following:

• Using a mirror for visual feedback of place and movement of articulators
• Using gestural cueing for place or manner of production (e.g., using a long, sweeping hand gesture for fricatives vs. a short, "chopping" gesture for stops)
• Using ultrasound imaging (placement of an ultrasound transducer under the chin) as a biofeedback technique to visualize tongue position and configuration (Adler-Bock, Bernhardt, Gick, & Bacsfalvi, 2007; Lee, Wrench, & Sancibrian, 2015; Preston, Brick, & Landi, 2013; Preston et al., 2014)
• Using palatography (various coloring agents or a palatal device with electrodes) to record and visualize contact of the tongue on the palate while the child makes different speech sounds (Dagenais, 1995; Gibbon, Stewart, Hardcastle, & Crampin, 1999; Hitchcock, McAllister Byun, Swartz, & Lazarus, 2017)
• Amplifying target sounds to improve attention, reduce distractibility, and increase sound awareness and discrimination—for example, auditory bombardment with low-level amplification is used with the cycles approach at the beginning and end of each session to help children perceive differences between errors and target sounds (Hodson, 2010)
• Providing spectral biofeedback through a visual representation of the acoustic signal of speech (McAllister Byun & Hitchcock, 2012)
• Providing tactile biofeedback using tools, devices, or substances placed within the mouth (e.g., tongue depressors, peanut butter) to provide feedback on correct tongue placement and coordination (Altshuler, 1961; Leonti, Blakeley, & Louis, 1975; Shriberg, 1980)

Considerations for Treating Bilingual/Multilingual Populations

When treating a bilingual or multilingual individual with a speech sound disorder, the clinician is working with two or more different sound systems. Although there may be some overlap in the phonemic inventories of each language, there will be some sounds unique to each language and different phonemic rules for each language.

One linguistic sound system may influence production of the other sound system. It is the role of the SLP to determine whether any observed differences are due to a true communication disorder or whether these differences represent variations of speech associated with another language that a child speaks.

Strategies used when designing a treatment protocol include

• determining whether to use a bilingual or cross-linguistic approach (see ASHA's Practice Portal page on Multilingual Service Delivery in Audiology and Speech-Language Pathology );
• determining the language in which to provide services, on the basis of factors such as language history, language use, and communicative needs;
• identifying alternative means of providing accurate models for target phonemes that are unique to the child's language, when the clinician is unable to do so; and
• noting if success generalizes across languages throughout the treatment process (Goldstein & Fabiano, 2007).

Considerations for Treatment in Schools

Criteria for determining eligibility for services in a school setting are detailed in the Individuals with Disabilities Education Improvement Act of 2004 (IDEA). In accordance with these criteria, the SLP needs to determine

• if the child has a speech sound disorder;
• if there is an adverse effect on educational performance resulting from the disability; and
• if specially designed instruction and/or related services and supports are needed to help the student make progress in the general education curriculum.

Examples of the adverse effect on educational performance include the following:

• The speech sound disorder affects the child's ability or willingness to communicate in the classroom (e.g., when responding to teachers' questions; during classroom discussions or oral presentations) and in social settings with peers (e.g., interactions during lunch, recess, physical education, and extracurricular activities).
• The speech sound disorder signals problems with phonological skills that affect spelling, reading, and writing. For example, the way a child spells a word reflects the errors made when the word is spoken. See ASHA's resource language in brief and ASHA's Practice Portal pages on Spoken Language Disorders and Written Language Disorders for more information about the relationship between spoken and written language

Eligibility for speech-language pathology services is documented in the child's individualized education program, and the child's goals and the dismissal process are explained to parents and teachers. For more information about eligibility for services in the schools, see ASHA's resources on eligibility and dismissal in schools , IDEA Part B Issue Brief: Individualized Education Programs and Eligibility for Services , and 2011 IDEA Part C Final Regulations .

If a child is not eligible for services under IDEA, they may still be eligible to receive services under the Rehabilitation Act of 1973, Section 504. 29 U.S.C. § 701 (1973) . See ASHA's Practice Portal page on Documentation in Schools for more information about Section 504 of the Rehabilitation Act of 1973.

Dismissal from speech-language pathology services occurs once eligibility criteria are no longer met—that is, when the child's communication problem no longer adversely affects academic achievement and functional performance.

Children With Persisting Speech Difficulties

Speech difficulties sometimes persist throughout the school years and into adulthood. Pascoe et al. (2006) define persisting speech difficulties as "difficulties in the normal development of speech that do not resolve as the child matures or even after they receive specific help for these problems" (p. 2). The population of children with persistent speech difficulties is heterogeneous, varying in etiology, severity, and nature of speech difficulties (Dodd, 2005; Shriberg et al., 2010; Stackhouse, 2006; Wren, Roulstone, & Miller, 2012).

A child with persisting speech difficulties (functional speech sound disorders) may be at risk for

• difficulty communicating effectively when speaking;
• difficulty acquiring reading and writing skills; and
• psychosocial problems (e.g., low self-esteem, increased risk of bullying; see, e.g., McCormack, McAllister, McLeod, & Harrison, 2012).

Intervention approaches vary and may depend on the child's area(s) of difficulty (e.g., spoken language, written language, and/or psychosocial issues).

In designing an effective treatment protocol, the SLP considers

• teaching and encouraging the use of self-monitoring strategies to facilitate consistent use of learned skills;
• collaborating with teachers and other school personnel to support the child and to facilitate his or her access to the academic curriculum; and
• managing psychosocial factors, including self-esteem issues and bullying (Pascoe et al., 2006).

Transition Planning

Children with persisting speech difficulties may continue to have problems with oral communication, reading and writing, and social aspects of life as they transition to post-secondary education and vocational settings (see, e.g., Carrigg, Baker, Parry, & Ballard, 2015). The potential impact of persisting speech difficulties highlights the need for continued support to facilitate a successful transition to young adulthood. These supports include the following:

• Transition Planning —the development of a formal transition plan in middle or high school that includes discussion of the need for continued therapy, if appropriate, and supports that might be needed in postsecondary educational and/or vocational settings (IDEA, 2004).
• Disability Support Services —individualized support for postsecondary students that may include extended time for tests, accommodations for oral speaking assignments, the use of assistive technology (e.g., to help with reading and writing tasks), and the use of methods and devices to augment oral communication, if necessary.

The Americans with Disabilities Act of 1990 (ADA) and Section 504 of the Rehabilitation Act of 1973 provide protections for students with disabilities who are transitioning to postsecondary education. The protections provided by these acts (a) ensure that programs are accessible to these students and (b) provide aids and services necessary for effective communication (U.S. Department of Education, Office for Civil Rights, 2011).

For more information about transition planning, see ASHA's resource on Postsecondary Transition Planning .

Service Delivery

See the Service Delivery section of the Speech Sound Disorders Evidence Map for pertinent scientific evidence, expert opinion, and client/caregiver perspective.

In addition to determining the type of speech and language treatment that is optimal for children with speech sound disorders, SLPs consider the following other service delivery variables that may have an impact on treatment outcomes:

• Dosage —the frequency, intensity, and duration of service
• Format —whether a person is seen for treatment one-on-one (i.e., individual) or as part of a group
• Provider —the person administering the treatment (e.g., SLP, trained volunteer, caregiver)
• Setting —the location of treatment (e.g. home, community-based, school [pull-out or within the classroom])
• Timing —when intervention occurs relative to the diagnosis.

Technology can be incorporated into the delivery of services for speech sound disorders, including the use of telepractice as a format for delivering face-to-face services remotely. See ASHA's Practice Portal page on Telepractice .

The combination of service delivery factors is important to consider so that children receive optimal intervention intensity to ensure that efficient, effective change occurs (Baker, 2012; Williams, 2012).

ASHA Resources

• Consumer Information: Speech Sound Disorders
• Interprofessional Education/Interprofessional Practice (IPE/IPP)
• Let's Talk: For People With Special Communication Needs
• Person- and Family-Centered Care
• Person-Centered Focus on Function: Speech Sound Disorder [PDF]
• Phonemic Inventories and Cultural and Linguistic Information Across Languages
• Postsecondary Transition Planning
• Selected Phonological Processes (Patterns)

Other Resources

• Age of Acquisition of English Consonants (Crowe & McLeod, 2020) [PDF]
• American Cleft Palate–Craniofacial Association
• English Consonant and Vowel Charts (University of Arizona)
• Everyone Has an Accent
• Free Resources for the Multiple Oppositions approach - Adventures in Speech Pathology
• Multilingual Children's Speech: Overview
• Multilingual Children's Speech: Intelligibility in Context Scale
• Multilingual Children's Speech: Speech Participation and Activity Assessment of Children (SPAA-C)
• Phonetics: The Sounds of American English (University of Iowa)
• Phonological and Phonemic Awareness
• Place, Manner and Voicing Chart for English Consonants (Roth & Worthington, 2018)
• RCSLT: New Long COVID Guidance and Patient Handbook
• The Development of Phonological Skills (WETA Educational Website)
• The Speech Accent Archive (George Mason University)

Adler-Bock, M., Bernhardt, B. M., Gick, B., & Bacsfalvi, P. (2007). The use of ultrasound in remediation of North American English /r/ in 2 adolescents. American Journal of Speech-Language Pathology, 16, 128–139.

Altshuler, M. W. (1961). A therapeutic oral device for lateral emission. Journal of Speech and Hearing Disorders, 26, 179–181.

American Speech-Language-Hearing Association. (2016a). Code of ethics [Ethics]. Available from www.asha.org/policy/

American Speech-Language-Hearing Association. (20016b). Scope of practice in speech-language-pathology [Scope of Practice]. Available from www.asha.org/policy/

Americans with Disabilities Act of 1990, P.L. 101-336, 42 U.S.C. §§ 12101 et seq.

