Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jan 24.
Published in final edited form as: J Gerontol B Psychol Sci Soc Sci. 2026 Jan 12;81(2):gbaf239. doi: 10.1093/geronb/gbaf239

The Advantages of Aging: Greater Stability of Self-Representation in Older Adults is Linked to Higher Well-Being

Sarah Hennessy 1,*, Jessica R Andrews-Hanna 1,2, Matthew D Grilli 1,2
PMCID: PMC12829465  NIHMSID: NIHMS2133379  PMID: 41325187

Abstract

Objectives:

Maintaining a coherent, stable sense of self is critical to well-being, particularly in older adulthood. Yet, little research has examined the objective stability of self-representation across the lifespan. In this study, we investigated how self-representation stability differs between younger and older adults, is supported by autobiographical memory in the laboratory and daily life, and predicts well-being.

Methods:

This observational study included younger (n = 51, aged 18-35) and older (n = 126, aged 60-90) adults. Outcome variables included personality, well-being, in-lab autobiographical memory, and naturalistic autobiographical thought. Using robust quantile regression, analyses examined age-group differences and interactions across self-representation stability, well-being, and autobiographical memory.

Results:

We observed that self-representation stability: 1) was higher in older adults, 2) predicted well-being across participants, 3) was not significantly related to in-lab autobiographical memory, and 4) was higher in younger adults who had more autobiographical thoughts in daily life.

Discussion:

These findings suggest a link between self-representation stability and well-being, with age-related differences in its cognitive mechanisms. These findings further underscore an important positive of aging and offer insight into the connection between autobiographical processes and self-representation stability.

Keywords: autobiographical memory, self, self-concept, ecological momentary assessment

One’s “sense of self”, their representation of who they are, is composed of traits (e.g., “kind”), roles (e.g., “scientist”), and behaviors (e.g., “hard-working”) (Markus & Wurf, 1987). Self-representations help guide goals and interpret experiences and memories (Conway, 2005), and their stability enables a consistent sense of self across situations and time (Baumeister, 1998). Stable self-representations may help explain older adults’ well-documented advantages in well-being (Burns, 2020; Carstensen & Mikels, 2005). However, prior research has not yet answered fundamental questions on this topic: Do younger and older adults differ in their acute self-representation stability? Does well-being vary as a function of self-representation stability? In the present study, we sought to answer these questions and further explore the long-theorized (Conway, 2005; Rathbone et al., 2012a; Sow et al., 2023) role of autobiographical memory in the presence of a stable self-representation, using both traditional in-laboratory and novel, naturalistic ecological momentary assessment methods.

Stability of self-representations across adulthood

Stable self-representations support health and well-being, with individuals perceiving a continuous and clear sense of self, reporting greater well-being (Diehl & Hay, 2011; Ritchie et al., 2011; Sedikides et al., 2016) and less psychopathology or loneliness (Butzer & Kuiper, 2006; Richman et al., 2016). This is particularly relevant in older age amid social, physical, and cognitive changes. Theoretical frameworks suggest that older adults are particularly adept at maintaining stable self-representations (Brandtstädter & Greve, 1994; McAdams, 2013). Erikson (1963) framed adolescence as a time of identity formation, and older adulthood as focused on coherence, placing self-stability as central to identity development. This may be supported by crystallized self-knowledge and shifts in memory (Grilli & Sheldon, 2022) and narrative (Liao & Bluck, 2023; Lind et al., 2021; McAdams, 2013) processes.

Yet, insufficient evidence evaluating age-related differences in self-representation stability exists. While previous studies describe lifetime personality development (Bleidorn et al., 2022), daily trait-related behavioral fluctuations (Fleeson & Gallagher, 2009), and perceived sense of self-stability (i.e., self-concept stability, self-continuity) (Diehl & Hay, 2011; Fuentes & Desrocher, 2012; Lampraki et al., 2023), none have examined objective stability through acute self-evaluation consistency. Measuring whether individuals answer trait-related questions consistently across short periods instead offers an implicit metric fundamentally differing from subjective self-reports of perception and behavior. Older adults may perceive greater self-continuity (e.g., “I am the same me as I was last week”), but this may not reflect consistency in actual self-evaluations (e.g., “I am kind”). Subjective self-continuity and objective self-representation stability are conceptually related but differ in their measurement approaches. Self-continuity is a phenomenological measure capturing an individual’s reported sense of self-consistency over time (Sedikides et al., 2023). Conversely, objective self-representation stability is an empirical measure of the consistency in individuals' self-evaluative responses, independent of participants' conscious appraisals of their own temporal stability. Despite their theoretical grounding, objective measures of self-representation stability remain underexplored.

Cognitive mechanisms of stable self-representations

Additionally, limited work investigates the cognitive mechanisms supporting objective self-representation stability across the lifespan. The self is deeply intertwined with autobiographical memory (Conway, 2005; Rathbone et al., 2012b; Sow et al., 2023), relying on selective encoding and retrieval to build and maintain coherence (Conway, 2005; Fivush & Grysman, 2023).

