N = 348
Note . M age = mean age, SD = standard deviation. The discrepancy in sample sizes between the 1988–1989 and the 2000–2002 assessments is a result of major efforts to re-engage participants from Cohorts 1 and 2 in the early 2000’s, facilitated by the emergence of the internet. Further detail on these efforts is reported in Whitbourne et al. (2009) . Although the age range for Wave 3 includes participants with ages 40–58, only two participants had an age reported outside of the intended 40–49 age range. One participants’ age is reported as 55, another’s is 58. All other remaining participants’ ages fell within the range of 40–48. Sensitivity analyses excluding these two participants are reported in Supplemental Table S8 , and had no substantive differences from the main results.
We checked the intraclass correlation coefficient (ICC) for each outcome within each cohort, at each wave. The ICC was generally very close to zero for these comparisons, suggesting little within-cohort homogeneity on the outcomes of intimacy, generativity, and integrity; thus we combined all four cohorts to increase sample size and power. The total sample size was N = 1224.
The RALS includes planned missingness due to the cohort sequential design, as well as missingness due to attrition. Among the original sample of 348 recruited in 1965–1968, 47% completed the most recent 2012–2014 assessment. Attrition within the other cohorts ranges from 65% to 62%. However, analyses comparing participants who remained in the sample to those who dropped out revealed no significant association between completion status in 2012–2014 and the main predictor of interest, i.e. emerging adulthood identity resolution (r = .02, p > .05), nor for the focal outcomes of intimacy ( r s range from −.04 to .08 across waves, all p s > .05), generativity ( r s range from −.08 to .11, all p s > .05), or integrity ( r s range from −.02 to .04, all p s > .05), consistent with prior attrition analyses of RALS data ( Whitbourne et al., 2009 ) 1 . We use Full Information Maximum Likelihood (FIML) estimation to account for missingness in all longitudinal analyses.
The Inventory of Psychosocial Development (IPD; Constantinople, 1969 ) was originally used to assess the first six psychosocial constructs in Erikson’s model. Starting in 1977, items to assess the remaining two constructs (generativity and integrity) were developed and added ( Whitbourne & Waterman, 1979 ). Each construct is assessed with ten items, including five positive and five negative items, each measured on a 7-point Likert scale. Participants are asked to indicate how characteristic or uncharacteristic each item is of them. Sample items for the subscales included in the present study include: “I know who I am and what I want out of life” (identity), “I have sympathetic concern for others” (intimacy), “I feel productive in my work” (generativity), and “I wouldn’t change my life if I lived it over” (integrity). Scale scores were calculated for each of these constructs by reverse scoring the negatively associated items and adding them to the positively associated items, yielding a sum score for each subscale that reflects a participant’s degree of resolution for each stage of Erikson’s model. Item-level data were available for IPD assessments that occurred from Wave 3 (1988–1989) and beyond, but were unavailable for the two assessments in 1965–1968 and 1976–1977. The full IPD instrument is available in the Supplemental Material . Correlations among the baseline measures of each psychosocial construct are reported in Table 2 . Cronbach’s alpha and test-retest reliability are reported for each subscale where item-level data were available in Supplemental Tables S2 through S5 . Alpha was .64 for identity, and ranged from .66–.76 for intimacy, .36–.61 for generativity, and .69-.78 for integrity. Given the low internal consistency estimates for generativity, we ran sensitivity analyses using a version of this subscale that removed two items that were weakly correlated with the rest of the items (detailed in the Analysis Plan and Results sections). Test-retest reliability for the IPD subscales ranged from r = .24 to r = .81, with stronger correlations among measurements that occurred closer in time. This is consistent with our expectations for data collected at approximately ten-year intervals.
Descriptive Statistics and Correlations among Variables
Variable | 1 | 2 | 3 | 4 | 5 | Mean (SD) | Observed Range |
---|---|---|---|---|---|---|---|
1. Cohort | - | ||||||
2. Sex | −.01 | - | |||||
3. BL Identity | −.04 | .07 | - | 7.30 (6.83) | −24, 30 | ||
4. BL Intimacy | <.01 | .13 | .56 | - | 11.49 (7.50) | −20, 29 | |
5. BL Generativity | −.03 | .15 | .45 | .55 | - | 7.24 (5.66) | −10, 23 |
6. BL Integrity | −.17 | .11 | .56 | .41 | .35 | 3.59 (8.31) | −22, 24 |
BL = Baseline. SD = standard deviation. Correlations among Wave 2–5 psychosocial stage variables are available in supplemental material . Scale scores for Identity, Intimacy, Generativity, and Integrity could theoretically range from −30 to 30.