Anthony, J. L., Aghara, R. G., Dunkelberger, M. J., Anthony, T. I., Williams, J. M., & Zhang, Z. (2011). What factors place children with speech sound disorders at risk for reading problems? American Journal of Speech-Language Pathology, 20, 146–160.

Baker, E. (2012). Optimal intervention intensity. International Journal of Speech-Language Pathology, 14, 401–409.

Baker, E., & Williams, A. L. (2010). Complexity approaches to intervention. In S. F. Warren & M. E. Fey (Series Eds.). & A. L. Williams, S. McLeod, & R. J. McCauley (Volume Eds.), Intervention for speech sound disorders in children (pp. 95–115). Baltimore, MD: Brookes.

Bernthal, J., Bankson, N. W., & Flipsen, P., Jr. (2017). Articulation and phonological disorders: Speech sound disorders in children . New York, NY: Pearson.

Blache, S., & Parsons, C. (1980). A linguistic approach to distinctive feature training. Language, Speech, and Hearing Services in Schools, 11, 203–207.

Blache, S. E., Parsons, C. L., & Humphreys, J. M. (1981). A minimal-word-pair model for teaching the linguistic significant difference of distinctive feature properties. Journal of Speech and Hearing Disorders, 46, 291–296.

Black, L. I., Vahratian, A., & Hoffman, H. J. (2015). Communication disorders and use of intervention services among children aged 3–17 years; United States, 2012 (NHS Data Brief No. 205). Hyattsville, MD: National Center for Health Statistics.

Bleile, K. (2002). Evaluating articulation and phonological disorders when the clock is running. American Journal of Speech-Language Pathology, 11, 243–249.

Byers Brown, B., Bendersky, M., & Chapman, T. (1986). The early utterances of preterm infants. British Journal of Communication Disorders, 21, 307–320.

Camarata, S. (2010). Naturalistic intervention for speech intelligibility and speech accuracy. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 381–406). Baltimore, MD: Brookes.

Campbell, T. F., Dollaghan, C. A., Rockette, H. E., Paradise, J. L., Feldman, H. M., Shriberg, L. D., . . . Kurs-Lasky, M. (2003). Risk factors for speech delay of unknown origin in 3-year-old children. Child Development, 74, 346–357.

Carrigg, B., Baker, E., Parry, L., & Ballard, K. J. (2015). Persistent speech sound disorder in a 22-year-old male: Communication, educational, socio-emotional, and vocational outcomes. Perspectives on School-Based Issues, 16, 37–49.

Carroll, J. M., Snowling, M. J., Stevenson, J., & Hulme, C. (2003). The development of phonological awareness in preschool children. Developmental Psychology, 39, 913–923.

Catts, H. W., McIlraith, A., Bridges, M. S., & Nielsen, D. C. (2017). Viewing a phonological deficit within a multifactorial model of dyslexia. Reading and Writing, 30, 613–629.

Coplan, J., & Gleason, J. R. (1988). Unclear speech: Recognition and significance of unintelligible speech in preschool children. Pediatrics, 82, 447–452.

Crowe, K., & McLeod, S. (2020). Children’s English consonant acquisition in the United States: A review. American Journal of Speech-Language Pathology, 29 (4), 2155-2169. https://doi.org/10.1044/2020_AJSLP-19-00168.

Dagenais, P. A. (1995). Electropalatography in the treatment of articulation/phonological disorders. Journal of Communication Disorders, 28, 303–329.

Dean, E., Howell, J., Waters, D., & Reid, J. (1995). Metaphon: A metalinguistic approach to the treatment of phonological disorder in children. Clinical Linguistics & Phonetics, 9, 1–19.

Dodd, B. (2005). Differential diagnosis and treatment of children with speech disorder . London, England: Whurr.

Dodd, B., Holm, A., Crosbie, S., & McIntosh, B. (2006). A core vocabulary approach for management of inconsistent speech disorder. International Journal of Speech-Language Pathology, 8, 220–230.

Eadie, P., Morgan, A., Ukoumunne, O. C., Eecen, K. T., Wake, M., & Reilly, S. (2015). Speech sound disorder at 4 years: Prevalence, comorbidities, and predictors in a community cohort of children. Developmental Medicine & Child Neurology, 57, 578–584.

Elbert, M., & McReynolds, L. V. (1978). An experimental analysis of misarticulating children's generalization. Journal of Speech and Hearing Research, 21, 136–149.

Ertmer, D. J. (2010). Relationship between speech intelligibility and word articulation scores in children with hearing loss. Journal of Speech, Language, and Hearing Research, 53, 1075–1086.

Everhart, R. (1960). Literature survey of growth and developmental factors in articulation maturation. Journal of Speech and Hearing Disorders, 25, 59–69.

Fabiano-Smith, L., & Goldstein, B. A. (2010). Phonological acquisition in bilingual Spanish–English speaking children. Journal of Speech, Language, and Hearing Research, 53, 1 60–178.

Felsenfeld, S., McGue, M., & Broen, P. A. (1995). Familial aggregation of phonological disorders: Results from a 28-year follow-up. Journal of Speech, Language, and Hearing Research, 38, 1091–1107.

Fey, M. (1986). Language intervention with young children . Boston, MA: Allyn & Bacon.

Flipsen, P. (2006). Measuring the intelligibility of conversational speech in children. Clinical Linguistics & Phonetics, 20, 202–312.

Flipsen, P. (2015). Emergence and prevalence of persistent and residual speech errors. Seminars in Speech Language, 36, 217–223.

Fox, A. V., Dodd, B., & Howard, D. (2002). Risk factors for speech disorders in children. International Journal of Language and Communication Disorders, 37, 117–132.

Gibbon, F., Stewart, F., Hardcastle, W. J., & Crampin, L. (1999). Widening access to electropalatography for children with persistent sound system disorders. American Journal of Speech-Language Pathology, 8, 319–333.

Gierut, J. A. (1989). Maximal opposition approach to phonological treatment. Journal of Speech and Hearing Research, 54, 9–19.

Gierut, J. A. (1990). Differential learning of phonological oppositions. Journal of Speech and Hearing Research, 33, 540–549.

Gierut, J. A. (1992). The conditions and course of clinically induced phonological change. Journal of Speech and Hearing Research, 35, 1049–1063.

Gierut, J. A. (2007). Phonological complexity and language learnability. American Journal of Speech-Language Pathology, 16, 6–17.

Glaspey, A. M., & Stoel-Gammon, C. (2007). A dynamic approach to phonological assessment. Advances in Speech-Language Pathology, 9, 286–296.

Goldstein, B. A., & Fabiano, L. (2007, February 13). Assessment and intervention for bilingual children with phonological disorders. The ASHA Leader, 12, 6–7, 26–27, 31.

Grunwell, P. (1987). Clinical phonology (2nd ed.). London, England: Chapman and Hall.

Hitchcock, E. R., McAllister Byun, T., Swartz, M., & Lazarus, R. (2017). Efficacy of electropalatography for treating misarticulations of /r/. American Journal of Speech-Language Pathology, 26, 1141–1158.

Hodson, B. (2010). Evaluating and enhancing children's phonological systems: Research and theory to practice. Wichita, KS: PhonoComp.

Howell, J., & Dean, E. (1994). Treating phonological disorders in children: Metaphon—Theory to practice (2nd ed.). London, England: Whurr.

Individuals with Disabilities Education Improvement Act of 2004, P. L. 108-446, 20 U.S.C. §§ 1400 et seq. Retrieved from http://idea.ed.gov/

Kent, R. D., Miolo, G., & Bloedel, S. (1994). The intelligibility of children's speech: A review of evaluation procedures. American Journal of Speech-Language Pathology, 3, 81–95.

Law, J., Boyle, J., Harris, F., Harkness, A., & Nye, C. (2000). Prevalence and natural history of primary speech and language delay: Findings from a systematic review of the literature. International Journal of Language and Communication Disorders, 35, 165–188.

Lee, A. S. Y., & Gibbon, F. E. (2015). Non-speech oral motor treatment for children with developmental speech sound disorders. Cochrane Database of Systematic Reviews, 2015 (3), 1–42.

Lee, S. A. S., Wrench, A., & Sancibrian, S. (2015). How to get started with ultrasound technology for treatment of speech sound disorders. Perspectives on Speech Science and Orofacial Disorders, 25, 66–80.

Leitão, S., & Fletcher, J. (2004). Literacy outcomes for students with speech impairment: Long-term follow-up. International Journal of Language and Communication Disorders, 39, 245–256.

Leonti, S., Blakeley, R., & Louis, H. (1975, November). Spontaneous correction of resistant /ɚ/ and /r/ using an oral prosthesis. Paper presented at the annual convention of the American Speech and Hearing Association, Washington, DC.

Lewis, B. A., Avrich, A. A., Freebairn, L. A., Hansen, A. J., Sucheston, L. E., Kuo, I., . . . Stein, C. M. (2011). Literacy outcomes of children with early childhood speech sound disorders: Impact of endophenotypes. Journal of Speech, Language, and Hearing Research, 54, 1628–1643.

Locke, J. (1980). The inference of speech perception in the phonologically disordered child. Part I: A rationale, some criteria, the conventional tests. Journal of Speech and Hearing Disorders, 45, 431–444.

McAllister Byun, T., & Hitchcock, E. R. (2012). Investigating the use of traditional and spectral biofeedback approaches to intervention for /r/ misarticulation. American Journal of Speech-Language Pathology, 21, 207–221.