While both episodic (Conway, 2005; Cuervo-Lombard et al., 2021; Duval et al., 2012) and personal semantic memory (Grilli & Verfaellie, 2015; Klein & Loftus, 1993) may be important for self-representation stability, their relative contributions may shift across the lifespan. Younger adults, with newer and more malleable self-concepts (Chessell et al., 2014; Waterman, 1982), may rely more on episodic memory for self-judgments, recalling trait-relevant experiences (“yesterday, I was hard-working”) to update the self-concept.

In contrast, older adults may rely more on personal semantic memory for self-concept stability, as they have well-established self-knowledge (“I am hardworking”), (Chessell et al., 2014) consolidated from episodic memories (Conway, 2005; Prebble et al., 2013). When making trait judgments about themselves (Klein & Loftus, 1993; Robinson & Clore, 2002), generating self-descriptive statements (Grilli & Verfaellie, 2015), or describing identity representations (Grilli, 2017), individuals rely primarily on personal semantic memory rather than episodic recollections. Crystallized self-knowledge may support self-stability by serving as a heuristic for consistent self-appraisals and as a scaffold that, especially for long-standing self-knowledge, organizes new and past experiences into a coherent sense of self. This shift may coincide with age-related changes in autobiographical memory, with older adults recalling fewer episodic and more semantic details (Simpson et al., 2023), accompanied by differences in neural activation (Addis et al., 2011; Devitt et al., 2023). However, further research is needed to understand the cognitive mechanisms of self-representation stability.

Measurement challenges in autobiographical memory research related to the self

Our understanding of cognitive mechanisms of self-representation stability may be limited by challenges with how researchers traditionally assess autobiographical memory. Lab measures (e.g., Levine et al., 2002) assess retrieval capacity, requiring specific event recall rich in episodic detail. However, they may miss how people naturally engage with memories in daily life. Ecological momentary assessment (EMA) may better capture autobiographical thought tendencies, including episodic memories (Andrews-Hanna & Grilli, 2021), in daily life, where self-knowledge is more likely to be abstracted and updated. Thus, daily episodic memory engagement may support self-representation stability, particularly during young adulthood, but traditional lab measures may overlook this.

Current Study

The goal of this preregistered study was to understand how acute, implicit self-representation stability relates to well-being and autobiographical memory in younger and older adults. We operationalized self-representation stability as acute consistency in self-evaluated personality, expecting individuals with stable self-concepts to provide similar trait ratings today as their own responses two weeks prior.

First, we aimed to clarify age-related differences in self-representation stability, hypothesizing that self-representation is more stable in older adults, reflecting an adaptive developmental shift. Secondly, we evaluated potential benefits to well-being associated with self-representation stability. We hypothesized that, across the lifespan, individuals with higher self-concept stability would report greater subjective well-being, particularly in indices related to the self (i.e., self-acceptance, purpose in life, autonomy). Lastly, we investigated whether the cognitive mechanism of self-representation stability differs across age groups. We hypothesized that self-representation stability is supported by 1) semantic self-knowledge retrieval capacity, particularly in older adults, and 2) frequent returns to autobiographical episodes in daily life (and secondarily, frequent general autobiographical thoughts), particularly for younger adults. We used both in-lab and EMA-based assessments of autobiographical memory.

Methods

The Institutional Review Board of the University of Arizona approved this research (reference no.: 2001274054). We obtained written informed consent from each participant before engaging in the research procedures. We preregistered the methods, hypotheses, and analyses for this study. We note that other studies from our research group have examined the sample assessed in this study, while none have addressed the research questions posed in the present study. The details of these studies are provided in Supplementary Material.

Participants

This study included 51 younger (aged 19-34) and 126 older adults (aged 60-85). We pooled these participants from two independent samples who completed similar procedures (Sample 1 = 55; Sample 2 = 122). Table 1 provides detailed demographic information for the pooled sample. For both samples, we recruited participants from Tucson, Arizona, via participant databases, flyers, advertisements, social media, and email lists, and exclusion criteria were as follows: not being proficient in English, history of psychiatric or neurological disorder, resting outside of specified age ranges, and performing below normal cutoffs on neuropsychological tests and cognitive screening measures. Our neuropsychological and cognitive screening measures include measures of memory (California Verbal Learning Test, Rey Complex Figure Test, language (Boston Naming Task), animal fluency task), attention/executive function (trail making task), and visuospatial abilities (block design, matrix reasoning from the Wechsler Adult Intelligence Scale). Consistent with our prior research, we used two scores more than one standard deviation below the age-corrected mean in one domain or three scores across three domains as an exclusionary screening criterion.

Table 1. Demographic information of participants, split by age group.