Participants were recruited for their first assessment during college. Subsequently, alumni (i.e., individuals who attended the University of Rochester, regardless of whether they graduated) were contacted for follow-up assessments using information in the University of Rochester alumni directory. In 2000, fee-based services (Find a Friend; Online Detective) were used and by 2002, and particularly by 2012, Internet searches became available for more thorough identification of past participants. These procedures for identifying and contacting alumni are described in more detail elsewhere (e.g., Sneed et al., 2012 ; Whitbourne et al., 2009 ). Alumni received a letter describing the study, a questionnaire, and a stamped envelope for returning the completed questionnaire. In the most recent assessments (2002, 2012–14), all questionnaires were completed using online survey tools. The present study was declared exempt from review by the Minneapolis VA IRB (VAM-19–00430).
Latent growth curve modeling (LGM) was used to test our main hypotheses. LGM is a longitudinal form of structural equation modeling ( Singer & Willett, 2003 ; Tomarken & Waller, 2005 ), in which parameters such as an intercept (I), linear slope (S), and quadratic slope (Q) are estimated to define an average trajectory of growth over time. Predictors such as emerging adulthood identity resolution can be entered as time-invariant covariates to examine their association with the growth parameters, thus indicating whether such predictors have a significant influence on trajectories of growth for outcomes.
For each of our focal outcomes – intimacy, generativity, and integrity – we developed a separate model to describe an average trajectory of change, and to determine whether trajectories varied by level of emerging adulthood identity resolution. First, we compared several unconditional growth models (i.e., models with no covariates included) to identify the best fitting functional form for each outcome. For each outcome, we considered the following functional forms: intercept-only, linear, quadratic, and basis (i.e., freely estimating the pattern of change; Preacher, 2010 ; Grimm et al., 2011 ) models. For basis models, the loading of Wave 1 scores on the shape factor was set at 0, the loading of the Wave 5 scores on the shape factor was set at 1, and intermediate loadings were allowed to vary freely, allowing estimation of non-linear growth. Thus, the fitted values for outcomes within a basis model are obtained by multiplying the basis shape factor by the basis factor loading for each wave, and adding the model intercept. The model with the best fit, according to visual inspection and fit statistics, was retained for each outcome. Lower values on comparative fit statistics (Akaike Information Criterion and Bayesian Information Criterion), Root Mean Square Error of Approximation (RMSEA) values less than .06, Comparative Fit Index (CFI) values greater than .95, and Standardized Root Mean Residual (SRMR) values less than .08 were taken as indicators of good fit ( Hooper, Coughlan, & Mullen, 2008 ; Hu & Bentler, 1999 ; Schreiber, Nora, Stage, Barlow, & King, 2006 ). After identifying the best fitting unconditional model, we added emerging adulthood identity resolution as a time-invariant covariate to examine its association with the growth parameters. We controlled for sex and cohort in these analyses. Identity resolution was mean-centered to facilitate interpretation of results.
In addition to testing our preregistered hypotheses, we conducted additional exploratory analyses to further probe our findings. We conducted exploratory analyses using mediation models to test whether relationships between identity and the later psychosocial stages were mediated by intervening stages. Specifically, we tested whether the relationship between identity and generativity was mediated by intimacy, and similarly whether the relationship between identity and integrity was mediated by generativity. We also conducted sensitivity analyses fitting a latent growth curve model for generativity using the alterative, shortened version of the generativity scale, and including a covariate reflecting whether participants obtained a graduate degree. Finally, we conducted analyses incorporating age as a covariate. All analyses were performed in R ( R Core Team, 2018 ). The package lavaan was used for latent growth curve modeling ( Rosseel, 2012 ).
Fit statistics for all unconditional models are reported in Table 3 . For intimacy, a linear growth curve model fit best. This model indicated that on average, participants started with a positive level of intimacy at baseline, and increased steadily over time (I=11.62, p <.001; S= 1.30, p <.001; see Figure 2a ). For generativity, a linear model also provided the best fit. This model suggested that participants generally began with a positive level of generativity, with linear increases afterward (I=7.45, p <.001; S= .63, p <.001; see Figure 2c ). For integrity, the basis model provided the best fit. The model produced a negative variance estimate for Wave 5 integrity, so we constrained this variance to zero; this constraint did not substantially affect model fit. The unconditional basis model suggested nonlinear growth in integrity, starting at a relatively low level in Wave 1 and increasing somewhat at Wave 2, declining in Wave 3, and then rising continuously through Wave 4 and Wave 5 (I=3.71, p <.001; S=4.75, p <.001; α 2 =.24, p =.001; α 3 =−.05, p =.65; α 4 =.35, p <.001; see Figure 2e ).
Note. Panels A, C, and E illustrate unconditional model-implied trajectories. Panels B, D, and F illustrate the model-implied trajectories for a hypothetical individual with mean-level identity resolution, as well as an individual one standard deviation above or below the mean.