McCauley, R. J., Strand, E., Lof, G. L., Schooling, T., & Frymark, T. (2009). Evidence-based systematic review: Effects of nonspeech oral motor exercises on speech. American Journal of Speech-Language Pathology, 18, 343–360.

McCormack, J., McAllister, L., McLeod, S., & Harrison, L. (2012). Knowing, having, doing: The battles of childhood speech impairment. Child Language Teaching and Therapy, 28, 141–157.

McDonald, E. T. (1974). Articulation testing and treatment: A sensory motor approach. Pittsburgh, PA: Stanwix House.

McLeod, S., & Crowe, K. (2018). Children's consonant acquisition in 27 languages: A cross-linguistic review. American Journal of Speech-Language Pathology, 27, 1546–1571.

McLeod, S., Verdon, S., & The International Expert Panel on Multilingual Children's Speech. (2017). Tutorial: Speech assessment for multilingual children who do not speak the same language(s) as the speech-language pathologist. American Journal of Speech-Language Pathology, 26, 691–708.

McNeill, B. C., & Hesketh, A. (2010). Developmental complexity of the stimuli included in mispronunciation detection tasks. International Journal of Language & Communication Disorders, 45, 72–82.

McReynolds, L. V., & Bennett, S. (1972). Distinctive feature generalization in articulation training. Journal of Speech and Hearing Disorders, 37, 462–470.

Morley, D. (1952). A ten-year survey of speech disorders among university students. Journal of Speech and Hearing Disorders, 17, 25–31.

National Institute for Literacy. (2009). Developing early literacy: Report of the National Early Literacy Panel. A scientific synthesis of early literacy development and implications for intervention. Washington, DC: U.S. Department of Education.

National Institute of Child Health and Human Development. (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction [Report of the National Reading Panel]. Washington, DC: Author.

Overby, M.S., Trainin, G., Smit, A. B., Bernthal, J. E., & Nelson, R. (2012). Preliteracy speech sound production skill and later literacy outcomes: A study using the Templin Archive. Language, Speech, and Hearing Services in Schools, 43, 97–115.

Pascoe, M., Stackhouse, J., & Wells, B. (2006). Persisting speech difficulties in children: Children's speech and literacy difficulties, Book 3. West Sussex, England: Whurr.

Peña-Brooks, A., & Hegde, M. N. (2015). Assessment and treatment of articulation and phonological disorders in children . Austin, TX: Pro-Ed.

Peterson, R. L., Pennington, B. F., Shriberg, L. D., & Boada, R. (2009). What influences literacy outcome in children with speech sound disorder? Journal of Speech, Language, and Hearing Research, 52, 1175-1188.

Powell, T. W., & Miccio, A. W. (1996). Stimulability: A useful clinical tool. Journal of Communication Disorders, 29, 237–253.

Preston, J. L., Brick, N., & Landi, N. (2013). Ultrasound biofeedback treatment for persisting childhood apraxia of speech. American Journal of Speech-Language Pathology, 22, 627–643.

Preston, J. L., McCabe, P., Rivera-Campos, A., Whittle, J. L., Landry, E., & Maas, E. (2014). Ultrasound visual feedback treatment and practice variability for residual speech sound errors. Journal of Speech, Language, and Hearing Research, 57, 2102–2115.

Prezas, R. F., & Hodson, B. W. (2010). The cycles phonological remediation approach. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 137–158). Baltimore, MD: Brookes.

Section 504 of the Rehabilitation Act of 1973, P.L. No. 93-112, 29 U.S.C. § 794.

Roth, F. P., & Worthington, C. K. (2018). Treatment resource manual for speech-language pathology . San Diego, CA: Plural Publishing.

Ruder, K. F., & Bunce, B. H. (1981). Articulation therapy using distinctive feature analysis to structure the training program: Two case studies. Journal of Speech and Hearing Disorders, 46, 59–65.

Rvachew, S. (1994). Speech perception training can facilitate sound production learning. Journal of Speech, Language, and Hearing Research, 37, 347–357.

Rvachew, S., & Bernhardt, B. M. (2010). Clinical implications of dynamic systems theory for phonological development. American Journal of Speech-Language Pathology, 19, 34–50.

Rvachew, S., Nowak, M., & Cloutier, G. (2004). Effect of phonemic perception training on the speech production and phonological awareness skills of children with expressive phonological delay. American Journal of Speech-Language Pathology, 13, 250–263.

Rvachew, S., Rafaat, S., & Martin, M. (1999). Stimulability, speech perception skills, and treatment of phonological disorders. American Journal of Speech-Language Pathology, 8, 33–43.

Shriberg, L. D. (1980). An intervention procedure for children with persistent /r/ errors. Language, Speech, and Hearing Services in Schools, 11, 102–110.

Shriberg, L. D., & Austin, D. (1998). Comorbidity of speech-language disorders: Implications for a phenotype marker for speech delay. In R. Paul (Ed.), The speech-language connection (pp. 73–117). Baltimore, MD: Brookes.

Shriberg, L. D., Austin, D., Lewis, B., McSweeny, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722.

Shriberg, L. D., Fourakis, M., Hall, S. D., Karlsson, H. B., Lohmeier, H. L., McSweeny, J. L., . . . Wilson, D. L. (2010). Extensions to the speech disorders classification system (SDCS). Clinical Linguistics & Phonetics, 24, 795–824.

Shriberg, L. D., & Kwiatkowski, J. (1980). Natural Process Analysis (NPA): A procedure for phonological analysis of continuous speech samples. New York, NY: Macmillan.

Shriberg, L. D., & Kwiatkowski, J. (1982a). Phonological disorders II: A conceptual framework for management. Journal of Speech and Hearing Disorders, 47, 242–256.

Shriberg, L. D., & Kwiatkowski, J. (1982b). Phonological disorders III: A procedure for assessing severity of involvement. Journal of Speech and Hearing Disorders, 47, 256–270.

Shriberg, L. D., & Kwiatkowski, J. (1994). Developmental phonological disorders I: A clinical profile. Journal of Speech and Hearing Research, 37, 1100–1126.

Shriberg, L. D., Tomblin, J. B., & McSweeny, J. L. (1999). Prevalence of speech delay in 6-year-old children and comorbidity with language impairment. Journal of Speech, Language, and Hearing Research, 42, 1461–1481.

Silva, P. A., Chalmers, D., & Stewart, I. (1986). Some audiological, psychological, educational and behavioral characteristics of children with bilateral otitis media with effusion. Journal of Learning Disabilities, 19, 165–169.

Stackhouse, J. (2006). Speech and spelling difficulties: Who is at risk and why? In M. Snowling & J. Stackhouse (Eds.), Dyslexia speech and language: A practitioner's handbook (pp. 15–35). West Sussex, England: Whurr.

Storkel, H. L. (2018). The complexity approach to phonological treatment: How to select treatment targets. Language, Speech, and Hearing Sciences in Schools, 49, 463–481.

Storkel, Holly. (2019). Using Developmental Norms for Speech Sounds as a Means of Determining Treatment Eligibility in Schools. Perspectives of the ASHA Special Interest Groups, 4, 67-75.

Teele, D. W., Klein, J. O., Chase, C., Menyuk, P., & Rosner, B. A. (1990). Otitis media in infancy and intellectual ability, school achievement, speech, and language at 7 years. Journal of Infectious Disease, 162, 685–694.

Tyler, A. A., & Tolbert, L. C. (2002). Speech-language assessment in the clinical setting. American Journal of Speech-Language Pathology, 11, 215–220.

U.S. Department of Education, Office for Civil Rights. (2011). Transition of students with disabilities to postsecondary education: A guide for high school educators . Washington, DC: Author. Retrieved from http://www2.ed.gov/about/offices/list/ocr/transitionguide.html

Wagner, R. K., & Torgesen, J. K. (1987). The nature of phonological processing and its causal role in the acquisition of reading skills. Psychological Bulletin, 101, 192–212.

Weiner, F. (1981). Treatment of phonological disability using the method of meaningful minimal contrast: Two case studies. Journal of Speech and Hearing Disorders, 46, 97–103.

Williams, A. L. (2000a). Multiple oppositions: Case studies of variables in phonological intervention. American Journal of Speech-Language Pathology, 9, 289–299.

Williams, A. L. (2000b). Multiple oppositions: Theoretical foundations for an alternative contrastive intervention approach. American Journal of Speech-Language Pathology, 9, 282–288.

Williams, A. L. (2003a). Speech disorders resource guide for preschool children. Clifton Park, NY: Delmar Learning.

Williams, A. L. (2003b). Target selection and treatment outcomes. Perspectives on Language Learning and Education, 10, 12–16.

Williams, A. L. (2012). Intensity in phonological intervention: Is there a prescribed amount? International Journal of Speech-Language Pathology, 14, 456–461.

Williams, A. L., McLeod, S., & McCauley, R. J. (2010). Direct speech production intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 27–39). Baltimore, MD: Brookes.

Wolfe, V., Presley, C., & Mesaris, J. (2003). The importance of sound identification training in phonological intervention. American Journal of Speech-Language Pathology, 12, 282–288.

World Health Organization. (2001). International classification of functioning, disability and health. Geneva, Switzerland: Author.

Wren, Y., Miller, L. L., Peters, T. J., Emond, A., & Roulstone, S. (2016). Prevalence and predictors of persistent speech sound disorder at eight years old: Findings from a population cohort study. Journal of Speech, Language, and Hearing Research , 59, 647–673.

Wren, Y. E., Roulstone, S. E., & Miller, L. L. (2012). Distinguishing groups of children with persistent speech disorder: Findings from a prospective population study. Logopedics Phoniatrics Vocology, 37, 1–10.

About This Content

Acknowledgements .