Younger Adults Older Adults Total Sample
Variable N (%) Mean (SD) N (%) Mean (SD) N (%) Mean (SD)
Sex
  Female 35 (69%) 92 (73%) 127 (72%)
  Male 16 (31%) 34 (27%) 50 (28%)
Age 24.5 (4.3) 69 (5.5) 56.1 (20.8)
Subjective SES 5.6 (1.8) 6.9 (1.5) 6.5 (1.7)
Years of education 15.8 (2.3) 17.1 (2.1) 16.7 (2.2)
Ethnicity
  Not Hispanic or Latino 41 (80%) 119 (94%) 160 (90%)
  Hispanic or Latino 10 (20%) 7 (6%) 17 (10%)
Race
  African American 0 (0%) 2 (2%) 2 (1%)
  Asian 7 (14%) 0 (0%) 7 (4%)
  Caucasian 39 (76%) 124 (98%) 163 (92%)
  Native American 0 (0%) 0 (0%) 0 (0%)
  Other 5 (10%) 0 (0%) 5 (3%)
Open in a new tab

Note. SES = socioeconomic status. Subjective SES indicates socioeconomic status as measured with the MacArthur Scale of Subjective Social Status.

Materials

Big-Five Factor Personality Assessment.

We measured personality with the Big-Five Factor Personality Assessment 2, Short Form (BFI-2-S; α = 0.77-0.78) (Soto & John, 2017). This is a self-report measure of the Big Five personality traits (openness, conscientiousness, extraversion, agreeableness, and neuroticism), comprising 30 items on a 5-point Likert scale with both standard and reverse-scored items. Participants indicated agreement or disagreement to statements that reflected each of the Big Five traits (e.g., “I am someone who…”: “tends to feel depressed, blue”; neuroticism, “keeps things neat and tidy”; conscientiousness, “is dominant, acts as a leader”; extraversion, “is fascinated by art, music, literature”; openness, “tends to find fault with others”; agreeableness). Participants completed this measure twice, spaced approximately 14 days apart (mean days between sessions = 13.86, median = 13, SD = 5.12, range = 8-42). There was no significant difference between age groups in the number of days between sessions (t(85) = 1.51, p = 0.14; Older adult mean days = 14.8, Younger adult mean days = 13.5).

Ryff Scales of Psychological Well-Being:

To assess subjective well-being, we utilized Ryff Scales of Psychological Well-Being (Ryff, 1989). This self-report scale comprised 42 items (α = 0.71-0.78) on a 7-point Likert scale that measure six aspects of well-being: autonomy (example item: “I have confidence in my opinions, even if they are contrary to the general consensus”), environmental mastery (example item: “In general, I feel I am in charge of the situation in which I live”), personal growth (example item: “I think it is important to have new experiences that challenge how you think about yourself and the world.”, positive relations with others (example item: “People would describe me as a giving person, willing to share my time with others.), purpose in life (example item: “Some people wander aimlessly through life, but I am not one of them.”), and self-acceptance (example item: “I like most aspects of my personality”).

Autobiographical Interview.

We assessed autobiographical memory using the standard procedures for the Autobiographical Interview (AI; Levine et al., 2002). In this task, participants retrieve five autobiographical memories across five lifetime periods. For younger adults, participants recall one memory from early childhood (< age 11), one from their teenage years (age 11-17), two memories from early adulthood (age 17-35), and one memory from last year. Older adult participants retrieve one memory each from early childhood, teenage years, early adulthood, middle adulthood (age 35-55), and the last year. For each retrieval, participants select one specific event that occurred in less than one day. Participants are then cued with a general probe, using the standard AI instructions. However, diverging from AI procedures, we did not include additional probing for specific details.

We audio-recorded, transcribed, and scored the memories using AI scoring procedures (Levine et al., 2002). For each memory, we segmented it into individual details and scored each as either internal (episodic details about the narrative’s main event) or external (non-episodic details). We assigned sub-scores to further categorize internal (i.e., time, place, emotion, perceptual, event) and external (i.e., semantic knowledge, repetitions, meta-cognitive statements), and further categorized semantic knowledge details into either personal semantics (information regarding the self; i.e., “I always hated yams”) or general semantics (general knowledge; i.e., “Paris is the capital of France”). Two trained raters scored each memory, each of whom received a high inter-rater reliability (ICC > 0.90) on a standard training set of memories with an expert scorer. We then averaged across raters for each participant to obtain a more reliable estimate of the participant’s internal and external detail use.

Mind Window mobile application.

Mind Window is a freely available mobile application (Andrews et al., 2024; Puig Rivera et al., 2025) for ecological momentary assessment (EMA) of cognition. The application is formatted with a home page and a notification system and is compatible across mobile phone operating systems. Upon enrollment, participants completed a set of baseline surveys, including socioeconomic status, using the MacArthur Subjective Social Status scale, level of education, and age. Participants who signed up to receive notifications received alerts six times per day at random intervals during participant-set waking hours. When notified, participants completed a Mind Window “check-in”, comprised of survey questions that asked about the nature of participants’ thoughts, mood, and environmental context immediately before receiving the notification. Specifically, participants reported where the focus of their attention was just before the notification, using a proportional slider (thoughts/memories, external perceptions, or bodily sensations). Participants then reported how specific their thoughts were, in terms of time and place, using a sliding scale where the lowest value is “no time/place”, the middle value is “general or vague time/place”, and the highest value is “very specific time and place”. Participants also reported the temporal orientation of their thoughts (more than 2 weeks from now/ago, less than two weeks from now/ago, now, no time at all). To encourage participants to respond to the notification as soon as possible, the application gave participants 10 minutes to complete the check-in after the notification. Additional details on the Mind Window procedure are reported in Puig Rivera et al., 2025.