Fit Indices for Latent Growth Curve Models.
Model | AIC | BIC | Chi-square | RMSEA | CFI | SRMR |
---|---|---|---|---|---|---|
Intimacy | ||||||
Unconditional intercept-only | 16085 | 16121 | χ = 230.09, = 13, <.001 | .12 | .56 | .19 |
Unconditional linear | 15887 | 15938 | χ = 26.44, = 10, =.003 | .04 | .97 | .08 |
Unconditional basis | 15886 | 15952 | χ = 19.22, = 7, =.008 | .04 | .98 | .08 |
Unconditional quadratic | 15878 | 15949 | χ = 9.52, = 6, =.23 | .02 | .99 | .03 |
Conditional linear | 15372 | 15454 | χ = 80.05, = 19, <.001 | .05 | .94 | .07 |
Generativity | ||||||
Unconditional intercept-only | 12734 | 12769 | χ = 95.48, = 13, <.001 | .08 | .74 | .13 |
Unconditional linear | 12661 | 12711 | χ = 16.54, = 10, =.09 | .02 | .98 | .04 |
Unconditional basis | 12656 | 12721 | χ = 5.93, = 7, =.55 | <.001 | 1.00 | .09 |
Unconditional quadratic | 12665 | 12735 | χ = 12.85, = 6, =.05 | .03 | .98 | .04 |
Conditional linear | 12324 | 12406 | χ = 32.60, = 19, =.03 | .02 | .96 | .04 |
Integrity | ||||||
Unconditional intercept-only | 14008 | 14043 | χ = 132.12, = 13, <.001 | .09 | .72 | .16 |
Unconditional linear | 13970 | 14020 | χ = 88.59, = 10, <.001 | .08 | .81 | .11 |
Unconditional basis | 13915 | 13975 | χ = 29.59, = 8, <.001 | .05 | .95 | .08 |
Unconditional quadratic | 13939 | 14004 | χ = 51.35, = 7, <.001 | .08 | .89 | .07 |
Conditional basis | 13469 | 13561 | χ = 108.23, = 17, <.001 | .07 | .78 | .11 |
Note. AIC = Akaike information criterion. BIC = Bayesian information criterion. RMSEA = root mean square error of approximation. CFI = comparative fit index. SRMR = root mean square residual.
Conditional models for each outcome incorporated emerging adulthood identity, cohort, and gender as covariates (see Table 4 ). For intimacy, emerging adulthood identity was significantly associated with intercept and slope (see Table 4 , Figure 2b ). Specifically, each point above the mean level of identity resolution was associated with a .56-point higher level of intimacy at baseline. However, individuals with greater identity resolution in college increased their intimacy more slowly over time than those with lower identity resolution in college. Sex was also associated with baseline intimacy, with women starting out on average 1.61 points higher than men. Cohorts did not differ significantly on baseline intimacy. Neither sex nor cohort was significantly associated with slope.
Latent Growth Curve Models Predicting Psychosocial Outcomes on Emerging Adulthood Identity
Intimacy | Generativity | Integrity | ||||
---|---|---|---|---|---|---|
Coefficient | SE | Coefficient | SE | Coefficient | SE | |
For intercept | ||||||
Intercept | 10.49 | .44 | 7.49 | .53 | - | - |
EA Identity | .56 | .03 | .34 | .02 | - | - |
Cohort | .12 | .15 | −.24 | .17 | - | - |
Female | 1.61 | .35 | 1.25 | .32 | - | - |
For linear slope | ||||||
Intercept | 1.06 | .25 | .28 | .25 | - | - |
EA Identity | −.10 | .02 | −.03 | .01 | - | - |
Cohort | .15 | .13 | .17 | .12 | - | - |
Female | .13 | .19 | .08 | .18 | - | - |
For basis intercept | ||||||
Intercept | - | - | - | - | 4.54 | .92 |
EA Identity | - | - | - | - | .62 | .05 |
Cohort | - | - | - | - | −.60 | .28 |
Female | - | - | - | - | 1.23 | .45 |
For basis shape factor | ||||||
Intercept | - | - | - | - | 5.30 | 2.38 |
EA Identity | - | - | - | - | −.38 | .12 |
Cohort | - | - | - | - | −1.15 | .64 |
Female | - | - | - | - | −.44 | 1.13 |
Basis factor loadings | ||||||
Wave 2 | - | - | - | - | .49 | .08 |
Wave 3 | - | - | - | - | .37 | .24 |
Wave 4 | - | - | - | - | .57 | .12 |
Intercept | 15.89 | 2.51 | 12.23 | 2.23 | - | - |
Slope | 1.18 | .50 | 1.74 | .53 | - | - |
Basis intercept | - | - | - | - | 22.29 | 13.63 |
Basis slope | - | - | - | - | 62.01 | 53.87 |
Note. All coefficients are unstandardized.