Content for ASHA's Practice Portal is developed through a comprehensive process that includes multiple rounds of subject matter expert input and review. ASHA extends its gratitude to the following subject matter experts who were involved in the development of the Speech Sound Disorders:  Articulation and Phonology page:

• Elise M. Baker, PhD
• John E. Bernthal, PhD, CCC-A/SLP
• Caroline Bowen, PhD
• Cynthia W. Core, PhD, CCC-SLP
• Sharon B. Hart, PhD, CCC-SLP
• Barbara W. Hodson, PhD, CCC-SLP
• Sharynne McLeod, PhD
• Susan Rvachew, PhD, S-LP(C)
• Cheryl C. Sancibrian, MS, CCC-SLP
• Holly L. Storkel, PhD, CCC-SLP
• Judith E. Trost-Cardamone, PhD, CCC-SLP
• Lynn Williams, PhD, CCC-SLP

Citing Practice Portal Pages 

The recommended citation for this Practice Portal page is:

American Speech-Language-Hearing Association (n.d.) Speech Sound Disorders: Articulation and Phonology. (Practice Portal). Retrieved month, day, year, from www.asha.org/Practice-Portal/Clinical-Topics/Articulation-and-Phonology/ .

In This Section

• Practice Portal Home
• Clinical Topics
• Professional Issues
• Advertising Disclaimer
• Advertise with us

Evidence Maps

• ASHA Evidence Maps

Peer Connections

• Connect with your colleagues in the ASHA Community
• ASHA Special Interest Groups

ASHA Related Content

• Find related products in ASHA's Store
• Search for articles on ASHAWire
• ASHA Stream

Content Disclaimer: The Practice Portal, ASHA policy documents, and guidelines contain information for use in all settings; however, members must consider all applicable local, state and federal requirements when applying the information in their specific work setting.

ASHA Corporate Partners

• Become A Corporate Partner

The American Speech-Language-Hearing Association (ASHA) is the national professional, scientific, and credentialing association for 234,000 members, certificate holders, and affiliates who are audiologists; speech-language pathologists; speech, language, and hearing scientists; audiology and speech-language pathology assistants; and students.

• All ASHA Websites
• Work at ASHA
• Marketing Solutions

Information For

Get involved.

• ASHA Community
• Become a Mentor
• Become a Volunteer
• Special Interest Groups (SIGs)

Connect With ASHA

American Speech-Language-Hearing Association 2200 Research Blvd., Rockville, MD 20850 Members: 800-498-2071 Non-Member: 800-638-8255

MORE WAYS TO CONNECT

Media Resources

• Press Queries

Site Help | A–Z Topic Index | Privacy Statement | Terms of Use © 1997- American Speech-Language-Hearing Association

• Previous Article
• Next Article

I. INTRODUCTION

Ii. methods, b. analysis, iii. bkb, hint, and ieee sentences, a. syntactic variation, b. lexical diversity, d. information sources and listener task, iv. spontaneous speech, a. communicative purpose, c. pauses and gaps, d. slowing and lengthening, e. echoing and discourse context, g. listener task, v. discussion, vi. conclusion, acknowledgments, is speech intelligibility what speech intelligibility tests test a).

Electronic mail: [email protected]

This paper is part of a special issue on Reconsidering Classic Ideas in Speech Communication.

• Split-Screen
• Article contents
• Figures & tables
• Supplementary Data
• Peer Review
• Open the PDF for in another window
• Reprints and Permissions
• Cite Icon Cite
• Search Site

Timothy Beechey; Is speech intelligibility what speech intelligibility tests test?. J. Acoust. Soc. Am. 1 September 2022; 152 (3): 1573–1585. https://doi.org/10.1121/10.0013896

Download citation file:

• Ris (Zotero)
• Reference Manager

Natural, conversational speech signals contain sources of symbolic and iconic information, both of which are necessary for the full understanding of speech. But speech intelligibility tests, which are generally derived from written language, present only symbolic information sources, including lexical semantics and syntactic structures. Speech intelligibility tests exclude almost all sources of information about talkers, including their communicative intentions and their cognitive states and processes. There is no reason to suspect that either hearing impairment or noise selectively affect perception of only symbolic information. We must therefore conclude that diagnosis of good or poor speech intelligibility on the basis of standard speech tests is based on measurement of only a fraction of the task of speech perception. This paper presents a descriptive comparison of information sources present in three widely used speech intelligibility tests and spontaneous, conversational speech elicited using a referential communication task. The aim of this comparison is to draw attention to the differences in not just the signals, but the tasks of listeners perceiving these different speech signals and to highlight the implications of these differences for the interpretation and generalizability of speech intelligibility test results.

Speech signals are of great importance for assessing hearing functioning and disability as well as the benefits of hearing devices. The use of speech signals are, in general, well motivated, since speech understanding is probably the most important function of the human auditory system. Probing the auditory system with speech signals should, it is reasoned, provide a valid measure of hearing function in everyday life. In fields such as audiology and hearing science, speech intelligibility is often quantified using standardized speech tests made up of either single words (e.g., Boothroyd, 1968 ) or simple sentences (e.g., Bench et al. , 1979 ; IEEE, 1969 ; Nilsson et al. , 1994 ). The purpose of this paper is to illustrate how such standardized speech test materials differ from spontaneous, conversational speech which is more representative of speech encountered by people in daily life. This paper is focused on one specific category of difference: information content; and one specific consequence of this difference: the extent to which a listener's task of extracting information from standard speech test materials is representative of understanding conversational speech. In particular, this paper will describe some information sources and associated functions in spontaneous speech which are absent from standard speech test materials.

Measuring speech perception involves the use of a model ( Bowman and Targowski, 1987 ). Speech perception is a highly complex behavior that is affected by many factors, all of which could not feasibly be replicated in the clinic or laboratory. It is necessary to simplify reality to measure speech perception, as is the case with any complex phenomenon. Models of speech perception may take many different forms and levels of specificity, may be defined overtly, or may be tacitly assumed. One influential model in the study of speech intelligibility is the Transmission Communication model ( Shannon, 1948 ). Using the Transmission Communication model to understand or explain speech intelligibility assumes that communication between a talker and a hearer is analogous to the transmission of a signal via a physical channel, such as a telephone line ( Cherry, 1966 ). As a model of human speech intelligibility, the Transmission Communication model involves simplifying assumptions: it defines information in terms of an alphabet of pre-defined symbols; and it explicitly excludes meaning and interpretation of these symbols as psychological rather than physical factors ( Hartley, 1928 ). The useful simplification provided by the Transmission Communication model is treating speech communication as purely syntactic–as the accurate transmission of symbols from a “dictionary” ( Cherry, 1951 ). Understanding of speech, from this perspective, might be summarized as accurate reception of units, such as phonemes or words, and their arrangement. A more general model treats written language as representative of spoken language (see Linell, 2005 , for a review). The written language model leads to a similar view of what constitutes speech perception: recovery of those linguistics units that can be represented as written symbols from the speech signal. Regardless of the particular model used, the useful simplifications of a model should not lead us to assume that the process of speech perception is correspondingly simple. As Tucker and Ernestus (2016) argue, the “common belief that language outside the laboratory is very similar to careful [read or laboratory] speech” is likely false.

Speech test results are informative, in absolute terms, to the extent that they are representative of, or at least proportional to, true speech perception abilities. However, discrepancies between clinical or laboratory results and real-world speech processing, hearing ability, hearing device benefit, and satisfaction have long been recognized (see Cord et al. , 2004 ; Cox et al. , 2016 ; Souza et al. , 2000 ; Working Group on Speech Understanding and Aging, 1988 ; Wu et al. , 2019 , for examples), along with the importance of considering more realistic speech signals (e.g., Hamilton and Huth, 2020 ; Nastase et al. , 2020 ; Tucker and Ernestus, 2016 ). To the extent that speech test results diverge from real-world speech perception ability and device benefit or satisfaction, we need to consider our models and their assumptions. It is necessary to critically evaluate what is included and what is excluded in the necessary simplification of reality.

A fundamental simplification strategy used in the design of speech testing methods and materials is to represent speech and language as consisting of discrete, atomic units which may be combined to form progressively larger discrete units ( Gatehouse, 1998 ). Phonemes are combined to form morphemes which are combined to form words and finally sentences. Representing language in this way makes it possible to score speech tests quickly and easily since the recognition of discrete units can be quantified using a simple count to produce a proportion. And, in the design of speech test materials, discrete units may be combined in specific and controlled ways to produce a desired number of test items and desired frequency distributions. This method of constructing speech test materials also makes it possible to script materials in written form. The possibility of scripting speech tests in written form indicates that auditory speech test stimuli have identical syntactic and semantic content to written language ( Tucker and Ernestus, 2016 ). It follows that auditory speech test stimuli have the same information content as written language which is itself fundamentally different from spontaneous speech ( Brazil, 1995 ). This may appear to be a common sense strategy, or even the only logical manner with which to construct speech test stimuli. However, because of what is left out by this representation of language, this simplification strategy is a primary cause of differences between standard speech test stimuli and the types of speech signals encountered by people outside the laboratory.