Procedure

This observational study consisted of 1) an assessment session with a research assistant, 2) ten days in which participants used an ecological momentary assessment mobile application (Mind Window), which notified participants six times per day, and 3) a second assessment session with a research assistant, completed approximately 14 days after the first assessment. The two samples’ protocols were the same with three exceptions: 1) we conducted sessions with Sample 1 virtually on Zoom (except Mind Window procedures), and with Sample 2 in person (except Mind Window procedures), 2) The Autobiographical Interview time limit was four minutes in Sample 1 and five minutes in Sample 2, per memory, 3) While in Sample 2, we used neuropsychological testing to screen both younger and older adults, we only screened older adults in Sample 1.

For both samples, participants completed all assessments as listed in Materials, and several other cognitive and affective measures (to be reported in separate publications) across several visits. Participants in Sample 1 completed assessments over three two-hour Zoom visits. Participants in Sample 2 completed data collection procedures in person at the University of Arizona with a trained research assistant in a private testing room, over three two-hour sessions (including initial baseline cognitive testing). Participants in Sample 2 additionally participated in an MRI scan, the results of which we will describe in a later report.

Analysis

Calculation of indices.

We calculated self-representation stability from two measurements of the BFI-2-S, which were assessed two weeks apart. First, we calculated trait-level personality indices from the BFI-2-S, using the BFI’s standard scoring procedures (Soto & John, 2017), resulting in one average score ranging from one to five for each of the five personality traits. Then, we calculated self-representation stability using double-entry intraclass correlation (ICC) for each participant (Dunkel & Worsley, 2016; Terracciano et al., 2018). This calculation measures ipsative (profile-level) stability of personality ratings and consists of the correlation between two arrays of personality trait scores: 1) scores at assessment 1, and 2) scores at assessment 2, order reversed. The resulting ICC score is a measure of holistic consistency between assessment 1 and assessment 2, ranging from −1 (complete reversal of profile from assessment 1 to assessment 2) to 1 (identical profiles at assessment 1 and assessment 2). We chose this metric, rather than rank-order stability or mean-level change as sometimes done in past research on personality development, to better capture the profile-level stability. In other words, for the purposes of this study, we were less interested in the change or stability of individual traits and more interested in how one’s holistic personality configuration remained similar between assessments. However, for descriptive purposes and comparability to other research involving lifespan development of the Big Five Inventory, we additionally calculated rank-order stability as the test-retest correlation between scores at assessment 1 with assessment 2, across participants within each age group, by individual trait.

We calculated well-being indices using the standard Ryff’s scoring procedures, in which we calculated summed subscales for each of the six facets and the total “sum score” as the sum of all scores across facets (Ryff, 1989). While the sum score was not explicitly mentioned in our preregistration, we calculated and assessed it as it is a standard step in assessing Ryff scores.

We calculated indices of autobiographical memory for both the in-lab and ecological-momentary-assessment-based memory measures. For the in-lab Autobiographical Interview, we calculated episodic specificity as the ratio of internal to total details, and personal semantic details as the ratio of personal semantic to total details. We averaged scores across memories within participants. In the naturalistically assessed autobiographical thought measure (Mind Window), we calculated autobiographical thought frequency as the proportion of Mind Window assessments (out of all assessments) during which: 1) focus of attention was predominantly internally oriented and 2) thoughts were at least minimally self-focused (≥ 0.1 on a scale of 0 to 1). We calculated episodic memory frequency as the proportion of all autobiographical thoughts in which the thought was focused on the past and at least 50% specific in either time or place. These criteria were chosen to ensure 1) the autobiographical thought was a memory (past-focused) and 2) the thought was episodic (specific in time or place), rather than a more general semantic memory, which would lack such specificity.

Statistical analysis

We conducted all preregistered statistical analyses using R statistics version 4.4.1 (“Race for Your Life”) in R Studio version 2023.9.01. We have made the analysis code and output available on our OSF repository. For all models, we assessed assumptions using the Breusch-Pagan test of heteroskedasticity and the Shapiro-Wilk test of non-normality. We assessed the impact of including Study ID (sample 1 vs. sample 2) as a covariate and included this variable in the final results only if it improved model fit. We factored age group as a categorical variable, with younger adults as the reference group. When present, we mean-centered within age group continuous predictor variables. We computed confidence intervals using the rank inversion method , and we corrected for multiple comparisons using the Benjamini-Hochberg method within each research question. We considered corrected p-values that were less than our alpha level of 0.05 as statistically significant.