EA = Emerging adulthood. Fixed effects represent the average trajectory across all participants, and random effects represent the variance of individual participants’ trajectories around the average trajectory.
For generativity, identity resolution was also associated with intercept and slope, in a similar pattern (see Table 4 , Figure 2d ). Each additional point on the identity resolution scale was associated with .34 additional points on the generativity scale at baseline. Higher identity resolution was also associated with slower growth in generativity over time. In addition, women had a higher initial level of generativity than men by 1.25 points. Cohort was not significantly associated with baseline generativity, and neither sex nor cohort predicted different rates of change in generativity. 2
Emerging adulthood identity resolution was also associated with both the initial level and pattern of growth for integrity (see Table 4 , Figure 2f ). Each point above the average for identity resolution was associated with a .62-point higher level of integrity at baseline. Higher identity resolution in college was also associated with a slower rate of growth from one wave to the next. Sex and cohort were also related to baseline levels of integrity. Women scored higher than men by 1.23 points on average. Later cohorts scored slightly lower than early cohorts, with each cohort starting .60 points below the previous cohort. Sex and cohort were not significantly associated with the growth parameters for integrity.
We conducted sensitivity analyses controlling for attainment of a graduate degree, as well as age. These analyses did not substantively change the results (see Supplemental Tables S7 - S10 ).
We used path analysis models to test whether the relationships between T1 identity and T3 generativity, as well as between T1 identity and T3 integrity, were mediated by the intervening psychosocial stages. Results of these models are reported in Figure 3 . The indirect path from identity to generativity via intimacy was significant ( b = .07, p <.001), though smaller than the direct path from identity to generativity ( b = .18, p <.001). A similar pattern was found for the indirect path from identity to integrity via generativity ( b = .13, p <.001), relative to the direct path from identity to integrity ( b = .23, p <.001). These findings suggest that part, but not all, of the relationship between identity and the later psychosocial stages can be explained by the intervening stages.
Note . Unstandardized coefficients are reported. *** p <.001, ** p <.01, * p <.05
Erikson’s classic psychosocial stage model suggests that identity development in adolescence and emerging adulthood shapes an individual’s progression through later developmental tasks across the lifespan ( Erikson, 1950 ; 1968 ; Vaillant & Milofsky, 1980 ). The purpose of the present study was to test whether identity resolution in emerging adulthood predicted subsequent development of intimacy, generativity, and integrity across early, middle, and later adulthood. We estimated developmental trajectories for each of these three psychosocial constructs starting in the early twenties and continuing through the sixties. Although emerging adulthood identity resolution did predict higher levels of intimacy, generativity, and integrity, as well as significant differences in growth rates for each of these three constructs, the specific trajectories we found were quite different from what we predicted. We expected that individuals with high identity resolution in college would go on to experience substantial growth in intimacy, generativity, and integrity, while their peers with lower identity resolution would experience little growth in these areas. Instead, our findings suggest that strong identity resolution in emerging adulthood predicts consistently high levels of intimacy, generativity, and integrity across adulthood. In contrast, those with low identity resolution in emerging adulthood go on to experience faster rates of growth across their thirties, forties, and fifties, so that they start to “catch up” with their peers later in life, especially in the domains of intimacy and integrity 3 . Only data with multiple assessments across the adult years could reveal these developmental patterns - a notable strength of the RALS dataset.
Identity formation is often viewed as a key developmental task for adolescence and emerging adulthood, and a prerequisite for healthy adult functioning ( Arnett, 2000 ; Erikson, 1968 ; 1950 ; Marcia, 1966 ; Meeus, 2011 ). Individuals who have trouble forming a coherent, stable sense of self at these early times of life are expected to experience poor outcomes later in adulthood as a result. However, we found that having relatively low identity resolution in college did not totally preclude growth through subsequent psychosocial stages. These findings provide some reassurance that individuals who do not manage to form a coherent identity “on time” in emerging adulthood are not destined to fail at the key developmental tasks later across the lifespan – they may just take longer to arrive there. Indeed, some research on adults in their mid-twenties suggests that those who remain in identity diffusion (the least mature, least resolved identity status) can nonetheless progress forward in their identity development later on ( Carlsson, Wangqvist, & Frisen, 2015 ). Prior research with the RALS has also suggested that life experiences such as entering a committed relationship or becoming a parent can contribute to “catching up” on psychosocial growth for individuals who exhibited lower levels of intimacy and generativity early in adulthood ( Whitbourne et al., 2009 ). Course-corrections ( Stewart & Vandewater, 1999 ), or later-life choices made after revisiting one’s earlier-life regrets, may be another mechanism for gaining ground in these psychosocial domains through midlife.