Natural speech encountered by people in everyday life is highly ambiguous if considered at the scale of units such as segments or even words (e.g., Ernestus et al. , 2002 ; Lieberman, 1963 ; Pickett and Pollack, 1963 ; Pollack and Pickett, 1963 ; Tucker and Ernestus, 2016 ; Warner and Tucker, 2011 ; Winitz and LaRiviere, 1979 ). People's ability to understand running speech with apparent ease, even in the face of hearing impairment ( Beechey et al. , 2020b ), stems from a complex cognitive process of information extraction, not a simple linear process of piecing together well-formed units (e.g., Nusbaum and Magnuson, 1997 ). Many important sources of information that are available to hearers in natural speech signals cannot easily be represented in the form of discrete units. This information content is evident if we consider speech production. Talkers typically do not pre-plan or rehearse what they are about to say ( Brazil, 1995 ; Lind et al. , 2014 ). Rather, speech is a real-time cognitive act that requires cognitive resources to perform ( Lee et al. , 2017 ) and that reflects, in part, the current state and cognitive demands experienced by the talker ( Berthold and Jameson, 1999 ; Griffin, 1987 ; Mendoza and Carballo, 1998 ; Yap et al. , 2011 ). Therefore, speech signals convey information about talkers that is often necessary for hearers to fully understand speech. For example, speech signals carry information about talkers' knowledge or confidence ( Jiang and Pell, 2017 ; Pon-Barry and Shieber, 2011 ), stress level, anxiety, or cognitive load ( Chen et al. , 2016 ; Cook, 1969 ), emotion ( Cummins et al. , 2015 ), judgment of current mutual knowledge shared with the listener ( Brennan and Williams, 1995 ; Owens et al. , 2018 ; Smith and Clark, 1993 ) or intention to introduce new information ( Arnold et al. , 2003 ; Barr and Seyfeddinipur, 2010 ). Because these sources of information cannot easily be represented as units, when speech materials are constructed from discrete units, these sources of information are filtered out.

The written-language basis of standard speech materials and the cognitive markers that characterize natural speech illustrate a distinction made by Halliday (1985) between product and process . Written text is a product which is static; whatever cognitive processes occur in the writing of text are not accessible to a reader. Spontaneous speech, on the other hand, is a cognitive process that occurs and is perceived in real time, and its characteristics reflect this ( Chafe, 1979 ). Standard speech test stimuli, which are readings or productions of memorized written text, are more similar to written language than spontaneous speech because the talker is not constructing the message during the speech production process. Read or memorized speech therefore lacks any meaningful information about cognitive state or process.

The information that is present in read or memorized speech is purely symbolic –consisting of words and grammatical rules for combining words into (usually) simple sentences. Symbolic links between words and syntactic rules, and the things in the world that they represent are arbitrary . This is a central tenet of linguistic theory ( de Saussure et al. , 1986 ). For example, lexical meaning of a word such as “cat” is based on convention; all English speakers have agreed that the speech sounds broadly transcribed as [kæt] represent a feline, a furry, four-legged domestic animal. However, there is no inherent connection between this sequence of speech sounds and its meaning. There is no connection between the sounds and the world.

In addition to symbolic information, spontaneous speech contains iconic information; that is, information that is inherently connected to, or informative about the listener's environment ( Perry et al. , 2018 ; Van Langendonck, 2010 ). For example, the length of a pause in a talker's speech is not informative because we agreed on its meaning but because the duration of a pause is likely proportional to the time required by the talker's cognitive process, or to their degree of uncertainty or distraction ( O'Connell and Kowal, 1983 ). To the extent that hearing science is concerned with perception of the auditory environment (including talkers) rather than language, it should be surprising that our speech tests present only symbolic information. While far less common than studies based on symbolic speech and language, iconic aspects of speech and speech perception have been investigated. For example, emotion discrimination is an important subject of investigation in hearing impaired individuals, particularly people who rely on cochlear implants, which provide highly degraded prosodic cues. Studies of cochlear implantees' perception of prosody have demonstrated reduced ability to discriminate between statement-like and question-like prosody ( Chatterjee and Peng (2008) , poorer than normal discrimination of stress ( Meister et al. , 2009 ), and the production of smaller than normal contrasts in intonation, likely as a result of degraded speech input ( Chatterjee et al. , 2019 ). Such studies are an important exception to typical investigations of speech perception in hearing science in that they seek to measure perception of aspects of speech that are not based in written language and which convey iconic information about talkers.

Although it might be argued that audiology and hearing science are concerned with understanding the auditory system rather than understanding speech signals, understanding auditory function in relation to speech depends on how well we approximate the actual task of speech perception in our assessments. If speech perception is the extraction of information from the speech signal ( Gibson, 1966 , 2015 ) then the information content of speech signals defines the task. If we fundamentally alter the information content of the signal, then the information extraction task becomes unrepresentative and measurement may be inaccurate or misleading. Understanding the characteristics of speech signals also leads to a better understanding of the process and demands of speech perception which is necessary to understand what can go wrong with speech perception, what is required for successful speech perception, and on what basis to judge whether speech perception is successful. Without an understanding of the task of speech perception, we are not in a good position to improve it. If we view the task of speech perception as limited to the extraction of symbolic information, then our definition of speech perception is simply what speech tests test . The danger of this circular definition is that it leaves us unable to detect many possible speech perception problems in patients and research participants. The result is overconfidence in our test results and an inability to predict auditory function outside clinical or laboratory settings.

The existence of non-symbolic information in natural speech has implications for the interpretation of speech test results in terms of spectro-temporal processing, since spectro-temporal information is used by hearers for more than the identification of segmental cues. For example, accepting that listeners rely on intonation to distinguish between statements and questions might lead us to accord greater importance to low frequencies. These information sources also have implications for cognitive hearing science. The study of listening effort is particularly concerned with the cognitive demands of speech perception. However, the sources of cognitive load most commonly studied in the listening effort literature are either extrinsic to the task of speech perception, such as environmental noise (e.g., Sarampalis et al. , 2009 ), attention and competing tasks ( Gagné et al. , 2017 ), or are argued to be proxy measures related to speech perception, such as working memory ( Rönnberg et al. , 2008 ). The many demands related to understanding the talker listed in this section are intrinsic to speech perception. The consideration of such factors could provide a more direct measure of cognitive processes related to speech perception, and hearing-related cognitive load ( Tucker and Ernestus, 2016 ).

Using the definition of information originating in cybernetics which holds that information is “any difference that makes a difference” ( Bateson, 2002 ), this paper will compare the information content of speech signals from commonly used speech perception tests with spontaneous, conversational speech. This comparison will illuminate the extent to which speech testing methods and materials might represent the task of perceiving conversational speech. In the most general sense, the information content of speech signals can be considered in two broad ways: (i) the presence or absence of sources of information; and (ii) the variation or diversity within information sources. If a category of cues is entirely absent from a speech signal, that category of information is unavailable to the hearer. Similarly, if a category of cues is present but is highly, or even, completely predictable due to an absence of variation, then those cues convey little or no useful information to the hearer ( Bateson, 2002 ; Hartley, 1928 ; Shannon, 1948 ). As Cherry (1966) summarizes, “Information can be received only where there is doubt; and doubt implies the existence of alternatives—where choice, selection, or discrimination is called for.”

Using the Bamford–Kowal–Bench (BKB) sentence test ( Bench et al. , 1979 ), the Hearing in Noise Test (HINT) ( Nilsson et al. , 1994 ), and the Institute of Electrical and Electronic Engineers (IEEE) sentences ( IEEE, 1969 ) as examples, it will be demonstrated that many of the choices made in the design of speech intelligibility test materials systematically emphasize symbolic information and exclude iconic information. In contrast, analysis of spontaneous, conversational speech illustrates the importance of iconic information in the task of speech perception.

Recordings of BKB sentences lists were obtained from the National Acoustic Laboratories Speech and Noise for Hearing Aid Evaluation compact discs (CDs) ( Australian Hearing, 2000 ). The sentences are spoken by a single native Australian-English speaking adult male. HINT sentence recordings spoken by a native United States English speaking male were obtained from the HINT CD ( Nilsson et al. , 1994 ). IEEE sentences were obtained from a recording of a male, native British-English talker.

Conversational speech was obtained from a 5 minute recording of diadic conversation previously reported by Beechey (2019) . Conversational speech signals were produced by two young adult female native Australian-English talkers while jointly completing a referential communication puzzle task described by Beechey et al. (2019) . The conversation elicitation task required the talkers to complete a challenging puzzle, including the need to communicate about multiple abstract, often ambiguous referents in the form of tangram images. As a result, the conversation is not fully representative of conversational speech in general. In particular, the participants likely experienced more communication difficulty than in a natural conversation and their communication was motivated, to some extent, by an extrinsic goal. The speech produced by the talkers therefore reflects relatively great cognitive demands which likely resulted in more frequent disfluencies. That is, it is probable that the speech exemplified in this conversation is different from typical conversational speech in the density rather than the type of iconic information it contains.

The data presented here may be considered a case study since data from a single conversation are presented. The goals of presenting these data are: (i) to describe the presence of information sources versus absence of such sources; and (ii) to illustrate the content of conversational speech itself, rather than to summarize this content with numbers. Both these goals are better served by illustration than by quantification, since quantification of presence versus absence is not meaningful, and a numerical summary cannot show the actual characteristics or content of conversational speech.

Acoustic analyses were conducted using Praat (version 6.2) ( Boersma and Weenink, 2022 ). Transcription of conversational speech and identification of communication events followed the conventions of Conversation Analysis ( Sacks et al. , 1974 ). See Table I for a summary of Conversation Analysis transcription symbols.

Sentences used in standard speech intelligibility tests are characterized by uniformity of syntax, prosody, and speaking style. The exception to this uniformity is lexical content that is characterized by a high level of diversity. The combination of uniformity and diversity found in the sentence test materials has the effect of maximizing the importance of symbolic information sources available to listeners and ensuring the complete absence of non-symbolic information.