Sensitivity analysis.

The sample sizes for the samples included in this study were determined based on separate research questions that are or will be published in separate manuscripts. We conducted a sensitivity analysis using the Monte Carlo simulations with the “MASS” package in R to determine the smallest detectable effect size in this sample. For additional sensitivity analyses details, see Supplementary Methods. For our first research question, the minimum detectable effect size was β = 0.47, corresponding to f² = 0.25, representing a medium to large effect. For our remaining research questions, the minimum detectable effect size for the interaction effect was β = 0.49 (given a 0.5 intercept and 0.2 main effects of age group and predictor), corresponding to f² = 0.22, representing a medium interaction effect.

Primary Analyses.

In response to violated assumptions of normality with heavy skew, we used a robust quantile regression for all research questions, as planned in our preregistration. Quantile regressions are advantageous in cases of high skew as they are robust against outliers by estimating conditional quantiles of the outcome variable, while still taking all data points into account (Koenker & Bassett, 1978). Results of models are presented by quantile; however, this does not indicate that findings are limited to any specific quantile, but instead indicates which quantile most strongly drives the effect observed in the overall model. Quantile regression, in addition to controlling for Type II error, offers unique insight into individual differences in the relationship variables, relative to the observed outcome variable distribution- nuances that may otherwise be missed using less robust statistical methods.

We estimated standard errors and p-values using the sandwich method and the Hall-Sheather bandwidth rule. We used the “quantreg” package in R to fit models with four quantiles: 20%, 40%, 60%, and 80%.

To assess differences in self-representation stability between age groups, we fit a quantile regression model, with self-representation stability as the outcome variable and age group as the predictor. To investigate self-representation stability and well-being, we fit seven separate linear regression models, one for each of the dimensions of Ryff’s Scales of Psychological Well-Being, and one for the sum score (Ryff, 1989). These models included age group, self-representation stability, and the interaction of self-representation stability and age group as predictors. To examine the relationship between self-representation stability and autobiographical memory, we fit four regression models with self-representation stability as the outcome variable. Each model contained age group, predictor (each of in-lab episodic specificity, in-lab personal semantics, EMA autobiographical thought frequency, and EMA episodic memory frequency), and their interaction as predictors.

Exploratory Analyses.

As preregistered, we exploratorily assessed whether age effects, relationships between stability and well-being, or stability and autobiographical memory, may be driven by the stability of specific personality traits (openness, conscientiousness, extraversion, agreeableness, or neuroticism) more so than the entire “ipsative” personality profile. First, we calculated trait-level stability as negative one times the absolute value of mean-level change between assessment 2 and assessment 1 for each personality trait separately, where more positive values indicate greater stability. We then repeated all analyses above for each personality trait separately. Results for these analyses are reported in Supplementary Material.

Results

The overall distribution of self-representation stability showed substantial skew (skew = −2.77), resulting in non-normality (W = 0.68, p < 0.05), but variance remained homoscedastic (BP = 1.60, p = 0.21). The skew indicated that participants were, in general, quite stable between the two measurements of the Big Five Inventory (M = 0.90, SD = 0.12, range = 0.26-1). Socioeconomic status (SES) was significantly higher in older adults compared to younger adults (t(82.5) = 4.55, p < 0.001); thus, we ran an additional set of models with SES included as a covariate. Supplementary Table 2 presents descriptive statistics for all variables between and across age groups. Supplementary Table 3 provides rank order and mean-level stability descriptive statistics for comparability with past research that has primarily relied on these metrics.

Do older adults show higher self-representation stability?

Notably, as shown in Table 2 and Supplementary Table 3, participants commonly showed high levels of self-representation stability. In this context, as hypothesized, the quantile regression revealed a significant effect of age group on self-representation stability. Specifically, older adults demonstrated higher self-representation stability than younger adults, and this effect was prominent at higher levels of self-representation stability (e.g., at 80% of self-representation stability, p < 0.05) (see Figure 1). In the model including SES as a covariate, this effect was replicated (see Supplementary Table 4).

Table 2. Quantile regression model details: Age differences in self-representation stability.

20% 40% 60% 80%
Predictor B (SE) t 95% CI p B (SE) t 95% CI p B (SE) t 95% CI p B (SE) t 95% CI p
(Intercept) 0.82 (0.04) 20.35 [0.76, 0.90] <0.001 0.93 (0.01) 63.69 [0.90, 0.94] <0.001 0.95 (0.01) 136.14 [0.94, 0.96] <0.001 0.97 (0.00) 257.06 [0.96, 0.97] <0.001
Age group (older) 0.03 (0.04) 0.66 [−0.05, 0.10] 0.51 0.01 (0.02) 0.32 [−0.01, 0.04] 0.75 0.01 (0.01) 1.86 [−0.00, 0.03] 0.06 0.01 (0.00) 2.68 [0.00, 0.02] 0.01
Open in a new tab

Note: Results are split by quantile (20%, 40%, 60%, 80%). B = beta (β) coefficient. SE = standard error. 95% CI = 95% confidence interval. Bolded p values indicate significance at the 0.05 alpha level.