Perhaps what was most surprising, from a developmental perspective, was that individuals with high identity resolution in college also tended to score highly on intimacy, generativity, and integrity in their youth, and maintain those levels over time. Eriksonian theory posits that healthy psychosocial development is dynamic, with each psychosocial construct becoming especially salient at a different point in the adult lifespan ( Erikson, 1950 ; 1968 ; see also Zucker, Ostrove, & Stewart, 2002 ; McAdams & de St. Aubin, 1992 ). However, our findings suggest that individuals who had already attained a high level of identity resolution in their early twenties were also particularly mature on all of these dimensions in a stable, trait-like way. Though it is assumed that identity resolution serves as preparation for future growth in these areas, our findings suggest that instead, people who have highly mature identities in emerging adulthood may already have a strong sense of intimacy, generativity, and integrity, which remains strong across their lifespan.
In contrast to the conventional, strictly age-graded interpretation of Erikson’s psychosocial stage model (e.g., Vaillant & Milofsky, 1980 ), more recent interpretations have emphasized Erikson’s assertion that each psychosocial construct is present across the lifespan ( Whitbourne et al., 2009 ; Pratt, Lawford, Matsuba, & Villar, 2020 ). In support of this “matrix” interpretation, previous research with the RALS ( Whitbourne et al., 2009 ), as well as other longitudinal datasets ( Einolf, 2014 ) has demonstrated a surprising degree of stability in some psychosocial constructs over time, a finding reinforced by the present analyses. Our results also question the sequential, stage-based interpretation by demonstrating that the psychosocial constructs that are typically associated with adulthood can nonetheless be quite strong in the early twenties (see also, e.g., Pratt & Lawford, 2014 ). Our mediation models suggested that the effects of identity on the later psychosocial stages (e.g., generativity) may be partially explained by the intervening stages (e.g., intimacy). However, identity remained significantly associated with late psychosocial stages after accounting for the indirect paths through the middle stages, further calling into question a strictly sequential interpretation of Erikson’s model.
The nonlinear growth pattern we found for integrity is intriguing, as it contradicts theoretically derived expectations for integrity across the lifespan. However, the rise in integrity around age 30 and subsequent dip in the 40’s may reflect changing attitudes across these decades of life. In particular, the integrity subscale includes items such as “I have reached my goals,” “I am proud of what I’ve done,” “I take responsibility for my actions,” and the reverse-coded item, “I am afraid of getting old.” The thirties have been a relatively neglected time of the lifespan, but this time of life, recently coined “established adulthood,” (Mehta et al., 2020), is a time of increased stability and success, as adults settle into career and family roles, and attain a degree of financial stability. This may be a time when adults feel a greater sense of accomplishment and pride compared to their emerging adulthood years. At the same time, the emerging health concerns and awareness of aging that characterize midlife (Lachman et al., 2015) have not yet set in for most established adults. Thus, it is conceivable that integrity, as measured by the IPD, does indeed increase slightly in the 30’s, with a slight decrease afterward. It is important to note that the magnitude of these changes we detected was small, relative to the larger increase in integrity across the 50’s and 60’s. Nonetheless, they point to possible salience of integrity at earlier stages of life than theoretically expected. Further research examining psychosocial development in established adulthood would help elucidate these effects.
Prior research has demonstrated differences by gender and birth cohort in Eriksonian psychosocial development (e.g., Fadjukoff et al., 2010 ; Whitbourne et al., 2009 ; Vandewater & Stewart, 2006 ), and the present study sheds additional light on how these demographic factors may influence psychosocial growth over time. First, we found that women tended to score slightly higher than men on intimacy, generativity, and integrity in emerging adulthood, and there were no gender differences in slope, suggesting that gender differences persisted over time. These differences may reflect underlying associations between gender and relevant personality traits. For example, meta-analytic evidence suggests that women tend to score higher on warmth and nurturance, two traits that are closely related to intimacy and generativity ( Costa, Terracciano, & McCrae, 2001 ). Birth cohort was unrelated to all outcomes except that later cohorts tended to have lower levels of ego integrity in emerging adulthood, consistent with past RALS research ( Whitbourne et al., 2009 ).
An important direction for future research is understanding how some individuals attained a high level of intimacy, generativity, and integrity earlier in life than expected. Given our findings that high levels of intimacy, generativity, and integrity appear to persist beyond emerging adulthood, promoting the acquisition of these psychosocial attributes in emerging adulthood may have a lasting positive effect across the lifespan.