The BKB and HINT materials consist of mainly subject–verb–object sentences, such as “Children like strawberries,” and “They are buying some bread.” A smaller number of sentences have intransitive verbs that do not take a direct object, as in “The car engine's running.” All sentences are simple statements and are of similar length. IEEE sentences are similarly all statements, except for a small number of rhetorical questions (e.g., “What joy there is in living”). IEEE sentences are slightly longer and more complex, including constructions with multiple adjectives (e.g., “The sky that morning was clear and bright blue”), direct and indirect objects (e.g., “The spot on the blotter was made by green ink”), and conjunctions (e.g., “The sofa cushion is red and of light weight”). In each test, all sentences are complete, i.e., there are no sentence fragments, and well-formed, i.e., each sentence is produced without any disfluencies such as pauses, filled pauses, or false starts. The sentence structures found in speech intelligibility materials are not highly representative of spoken language which often contains relatively long and complex structures ( Brazil, 1995 ; Halliday, 1985 ), in addition to frequent disfluencies ( Fox Tree, 1995 ).

In contrast to syntactic structure, which is similar across sentences, there is high lexical diversity both within and across speech intelligibility test sentences. This is a more extreme instance of the previously documented tendency for read or rehearsed speech to exhibit greater lexical diversity than spontaneous speech ( Chafe and Tannen, 1987 ) due, in part, to the tendency of interlocutors to tacitly agree on and re-use particular lexical items within conversations ( Brennan, 1996 ; Brennan and Clark, 1996 ). For the speech intelligibility test materials, each sentence in each list is made up of different non-function words including nouns, verbs, and adjectives. While some words are repeated throughout the materials, for example, in the BKB sentences, “boy” occurs 15 times, and “dog” and “mother” 10 times each, non-function words are never repeated within sentences or even lists. Each word is produced in a phonetically un-reduced form characteristic of clear speech.

All BKB, HINT, and IEEE sentences are semantically well-formed. As a result, each lexical item is partially predictable from the sentence-level context. That is, lexical items within sentences are not independent. A lexical item may be more easily recognized, despite masking or hearing impairment, due to the context of preceding words. Similarly, a listener may be able to correctly guess an earlier missed word after hearing later words in a sentence. Sentence-level context in sentence test materials takes two forms. A word may be predictable in terms of lexical semantics and in terms of grammatical structure. For example, in the BKB sentence “she cut with her knife,” the words “cut” and “knife” are semantically related since a knife is an implement used for cutting. In addition, a listener can use their knowledge of English grammar to predict that the final word of this sentence will be a noun. In contrast, if a listener missed the final word of the BKB sentence “the little baby sleeps,” they could nevertheless predict that the final word is most likely a verb.

In contrast to the sentence-level context found in standard speech intelligibility test materials, there is a complete lack of discourse-level context in these tests. Each test list is a set of non-sequiturs. That is, each sentence is unrelated to, and independent of, all other sentences that precede and follow it. Understanding one sentence does not rely on a listener having understood previous sentences and does not affect understanding of subsequent sentences. This independence is ensured by the lack of repetition of lexical items within lists.

While the description of BKB, HINT, and IEEE sentence materials in this section may appear mundane, the outlined characteristics are informative about the types and sources of information contained within these sentence materials and the listener task that is entailed by these materials.

The least predictable aspect of each speech test sentence is lexical content, which is rarely repeated and is never cued by preceding sentences. The most predictable aspect of speech intelligibility test sentences is prosody. It follows that words carry most of the information available to listeners. This is not surprising given that sentences are scored by number of words correctly repeated in speech intelligibility tests. Words are what differ between speech test sentences and what make a difference in the identity of each sentence. In contrast, prosody does not differ greatly between sentences. For example, Fig. 1 shows intonation contours of 100 sentences from each of the BKB, HINT, and IEEE materials. Within each test, the pattern of intonation is very similar. Nor does prosody make a difference to the identity of sentences. As a result, prosody carries no information in speech intelligibility test sentences.

(Color online) Intonation contours of 100 BKB, HINT, and IEEE sentences.

The information that is available to listeners is therefore purely symbolic in the form of (i) lexical semantics, including parts of speech (e.g., noun, verb, adjective) and dictionary meaning of words; and (ii) grammatical structure, primarily in terms of grammatical roles, such as subject and object and relations between different parts of speech, both of which provide contextual information about lexical items. What is not available to listeners is any information about the cognitive state of the talker, including talker knowledge and shared knowledge. For example, there is no distinction between new and old information, which is typically marked by acoustic emphasis of new information and acoustic de-emphasis of old information, and is also signaled by disfluencies ( Fox Tree, 2001 ). The uniform prosody across sentences leaves little scope for contrastive emphasis, and the lack of discourse context and lexical repetition means that all information is necessarily new information. The uniform intonation pattern across sentences ensures that there is no distinction between statements and questions, and that there is no information about talker emotion.

It is clear that the task of the listener when presented with speech intelligibility test sentences is to understand the words that are spoken. The listener's task does not include any requirement to understand the talker's intentions, state of mind, knowledge, or emotion. This is not surprising since the listener is not engaged in a conversation and frequently is not listening to a live voice, but to a recording. It is also not surprising since we know that speech intelligibility is often measured in units of words correctly repeated. But, it is important to consider whether understanding words—symbolic information—sufficiently captures the task of speech intelligibility faced by listeners outside the laboratory. Can we understand the impact of hearing impairment or the benefits of hearing devices if we only measure word recognition?

This section provides a brief tour of information sources in spontaneous, conversational speech elicited using a referential communication task ( Beechey et al. , 2019 ). This is by no means the only type of conversational speech that has been used to investigate effects of hearing impairment on communication. For example, conversational speech with similar characteristics has been elicited using the Diapix task ( Baker and Hazan, 2011 ). The following discussion of spontaneous speech is not intended to be exhaustive, or to represent all the variety of natural spontaneous speech. See Tucker and Ernestus (2016) and O'Connell and Kowal (2008) for comprehensive reviews. All examples in this section are excerpts from the full conversation transcript (see supplementary material for excerpts). 1

Unlike BKB, HINT, and IEEE sentences, which are almost exclusively statements, conversational speech consists of portions of speech that have many different purposes, such as statements, open-ended questions, closed questions, rhetorical questions, and confirmation requests (see Fig. 2 for examples) among many others.

Talker purpose examples.

Conversation Analysis transcription symbols. Repeated symbols indicate degree.

Whereas the uniform purpose of all sentences in the BKB, HINT, and IEEE materials ensures that the listener need not discriminate talker purpose, this is not the case for listeners when exposed to spontaneous speech. During a conversation, part of a listener's task is to understand what an utterance is for . For example, the talker may wish to tell the listener something that the listener does not know (a statement), or the talker may wish to request that the listener tell them something that the listener knows (a question) ( Brazil, 1995 ). A notable exception to the statement content of speech test materials is the Helen Test ( Ewertsen, 1973 ) which consists of simple questions. Not only does this test consist of a different type of syntactic construction, the test paradigm requires a listener to engage with the purpose of utterances: they are required to answer the questions. A recent variation of this test presents listeners with question–answer pairs and requires them to judge whether the answer is correct ( Best et al. , 2016 ).

Consistent with the many diverse communicative purposes exhibited by utterances in conversational speech, many distinct intonation patterns can be observed. In contrast to the homogeneous intonation contours of BKB, HINT, and IEEE sentences, intonation carries information in conversational speech. Conversational intonation is a difference across sentences that contributes to differences in meaning. For example, Fig. 3 shows a level intonation contour of a statement; a contrastive construction in which the talker is drawing attention to a distinction between two things is characterized by two clear pitch peaks; an open question is characterized by an initial small positive pitch peak on a word such as “which” or “what,” followed by a steady fall in pitch; and a closed question that seeks to elicit a “yes” or “no” answer is characterized by a final rising pattern. The information content of intonation in these examples differs from the intonation contours of the speech test sentences that constitute a lack of difference across sentences and that do not make a difference since essentially, all such sentences are simple statements.

(Color online) Examples of intonation contours in conversational speech.

Prosody of read speech is subject to individual variability and choice on the part of a talker, though the function of intonation is typically different in read and spontaneous speech. For example, Ayers (1994) reported that changes in pitch range were used by talkers as cues to topic structure during reading, but during spontaneous speech intonation also functioned to signal talker corrections and turn-taking; functions that were absent from read speech.

Prosody also plays a central role in conveying information about talker emotion and affect, such as happiness, disappointment, and sarcasm. The information about talker emotion carried by spontaneous speech appears to be far greater than that carried by read speech ( Alghowinem et al. , 2013 ).

For example, in addition to the segmental content of “hee hee,” phonetically [hi hi:] (see Fig. 4 ), rapidly rising pitch, followed by a high pitch plateau, along with vowel lengthening conveys happiness or joy on the part of talker (see Fig. 5 ).

Prosodic marking of emotion and affect on “hee hee,” “oh,” and “damn.” Non-adjacent words are plotted along-side each other for comparison.

In contrast, talker disappointment is conveyed during the production of “oh” (Fig. 6 , line 3) by a low and level intonation contour, along with vowel lengthening (see Fig. 5 ) and sarcasm is conveyed in “damn” (Fig. 6 , line 3) by a fall, rise, fall contour (Fig. 5 ).

Disappointment and sarcasm.

Importantly, emotion is not unambiguously conveyed by lexical semantics in these cases. It is not just the words that are spoken, or the syntactic structure into which words are arranged; it is the way speech is produced–the acoustics–that carries iconic information content. This distinction can be seen if one considers the use of emoji as a primitive means to add affective meaning to written language that is otherwise often ambiguous ( Holtgraves and Robinson, 2020 ; Was and Hamrick, 2021 ).