Figure 1. Age differences in self-representation stability.

Figure 1.

Open in a new tab

Note: Older adults (blue) show higher self-representation stability than younger adults (red), and this effect was most prominent at higher levels of self-representation. Panel A: Stability coefficients by age group and tau level. Coefficient indicates estimate of self-representation stability, as measured by two-week consistency in self-appraised personality characteristics. Tau indicates quantile (20%, 40%, 60%, 80%). Shaded band indicates the 95% confidence interval of the coefficient. Panel B: Box plot depicting age differences in self-representation stability at Tau = 0.8. Dots indicate individual datapoints.

Does self-representation stability predict well-being?

Five participants did not complete Ryff’s Scales of Psychological Well-being, resulting in 172 participants included (51 younger adults, 121 older adults). We observed a significant positive effect of self-representation stability on overall well-being across age groups (pcorrected < 0.001); see Figure 2 Panel A. Supporting our preregistered hypothesis, across age groups, self-representation stability significantly positively predicted Autonomy, and the quantile analysis showed that this was particularly the case for those low in Autonomy (in the 20% quantile) (pcorrected < 0.01; Figure 2 Panel B). Across age groups, self-representation stability was also positively related to Positive Relations (pcorrected < 0.01) and Life Purpose (pcorrected < 0.01), both driven by those at the highest end of the distribution on this measure of well-being (80% quantile) (Figure 2, Panels C and D). Lastly, self-representation stability significantly positively predicted environmental mastery, especially among individuals with moderately high (i.e., in the 60% quantile) well-being on this measure (pcorrected < 0.05; Figure 2 Panel E). Table 3 presents a summary of significant results, and Supplementary Table 5 presents all model details. In the models including SES as a covariate, these results were largely replicated (see Supplementary Table 6) except: 1) neither sum well-being score nor life purpose was predicted by stability, 2) the main effect on Positive Relations was present in both the 80% (pcorrected < 0.001) and 20% (pcorrected < 0.01) quantile. No interaction effects were observed in any subscales in these models.

Figure 2. Self-representation stability predicting well-being.

Figure 2.

Open in a new tab

Note: Coefficient indicates the estimate change in outcome variable, given a one-unit increase in self-representation stability. Tau indicates quantile (20%, 40%, 60%, 80%). Shaded band indicates the 95% confidence interval of the coefficient. Panel A shows self-representation stability predicting Ryff’s sum score of well-being. We observed a significant positive effect of self-representation stability. Note that the y axis is scaled at 0-150 to better show effects across quantiles, but some confidence intervals (20% quantile, see Table 3) reach beyond these values. Panel B shows self-representation stability positively predicting Ryff’s autonomy facet of well-being across age groups, particularly at the lowest quantile. Panel C shows self-representation stability positively predicting Ryff’s environmental mastery facet of well-being across well-being, particularly at the 60% quantile. Panel D shows self-representation stability positively predicting Ryff’s positive relations facet of well-being, across age groups, particularly at the highest quantile. Panel E shows self-representation stability positively predicting Ryff’s life purpose facet of well-being, across age groups, particularly in the highest quantile.

Table 3. Summary of significant results for self-representation stability predicting well-being.

Predictor Outcome Quantile Direction B (SE) t 95% CI p corrected
Self-Representation Stability Sum Score 80% + 15.92 (2.57) 6.2 [−8.61, 23.58] < 0.001
Positive Relations 80% + 1.78 (0.59) 3.01 [−11.34, 7.77] < 0.01
Life Purpose 80% + 1.94 (0.62) 3.11 [−10.56, 9.00] < 0.01
Environmental Mastery 60% + 11.93 (3.72) 3.21 [6.15, 14.74] < 0.01
Autonomy 20% + 19.12 (5.81) 3.29 [−14.33, 32.96] < 0.01
Open in a new tab

Note: Supplementary Table hows only results where pcorrected < 0.05 for the main effect of self-representation stability or the self-representation stability X age group interaction. Full model results shown in Supplementary Table . B = beta (β) coefficient. SE = standard error. 95% CI = 95% confidence interval. p values are corrected using Benjamini-Hochberg correction by quantile.

Does in-lab autobiographical memory predict self-representation stability?

Sixteen participants were missing datapoints for the in-lab autobiographical interview, resulting in 161 participants included in this model (43 younger adults, 118 older adults). As hypothesized, episodic specificity did not significantly predict self-representation stability, nor interact with age group at any investigated quantile (pscorrected > 0.05). Contrary to our hypothesis, personal semantics also did not significantly predict self-representation stability, nor interact with age group at any investigated quantile (pscorrected > 0.05). We present model details in Table 4. In the models including SES as a covariate, these results were replicated (Table S7).