The present findings point to identity development as one potential mechanism contributing to psychosocial maturity in emerging adulthood. Engaging in identity work may involve trying out different educational and career possibilities, developing friendships and romantic relationships, experimenting with different ideological, religious, and political views, and making lasting commitments in these areas ( Marcia, 1966 ). These same activities could also lead youth to grapple with the same issues that are central to the developmental tasks of intimacy, generativity, and integrity, including forming close bonds with significant others, promoting the well-being and growth of younger people, and contemplating one’s lasting impact on the world. For example, volunteering in one’s neighborhood and community may contribute to emerging adults’ identity development ( Pancer, Pratt, Hunsberger, & Alisat, 2007 ), and also to the development of generative concern ( Soucie, Jia, Zhu, & Pratt, 2018 ). Thus, the activities involved in identity development may also promote the development of intimacy, generativity, and integrity all during emerging adulthood.
Another possibility is that these psychosocial constructs may all reflect underlying dispositional traits, or attributes of personality that are relatively stable across the lifespan. For example, multiple studies have found correlations between generativity and the Big Five personality traits of extraversion, openness to experience, and emotional stability (the opposite of neuroticism) among middle-aged adults ( Bradley & Marcia, 1998 ; Cox, Wilt, Olson, & McAdams, 2010 ; De St. Aubin & McAdams, 1995 ; Peterson & Duncan, 2007 ; Van Hiel, Mervielde, & de Fruyt, 2006 ). The pattern we observed, where some participants were persistently high on all of the psychosocial constructs we examined, may simply reflect especially adaptive personality trait profiles among those individuals. Furthermore, there is evidence supporting the existence of a general factor of psychosocial development ( Dunkel et al., 2012 ; Dunkel & Harbke, 2017 ), a latent factor capturing the shared variance among all eight psychosocial constructs. This general psychosocial factor may reflect an overall ability to successfully navigate psychosocial challenges, and could drive simultaneous high scores across multiple psychosocial dimensions.
The present study revealed some limitations of the IPD, the questionnaire measure used to assess Eriksonian constructs in the RALS. Notably, internal consistency for the generativity subscale was quite low for some assessments. Our sensitivity analysis, using a modified version of the generativity subscale with problematic items removed, somewhat mitigates these concerns, as the findings were largely similar to the main analysis. However, future research replicating the present findings using alternative measures of the Eriksonian psychosocial constructs (e.g., the Eriksonian Psychosocial Stage Inventory; Rosenthal, Gurney, & Moore, 1981 ; the Loyola Generativity Scale; McAdams & de St. Aubin, 1992 ) are warranted. Though replicating the full forty-year longitudinal models may not be possible, replication of the cross-sectional effects and shorter-scale longitudinal studies may bolster support for the connections between identity, intimacy, generativity, and integrity.
An important limitation of the present sample is its homogeneity in terms of education, race, and social class. The RALS participants were recruited from among students at a private university, and their scores on the IPD indicate that they are in general relatively well-adjusted on most of Erikson’s psychosocial dimensions. Investigating similar questions among individuals who have especially low levels of identity resolution may reveal whether developmental trajectories unfold differently for individuals with substantially lower identity resolution in emerging adulthood. Thus, extending this work to clinical samples (e.g., individuals with borderline personality disorder and other mental health concerns characterized by low identity resolution; Wilkinson-Ryan & Westen, 2000 ) is an important direction for future study. As much of the foundational work on lifespan psychosocial development has relied on relatively highly educated samples (e.g., the Mills College study, Helson, 1967 ; the Radcliffe College study, Stewart, 1978 ), investigation of more diverse and representative samples is also needed. Evidence suggests, for example, that emerging adults who do not attend college tend to become parents earlier than college students, and have substantially different patterns of employment and financial dependence across their early twenties – all factors that may affect opportunities for identity development at this time of life ( Mitchell & Syed, 2015 ). In general, research on ego identity development among youth from racial/ethnic minority backgrounds and youth who do not attend college has been sparse, and the need for more work with diverse groups is a longstanding gap within the ego identity development literature ( Syed & Mitchell, 2013 ). Nonetheless, the RALS’ strengths – the inclusion of men and women from multiple birth cohorts, and five waves of assessments spanning nearly fifty years – make it a unique and valuable source of information on developmental sequencing of Eriksonian psychosocial stages.
The consequences of adolescent and emerging adulthood identity formation processes for later-life outcomes have been regularly asserted, but insufficiently tested ( Erikson, 1950 ; 1968 ; Vaillant & Milofsky, 1980 ; McLean & Syed, 2015 ). The present study investigated the association between emerging adulthood identity resolution and subsequent development through the Eriksonian psychosocial stages associated with adulthood, namely intimacy, generativity, and integrity. We found that higher identity resolution was associated with persistently high levels on each of these psychosocial constructs, whereas lower identity resolution predicted lower initial levels and gradual increases over time. These trajectories appeared to nearly converge by the time participants were in their sixties, suggesting that one’s emerging adulthood identity has less importance over time, and that individuals who struggled more with identity formation in emerging adulthood are able to make up for it later in life. Our findings support the growing body of literature questioning the strictly sequential, age-graded interpretation of Erikson’s psychosocial stage model (e.g., Whitbourne et al., 2009 ; Pratt et al., 2020 ). Nonetheless, it appears that successfully resolving the identity-related challenges of emerging adulthood may have a lasting, positive impact for intimacy, generativity, and integrity across the lifespan.