Conversational speech is characterized by the frequent occurrence of periods of silence, both within talker turns and between talker turns in addition to filled pauses such as “umm” and “err.” A review of the extensive literature on what the authors term pausology is provided by O'Connell and Kowal (1983) . Early studies of pauses recognized uncertainty—specifically the need for a talker to make a choice—as underlying many pauses. Such studies investigated the locations of pauses to determine the size of grammatical units about which talkers must make decisions. For example, Goldman-Eisler (1958) reported that pauses occurred most frequently before the most unpredictable words, and argued that talkers produce pauses in order to make decisions about subsequent lexical items. In contrast, Boomer (1965) reported the most frequent position of pauses as immediately following the first word in a clause and argued that after producing an initial word, a talker is constrained in their choices as to how to complete a clause in a manner consistent with this initial word. In a study of filled pauses, “uh” and “umm,” Fox Tree (2001) considered pauses not only as markers of talker difficulty, but as sources of information that may aid listener comprehension. Fox Tree reported that “uh” signals a short delay and functions to draw the listener's attention to upcoming speech, whereas “umm” signals a longer delay and does not function to draw the listener's attention to immediately upcoming speech. Such disfluencies inform the listener that the talker is engaged in conceptualization that may influence the level of confidence the listener will have in what the talker says.

In Fig. 7 , numbers in parentheses indicate the duration of silent pauses within talker turns (lines 1 and 3) and a silent gap between talker turns. Silence, as it contrasts with periods of speech, conveys information, just as any absence may convey information in contrast to its opposite (see Bateson, 2002 , for discussion). The information conveyed by silence is iconic rather than symbolic. For example, the duration of a period of silence is a direct reflection of the time needed by a talker for cognitive processes, such as deciding how to describe an unfamiliar or complex referent (Fig. 7 , lines 1 and 3) or for a non-speech–related cognitive process, such as visual search (Fig. 7 , line 2).

Pauses and gaps.

Silent pauses within a talker's turn inform the listener that the talker is likely to introduce new or complex information, is experiencing uncertainty, or that they are distracted. Such information may inform the listeners' subsequent interpretation in terms of the reliability or trustworthiness of heard speech.

It is noteworthy that interlocutors do not attempt to avoid such silences during conversation. If such silences were immediately filled for the sake of temporal efficiency, the information that is carried by silence would be eliminated from discourse.

Changes in the temporal characteristics of speech convey information about a talker's cognitive state or process in a similar manner to silent pauses. For example, lengthening of speech sounds that do not constitute phonemic distinctions may reflect an ongoing thought process during speech production. That is, a talker who is thinking while talking may lengthen portions of their speech so that their cognitive process can keep pace with their speech production ( Chafe, 1980 ). It is expected, on this basis, that talkers will produce more slowing or lengthening of speech when speaking spontaneously than when reading; an expectation confirmed by Levin et al. (1982) . This can be seen on lines 3, 6, and 7 of Fig. 8 . As with silent pauses, lengthening of speech sounds carries iconic rather than symbolic information. Lengthening of speech sounds may be more informative to a listener than silence though because it is combined with both lexical semantics and prosody, such as stress placement or pitch changes. Such lengthening may sometimes be primarily useful to the listener rather than the talker. For example, elongation of a syllable may be used to indicate upcoming new information, giving the listener an opportunity to focus attention on immediately following speech ( Arnold et al. , 2003 ; Fox Tree, 2001 ). Lengthening does not appear to be an effort on the part of a talker to hold the floor in Fig. 8 since it co-occurs with silent pauses within turns and silent gaps between turns, which would be expected to be filled if the talker were attempting to maintain their turn.

Speech sound lengthening.

Whereas the primary source of difference between speech intelligibility test sentences is the lexical content of each sentence, words are just one of many sources of information available in conversational speech. Individual lexical items that occur in conversational speech may convey different information at different points in time, depending on discourse context. One way that lexical items are used quite differently in conversational speech in comparison to the BKB, HINT, and IEEE sentences is that they may be frequently repeated, with each repetition conveying somewhat different information.

For example, in Fig. 9 , talker 2 introduces the lexical item “llama” for the first time in the conversation in the form of a statement. Talker 1 then echoes the same word “llama” as a means of confirmation, and potentially to indicate acceptance of the particular term ( Brennan and Clark, 1996 ). In her next turn, talker 1 then repeats “llama” four more times. The repetition of this lexical item does not serve to introduce lexical information in this context, but rather. to demonstrate to the listener that the talker is engaged in an ongoing thought process about this referent.

During spontaneous speech. it is not uncommon for talkers to edit an utterance during production. An example of this process is illustrated on line 2 of Fig. 10 where, after beginning the word “blue,” the talker stops and inserts the adjective “dark” in order to disambiguate their utterance. This editing process illustrates the real-time nature of speech production that Brazil (1995) refers to as now-coding . Utterances are not fully formed before they are produced by talkers, in contrast to the BKB, HINT, and IEEE sentences that are scripted for the talker. Talkers do, however, monitor their own speech in real-time and are aware of the information needed by their interlocutor and the relative information content of their (ongoing) utterance.

Talker edit to include initially omitted adjective “dark.”

In addition to illustrating now-coding of speech, talker edits may also convey information to talkers about the importance of referents. For example, on line 3 of Fig. 10 , talker 1 responds to the edit produced by talker 2 by seeking to confirm the referent. That is, the edit signaled to the listener that the additional information provided by the edit was necessary and therefore should elicit attention.

The examples in this section have sought to demonstrate that the task of speech perception, in the case of spontaneous speech, is to understand the talker , not simply the words . Understanding the talker does involve recovering lexical semantics but this is only a fraction of what is required. A listener must understand the talker's purpose in saying something, such as whether the talker wishes to tell the listener something they know, or request that the listener tell them something they do not know. A listener also needs to incorporate information about the talker, including their current cognitive state, their knowledge, reliability, and emotion. Without gleaning this type of information, the bare words produced by the talker are often insufficient to ensure unambiguous meaning, as anyone who has inserted an emoji at the end of a text message has sensed.

The information content of conversational speech, upon which the task of speech perception can be defined, is based on what Bateson (2002) called “news of difference.” Differences in prosody, timing, and fluency, among other characteristics, carry information. The absence of these acoustic characteristics ensures corresponding absence of information. If categories of information are not present in a signal, they cannot be considered part of the task of understanding that signal. Where categories of information are present in a signal, it would be very surprising if they did not form part of the task of understanding that signal.

The function of perception is to inform an organism about its environment ( Gibson, 1966 ). Auditory perception of non-speech signals is typically thought of in this way. For example, spatial hearing allows an organism to perceive the location of a sound source–an essential skill for survival. Study of the perception of human language has often emphasized the aspect of language that most clearly separates it from non-human animal communication: the use of arbitrary sign systems. But assuming language is made up of only symbolic information, as in typical speech intelligibility test materials, treats the perception of speech as fundamentally different from other auditory signals; if language is purely symbolic, it has no direct relationship to the environment.

Previously, ecological approaches to speech perception that posit a link between signals and events in the environment have been limited to articulatory events in the form of the Motor Theory of speech perception ( Fowler, 1986 ; Rosenblum, 2004 ). However, a more expansive view of ecological perception of speech is that talkers form an important part of a listener's environment. Perceiving the auditory environment, on this view, includes perceiving the identities and states of talkers. Perceiving auditory events in the environment includes the perception of talkers' real-time cognitive processes. Such ecological perception of talker identity and state is often not carried by words, but rather by the iconic information overlaid on words and sentences.

The iconic aspects of spontaneous speech described in this paper are a major source of difference between conversational speech and what is referred to as clear speech in hearing science. Numerous studies have demonstrated a clear speech advantage; speech intelligibility scores are almost always higher when listeners are presented with clear speech signals than with conversational speech (e.g., Hargus Ferguson, 2012 ; Hargus Ferguson and Kewley-Port, 2002 ; Liu and Zeng, 2006 ; Payton et al. , 1994 ). However, the clear speech advantage must be considered in the context of the task implied by a given test paradigm. For example, Miles et al. (2022) reported results from an experiment in which utterances excised from a recording of diadic conversation were presented in a traditional sentence test paradigm and scored in a binomial fashion by words correctly repeated. The results obtained using excised conversational speech were poorer than for BKB sentences for the same listeners. In contrast, where participants encounter conversational speech signals within conversation, their communication is surprisingly robust. For example, Beechey et al. , (2020a , b) reported that participants with hearing impairment were uniformly able to converse successfully in realistic noise over 82 dBA, with and without hearing aid amplification. This level of functioning during conversation should be very surprising considering that many hearing impaired listeners would be expected to perform poorly in such adverse noise conditions when presented with easier clear speech sentences. One explanation for these results is that conversational speech is more difficult only when the listener does not benefit from the additional iconic information contained within it. Where the listener's task is to repeat back words, iconic information is not useful. Further, in a test that is scored in terms of symbolic information, such as words correct, any iconic information gleaned by a listener is not considered in the performance metric. This interpretation is supported by studies of speech comprehension that have compared levels of speech understanding when a participant passively overhears a dialogue or actively takes part in a dialogue. For example, Branigan et al. (2011) found greater comprehension when a participant took part in a conversation than when they merely overheard a conversation.