Table 4. Quantile regression model details: Self-representation stability predicted by in-lab autobiographical memory.

20% 40% 60% 80%
Predictor B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected
Episodic Specificity
 (Intercept) 0.81 (0.04) 20.07 [0.76, 0.86] <0.001 0.88 (0.03) 27.49 [0.85, 0.94] <0.001 0.94 (0.02) 48.92 [0.89, 0.96] <0.001 0.97 (0.01) 127.27 [0.96, 0.98] <0.001
 Episodic Specificity 0.35 (0.28) 1.24 [0.25, 0.88] 0.22 0.37 (0.19) 1.95 [−0.10, 0.52] 0.11 0.02 (0.11) 0.2 [−0.13, 0.38] 0.84 0.02 (0.05) 0.32 [−0.02, 0.10] 0.89
 age_group 0.04 (0.04) 0.87 [−0.02, 0.11] 0.77 0.06 (0.03) 1.73 [−0.01, 0.10] 0.17 0.02 (0.02) 1.21 [0.00, 0.07] 0.24 0.01 (0.01) 1.45 [0.00, 0.03] 0.17
 Episodic Specificity:age_group −0.27 (0.33) −0.83 [−0.82, −0.05] 0.41 −0.30 (0.22) −1.42 [−0.51, 0.20] 0.31 −0.03 (0.12) −0.23 [−0.37, 0.15] 0.82 −0.04 (0.05) −0.7 [−0.14, 0.04] 0.48
Personal Semantics
 (Intercept) 0.84 (0.03) 33.12 [0.81, 0.88] <0.001 0.90 (0.02) 37.92 [0.88, 0.92] <0.001 0.95 (0.01) 69.95 [0.92, 0.96] <0.001 0.97 (0.01) 155.92 [0.96, 0.98] <0.001
 Personal Semantics −0.50 (0.27) −1.87 [−0.91, −0.29] 0.13 −0.31 (0.27) −1.14 [−0.51, 0.04] 0.26 −0.12 (0.15) −0.77 [−0.38, 0.12] 0.84 0.01 (0.07) 0.13 [−0.15, 0.03] 0.89
 age_group 0.01 (0.03) 0.26 [−0.05, 0.05] 0.8 0.03 (0.03) 1.25 [0.01, 0.06] 0.21 0.02 (0.01) 1.18 [0.00, 0.05] 0.24 0.01 (0.01) 1.37 [0.00, 0.03] 0.17
 Personal Semantics:age_group 0.46 (0.39) 1.17 [0.11, 1.04] 0.41 0.26 (0.31) 0.83 [0.05, 0.61] 0.41 0.12 (0.17) 0.74 [−0.15, 0.44] 0.82 0.08 (0.08) 1.09 [−0.06, 0.28] 0.48
Open in a new tab

Note: Results are split by quantile (20%, 40%, 60%, 80%). B = beta (β) coefficient. SE = standard error. 95% CI = 95% confidence interval. Bolded p values indicate significance at the 0.05 alpha level. p values are corrected using Benjamini-Hochberg correction by quantile. Predictor variables of episodic specificity and personal semantics are mean-centered by age group.

Does autobiographical thought in daily life predict self-representation stability?

To be included in this analysis, participants were required to have completed at least ten Mind Window check-ins (94% of young adults and 100% of older adults). Three participants did not complete Mind Window protocols (i.e., did not complete at least 10 check-ins), resulting in 174 participants included (48 younger adults, 126 older adults). Among these participants, compliance in Mind Window protocols was very similar between age groups: during the study period, older adults completed 85% of surveys, and younger adults completed 83% of surveys. Several participants completed surveys beyond the 10-day window, but overall mean and median survey completion numbers were similar between age groups (older adults: mean = 77.47 surveys, median = 51 surveys; younger adults: mean = 56.96 surveys, median = 50 surveys). We observed a positive main effect of autobiographical thought frequency in daily life on self-representation stability (pcorrected < 0.05), driven by the 60% quantile. However, supporting our hypothesis, we observed a significant interaction effect between age group and autobiographical thought frequency in daily life (pcorrected < 0.05), indicating that thought frequency was positively associated with self-representation stability only in younger adults. This interaction effect was most prominent at moderate levels of self-representation stability (e.g., at 60% of self-representation stability) (Figure 3). Contrary to our hypothesis, we did not observe main or interaction effects in the model predicting self-representation stability from episodic memory frequency (pscorrected > 0.05). We present model details in Table 5. In models including SES as a covariate, these results were replicated (Table S8).

Figure 3. Self-representation stability predicted by autobiographical thought frequency in daily life.

Figure 3

Open in a new tab

Note: Coefficient indicates the estimate change in self-representation stability, given a one-unit increase in autobiographical thought frequency. Tau indicates quantile (20%, 40%, 60%, 80%). Shaded band indicates the 95% confidence interval of the coefficient. We observed a significant interaction effect between age group and autobiographical thought frequency in daily life, such that thought frequency was positively associated with self-representation stability only in younger adults, most prominent at moderate levels of self-representation stability.