Supplemental material, acknowledgments.
Project information including preregistration is available at https://osf.io/38sz2/?view_only=57b2e0d1e629483d964af493ab80f6ed . This material is based on work supported by the Office of Academic Affiliations, Department of Veterans Affairs, through the Health Services Research and Development Advanced Fellowship, TPH 67-000, PI: Diana Burgess. The present analyses have not been previously published or presented. Prior publications based on the Rochester Adult Longitudinal Study dataset are included in the literature review of the present study.
1 The complete correlation matrix for attrition at each wave is available in Supplemental Table S1 .
2 Due to the low internal consistency of the generativity subscale, we fit an additional model for generativity, after removing two weakly correlated generativity items from the subscale. The results of this model are reported in Supplemental Table S6 . The results were substantively very similar to those found with the full generativity subscale.
3 The current findings are distinct from those of Sneed et al. (2012) in demonstrating the association between emerging adulthood levels of identity resolution and long-term trajectories of intimacy; further explanation can be found in the Supplemental Materials .
Lauren L. Mitchell, Center for Care Delivery & Outcomes Research, Minneapolis VA Healthcare System.
Jennifer Lodi-Smith, Department of Psychology, Canisius College.
Erica N. Baranski, Department of Psychology, University of Houston.
Susan Krauss Whitbourne, Department of Psychological and Brain Sciences, University of Massachusetts.
IMAGES
VIDEO
COMMENTS
Many theories of motivation recognize a person's identity adds value to the prediction of behavior; albeit indirectly, through constructs such as social norms, self-efficacy, perceived behavioral control, and values. Recently, behavioral research has more directly assessed the influence of identity on behavior; however, it is unclear whether ...
From Erikson's psychosocial theory to identity process-oriented models. Erikson's (Citation 1950, Citation 1968) psychosocial theory has inspired research on personal identity up to the present time (Schachter & Galliher, Citation 2018).A fundamental tenet of Erikson's view is that identity formation is a core task that individuals strive to address throughout their entire life span ...
In summary, in research using functional neuroimaging, explicit self-evaluation as well as more indirect forms of social self-evaluation implicated in relational identity robustly engage the vmPFC and the rostral/perigenual ACC (as part of a broader network including the VS, TPJ, and dmPFC) in children and adolescents, often more so than in adults.
From Erikson's early writings, several broad approaches to identity theory and research have emerged, laying differential emphasis on the psychosocial, phenomenological, and the contextual nature of identity. This article has reviewed some of the writings and research that have sprung from the identity status model of James Marcia (1966, 1980 ...
Over the past two decades, identity has emerged as a concept framing studies of early career researcher experience. Yet, identity is an amorphous concept, understood and used in a range of ways. This systematic review aimed to unpack the underpinnings of the notion of researcher identity. The final sample consisted of 38 empirical articles published in peer-reviewed journals in the last 20 ...
There is an emergent identity work perspective that draws on multiple intertwined streams of established identities theorizing and identities-related research. This perspective is characterized loosely by five broad sets of assumptions: (i) selves are reflexive and identities actively worked on, both in soliloquy and social interaction; (ii) identities are multiple, fluid and rarely fully ...
In the current review, we discuss progress in the field of identity research between 2010 and 2020. One of the key developmental tasks in adolescence and young adulthood is to develop a coherent sense of self and identity (Erikson, 1968).Personal identity refers to one's sense of the person one genuinely is, including a subjective feeling of self‐sameness and continuity over contexts and time.
This introductory summary provides an overview of the content of the special issue entitled "Identity Development Process and Content: Toward an Integrated and Contextualized Science of Identity." The 16 theoretical and empirical articles that comprise this special issue were selected to highlight innovative methodologies, theoretical integration, and multicultural perspectives on the ...
In recent years the interest in qualitative identity research has been growing, as can be seen in the programs of the International Society for Research on Identity conferences in 2017 and 2019 (ISRI, Citation 2020), and in the increasing number of qualitative studies on identity (among many others Carlsson et al., Citation 2015; Kunnen & Bosma ...
Yet, identity is an amorphous concept, understood and used in a range of ways. This systematic review aimed to unpack the underpinnings of the notion of researcher identity. The final sample ...