Further, the clear speech advantage may lead researchers and clinicians to conclude that poorer speech intelligibility with conversational speech means that conversational speech contains acoustic or language elements that, since they appear detrimental, must not be useful to a listener. That is, many of the aspects of conversational speech that distinguish it from clear speech may be viewed as noise. In addition to being potentially detrimental to speech intelligibility, these characteristics may also be interpreted as contributing to variability and lack of control in speech testing. In the dialogue examples presented in Sec. IV , this author has sought to convince the reader that, far from being noise, these aspects of conversational speech convey information. Mistaking information for noise can stem from the theoretical framework through which we interpret observed data ( Hanson, 1958 ). If we define speech as made up of words, then anything other than words is easily seen as just variability. A telling example of how theory and definition can affect interpretation of data in this way is the debate in genetics over “junk” deoxyribose nucleic acid (DNA). If the function of DNA is defined as coding for proteins, a large fraction of human DNA might be seen as non-functional, i.e., simply taking up space. This was the view of junk DNA in which it was argued that large portions of DNA were remnants from evolution. However, to geneticists who considered a definition of DNA that was not limited to coding for proteins, these DNA remnants are seen to serve other functions ( Biémont and Vieira, 2006 ). This change in interpretation was associated with the development of epigenetics and the concept of gene expression. Epigeneticists saw additional dimensions of information and function where others had seen noise.

There are many good reasons why speech test materials have developed to take their current form. Standard speech tests are highly controlled and repeatable experimental instruments that can be administered and scored quickly. However, these virtues should not lead us to reify the task of perceiving and understanding controlled speech test signals as equivalent to the perception or understanding of natural, spontaneous speech. To do so is to conclude that speech intelligibility is what speech tests test. Such reasoning is analogous to treating intelligence quotient (IQ) scores as the same as intelligence. In the absence of a substantive theory of intelligence, psychology has been criticized for effectively defining intelligence as what intelligence tests test ( Boring, 1961 ). This has resulted in the failure to notice that IQ is uncorrelated with complex cognitive reasoning ability required for many real-world tasks ( Ceci and Liker, 1987 ). Importantly, this does not mean that standard speech tests should be abandoned. If it is accepted that the task of perceiving and understanding standard speech intelligibility test signals is a fraction of the task of perceiving and understanding natural, spontaneous speech, this does not imply that speech intelligibility tests lack any predictive or explanatory power. Improved performance measured by a standard speech intelligibility test reflects an improved ability to recognize lexical items but the effect of what is left out is not known. As a consequence, it must be concluded that speech intelligibility test results have at most an ordinal relationship to speech intelligibility performance outside the clinic or laboratory. Rather than analyzing and interpreting speech test results in absolute numerical terms, speech test results should be considered in terms of relative change, and analyzed using appropriate ordinal methods (e.g., Bürkner and Vuorre, 2019 ; Grice et al. , 2015 ). An improved speech test score following a hearing aid adjustment can fairly be considered a valuable improvement. It should be recognized, however, that interpreting a percentage change in speech score as reflecting a change in communication ability of the same magnitude is an over-interpretation.

As the work of researchers investigating perception of iconic information in speech, such as emotion perception (e.g., Chatterjee and Peng, 2008 ) demonstrates, tests of speech perception need not be limited to measuring reception of symbolic information. A possible future direction of research is to pursue the development of a fast, easy to administer and score test that measures individuals' iconic information reception. For example, the same string of words may be interpreted very differently, depending on the talker's prosody. A test might be developed using a forced-choice paradigm that presents listeners with sentences with natural prosody associated with emotion and affect, such as sarcasm, followed by a closed set of meanings from which to choose. Similarly, a test that requires a listener to discriminate a talker's cognitive state, such as fatigue or anxiety, could provide a valuable measure of real-world functioning beyond symbolic language perception. A major challenge associated with developing such tests, however, is the creation of sufficiently controlled stimuli that exemplify genuine, spontaneous instances of talker emotion, affect, or cognitive state.

This paper has sought to explore the question: should we be satisfied with tests of the auditory system that measure the processing of only symbolic information? The answer to this question is no doubt complex and dependent on the particular goals of speech testing and the field within which tests are administered. In linguistic studies where the aims may be to investigate grammatical structures and their mental representations, processing of symbolic information may be the object of study. In perceptual fields, including audiology and hearing science, there is growing interest in increasing the similarities between laboratory conditions and the real-world situations they seek to represent in order to improve the accuracy and generalizability of findings (see Keidser et al. , 2020 , for a recent review). Such efforts have considered how tasks are represented in auditory testing at a relatively high level within the framework of the World Health Organization (WHO) International Classification of Functioning, Disability and Health (ICF) ( World Health Organization, 2001 ). This work considers differences in auditory functioning, such as detection of sound, attentive listening, and taking part in activities related to hearing, including communicating with others. However, lower-level consideration of what is meant by speech intelligibility has received much less attention in the drive towards realism.

Given the widespread acceptance of standard speech intelligibility testing materials and methods, as well as the many real advantages of using such methods, including speed of testing and interpretation, it is hard to imagine that these methods will be abandoned or substantially modified in the near future. In the short term it is a change in interpretation rather than methodology that this paper hopes to inspire. In the longer term, research in hearing science and related fields may benefit from incorporating more research into the low-level details of the task of understanding spontaneous speech and hearing impaired individuals' perception of how talkers speak–not just the words they say.

This work was supported by a Medical Research Foundation Fellowship (MRF-049-0004-F-BEEC-C0899). The author thanks Jorg Buchholz for providing access to the data presented in this paper, and William Whitmer for helpful comments on an earlier draft.

See supplementary material at https://www.scitation.org/doi/suppl/10.1121/10.0013896 for a full transcript of the conversation presented in this paper.

Supplementary Material

Citing articles via.

• Online ISSN 1520-8524
• Print ISSN 0001-4966
• For Researchers
• For Librarians
• For Advertisers
• Our Publishing Partners  
• Physics Today
• Conference Proceedings
• Special Topics

pubs.aip.org

• Privacy Policy
• Terms of Use

Connect with AIP Publishing

This feature is available to subscribers only.

Sign In or Create an Account

Speech Intelligibility Index

Speech intelligibility index (SII) is a measure, between 0 and 1, that represents the intelligibility of speech under a variety of adverse listening conditions, such as noise masking, filtering, and reverberation. Speech cues interrupted by fewer of these conditions will be more available to the listener, and will thus have a higher SII value.

First established by ANSI/ASA S3.5-1997 (R2020) – American National Standard Methods for Calculation of the Speech Intelligibility Index , SII is defined as the “product of band importance function and band audibility function, summed over the total number of frequency bands”. In symbols, this can be understood as:

Where n is the number of SII computational bands, while I i and A i are the values of the band importance function and the band audibility function associated with the frequency band designated by the summation index i . The band referred to is a frequency band, which designates the high and low frequencies in which a sound is emitted. This concept is essential for establishing the basis of speech in its regular conditions.

Workers at Bell Telephone Laboratories in the 1960s conducted the earliest examination of the interaction of different noises influencing speech-recognition performance. Their extensive work led to the definition of the acoustical index, known as the articulation index (AI). The calculation for the AI was first presented in the original version of the S3.5 standard, ANSI/ASA S3.5-1969.

Through the guidance of ANSI/ASA S3.5-1997 (R2020) , SII may be computed through four different methods: critical frequency band, one-third octave frequency band, equally contributing critical band, and octave frequency band. The requirements and formulas needed for carrying out these calculations are addressed in the standard.

The applications of this standard are plentiful, as it is stated in the document that it extends to all listening conditions where the specified input variables exist. This has many different uses, such as in research, e.g. for determining the impact that hearing loss has on the audibility of speech , or for testing PA systems, which can be present in a variety of venues, public or private.

Due to the complexity of the topic, there are three different programs available for calculating speech intelligibility index that make use of the guidelines covered in this standard, However, before using the software, the Acoustical Society of America (ASA) recommends that users familiarize themselves with the standard. The programs can be downloaded from here: Programs for SII

1. Acoustical Society of America (ASA),  ANSI/ASA S3.5-1997 (R2017) – American National Standard Methods for Calculation of the Speech Intelligibility Index (New York: ASA, 2017), 11.

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed .

• Cambridge Dictionary +Plus

speech intelligibility

Meanings of speech and intelligibility.

Your browser doesn't support HTML5 audio

(Definition of speech and intelligibility from the Cambridge English Dictionary © Cambridge University Press)

• Examples of speech intelligibility

Word of the Day

left-handed

(of a person) using their left hand to write and do most things

Robbing, looting, and embezzling: talking about stealing

Learn more with +Plus

• Recent and Recommended {{#preferredDictionaries}} {{name}} {{/preferredDictionaries}}
• Definitions Clear explanations of natural written and spoken English English Learner’s Dictionary Essential British English Essential American English
• Grammar and thesaurus Usage explanations of natural written and spoken English Grammar Thesaurus
• Pronunciation British and American pronunciations with audio English Pronunciation
• English–Chinese (Simplified) Chinese (Simplified)–English
• English–Chinese (Traditional) Chinese (Traditional)–English
• English–Dutch Dutch–English
• English–French French–English
• English–German German–English
• English–Indonesian Indonesian–English
• English–Italian Italian–English
• English–Japanese Japanese–English
• English–Norwegian Norwegian–English
• English–Polish Polish–English
• English–Portuguese Portuguese–English
• English–Spanish Spanish–English
• English–Swedish Swedish–English
• Dictionary +Plus Word Lists

{{message}}

There was a problem sending your report.

• Definition of speech
• Definition of intelligibility
• Other collocations with speech

• Daily Crossword
• Word Puzzle
• Word Finder
• Word of the Day
• Synonym of the Day
• Word of the Year
• Language stories
• All featured
• Gender and sexuality
• All pop culture
• Writing hub
• Grammar essentials
• Commonly confused
• All writing tips
• Pop culture
• Writing tips

Advertisement

intelligibility

[ in-tel-i-j uh - bil -i-tee ]

• the quality or condition of being intelligible ; capability of being understood.
• something intelligible .

Word History and Origins

Origin of intelligibility 1

Related Words

• transparency