Abstract. This article introduces the reader to the concepts of identity, positionality and reflexivity and outlines their relevance to research paramedics. We outline how a researcher's identity and positionality can influence all aspects of research, including the research question, study design, data collection and data analysis.
In supraliminal or "conscious" priming, a participant is exposed to priming stimuli (here, the moral identity characteristics) as part of a conscious task (Bargh and Chartrand 2000). 5 Following prior research on moral identity activation (Aquino et al. 2007), the task was for subjects to write a story about themselves using either the ...
National identity on the other hand is a person's identity or sense of belonging to one or more states or to one or more nations. Thus, Kehily (2009) argues that while identity plays a huge role ...
Identity has been defined as "unitary" or. "multiple", "real" or "constructed", "stable" or "fluid", "personal" or "social", and in many other ways that often seem ...
Identity cues appear ubiquitously alongside content in social media today. Some also suggest universal identification, with names and other cues, as a useful deterrent to harmful behaviours online.
ABSTRACT Thepaperundertakesanordinarylanguageanalysisofthecurrentmeaningsof\iden-tity,"acomplicatedandunclearconceptthatnonethelessplaysacentralroleinongoing
Among the social determinants that affect the health and well-being of young people throughout the world, gender is a pivotal influence, with both subtle and overt, immediate as well as longer term influences on adolescent development, resources and opportunities, and ultimately, adolescent and adult health. Most societies are profoundly gendered; these gender roles and expectations affect ...
Your researcher identity is formed by combining all of your research outputs (e.g., articles, books, datasets, conference proceedings, blog posts, etc) and its impact that may be aligned with or act as an author profile. As a researcher, it is critical to have control in managing your researcher identity. Your researcher identifier can:
(Mis)diagnosing DID. The combination of insufficient training in recognising trauma-related dissociation, limited exposure to accurate scientific information about DID, symptom similarities with other disorders (such as schizophrenia, bipolar disorder and borderline personality disorder) and the aetiology debate has led to a reluctance to consider a diagnosis of DID, leading to under- and ...
The present research advances the current understanding by incorporating viewers' need for self-identity into their response model. Six studies, including an analysis of field data (14,725 Instagram photos by a top travel influencer) and five controlled experiments, find that the presence (vs. absence) of another human in the photo of an ...
The way in which intersectionality theory was taken up in vision impairment research is described using three overarching themes including: (1) as a lens for the interpretation of findings; (2) as a general conceptual framework for the article; and (3) as a tool for data analysis.
Autism can be considered both a personal and social identity. Identifying the factors contributing to positive Autistic identity development is crucial given the potential implications for mental health and wellbeing. In this systematic review, we aimed to synthesize quantitative literature on Autistic identity to identify the (individual and environmental) factors associated with Autistic ...
Identity: An International Journal of Theory and Research, under the sponsorship of the International Society for Research on Identity (ISRI), provides an interdisciplinary and international publication outlet for conceptual, empirical, and methodological developments and emerging trends in the field of identity research.. The Journal brings together leading research in identity undertaken by ...
Travers has received funding from the Social Sciences and Humanities Research Council of Canada for research related to transgender youth and from Sport Canada to fund research related to gender ...
A version of this article appears in print on , Section A, Page 11 of the New York edition with the headline: The Lesser-Known Side of Harris's Identity. Order Reprints | Today's Paper | Subscribe
A 20-page article may perfectly fit a researcher's needs. Sustaining that argument for 200 pages might be unnecessary -- or impossible. The quality of a research article and the legitimacy of its findings are verified by other scholars, prior to publication, through a rigorous evaluation method called peer-review. This seal of approval by other ...
PARIS (AP) — Exactly one week after thousands of athletes sailed through the opening ceremony at the Paris Olympics, a gold medal was finally won Friday by a member of the neutral collective of athletes.. Ivan Litvinovich was barred from boarding a boat on the River Seine — just like all 32 athletes from Russia and Belarus whose Olympic team identity has been made almost invisible because ...
Typically, adolescence and emerging adulthood are seen as an especially crucial time for identity development (Erikson, 1968; Arnett, 2015; Schwartz et al., 2013; Meeus, 2011; McLean & Syed, 2015).This is a time of life when youth have relatively ample opportunities and motivation to explore what kind of person they want to be, and what direction they would like their life to take (Arnett, 2000).
Vice President Kamala Harris condemned former President Donald J. Trump's remarks. But she also made it clear she would not engage in a debate with a white man critiquing her Black identity.
I hope my research project provides a great learning experience for my graduate school ambitions. I have enjoyed the research project and am looking at completing more research in graduate school and potentially pursue a research-related career. ... Brian Pennington offers insights on multi-religious identity for Religious News Service article.