Table 5. Quantile regression model details: Self-representation stability predicted by daily autobiographical thought and memory.

20% 40% 60% 80%
Predictor B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected B (SE) t 95%
CI		 p corrected
Episodic Memory in Daily Life
(Intercept) 0.85 (0.04) 20 [0.78, 0.87] <0.001 0.90 (0.02) 39.7 [0.88, 0.93] <0.001 0.95 (0.01) 108.4 [0.93, 0.96] <0.001 0.97 (0.01) 122.1 [0.95, 0.97] <0.001
mem_mc −0.26 (0.23) −1.17 [−0.92, 0.01] 0.48 −0.32 (0.22) −1.43 [−0.35, 0.10] 0.16 −0.04 (0.07) −0.48 [−0.23, 0.08] 0.63 −0.00 (0.08) −0.07 [−0.23, 0.14] 0.94
age_group 0.00 (0.05) 0.1 [−0.03, 0.10] 0.92 0.04 (0.02) 1.56 [−0.00, 0.06] 0.24 0.01 (0.01) 1.54 [0.00, 0.04] 0.13 0.01 (0.01) 1.33 [0.00, 0.03] 0.2
mem_mc:age_group 0.22 (0.32) 0.69 [0.01, 0.84] 0.49 0.23 (0.23) 0.98 [−0.28, 0.31] 0.57 0.01 (0.09) 0.13 [−0.10, 0.19] 0.89 −0.00 (0.08) −0.02 [−0.07, 0.24] 0.98
Autobiographical Thought in Daily Life
(Intercept) 0.84 (0.04) 20.3 [0.75, 0.88] <0.001 0.91 (0.02) 39.8 [0.88, 0.93] <0.001 0.94 (0.01) 97.18 [0.93, 0.95] < 0.001 0.97 (0.01) 120.4 [0.95, 0.97] <0.001
thought_mc 0.17 (0.25) 0.7 [−0.35, 0.26] 0.48 0.09 (0.05) 1.66 [0.00, 0.11] 0.16 0.05 (0.02) 2.67 [0.02, 0.08] 0.02 0.01 (0.02) 0.73 [−0.01, 0.09] 0.93
age_group 0.02 (0.05) 0.34 [−0.03, 0.10] 0.92 0.02 (0.03) 0.8 [−0.00, 0.06] 0.42 0.02 (0.01) 2.23 [0.01, 0.04] 0.05 0.01 (0.01) 1.28 [0.00, 0.03] 0.2
thought_mc:age_group −0.21 (0.28) −0.76 [−0.36, 0.46] 0.49 −0.04 (0.08) −0.57 [−0.19, 0.03] 0.57 −0.07 (0.03) −2.38 [−0.10, −0.03] 0.04 −0.04 (0.02) −1.99 [−0.10, −0.02] 0.1
Open in a new tab

Note: Results are split by quantile (20%, 40%, 60%, 80%). B = beta (β) coefficient. SE = standard error. 95% CI = 95% confidence interval. Bolded p values indicate significance at the 0.05 alpha level. p values are corrected using Benjamini-Hochberg correction by quantile. Predictor variables of daily autobiographical thought and episodic memory are mean-centered by age group.

Discussion

This study examined implicit self-representation stability, measured with self-appraised personality consistency across a two-week interval, and its relationship to autobiographical memory and well-being in younger and older adults. Results largely supported our pre-registered hypotheses: 1) older adults showed higher self-representation stability than younger adults, 2) self-representation stability was related to well-being across age groups, 3) in-lab episodic specificity was unrelated to self-representation stability, and 4) in young adults, self-representation stability was positively related to autobiographical thought frequency in everyday life. Our hypotheses that stability would be predicted by in-lab measures of personal semantics in older adults or episodic memory frequency in everyday life in young adults were not supported.

Older adults have greater self-representation stability

In line with our hypothesis, self-representation stability was higher in older than in younger adults. This was most prominent at the highest end of self-representation stability’s distribution, aligning with identity development models (Erikson, 1963), and suggestions that older adults utilize narrative (Liao & Bluck, 2023; Lind et al., 2021; McAdams, 2013) and adaptive (Brandtstädter & Greve, 1994) strategies to promote self-coherence. These findings complement previous work showing that the subjective sense of a clear (Diehl & Hay, 2011; Fuentes & Desrocher, 2012) and continuous (Lampraki et al., 2023; Rutt & Löckenhoff, 2016) self increases with age. Here, we offer novel insight into age differences in self-representation stability by utilizing an implicit, objective measurement of acute self-evaluation consistency, highlighting an important positive aspect of aging.

Self-representation stability is positively related to well-being

Older adults’ advantage in self-representation stability is notable due to its implications for well-being. Self-representation stability was positively related to overall well-being (the sum of all Ryff’s subscales) and, among individual subscales, we observed a positive relationship with stability for both self-related (i.e., autonomy, in line with our preregistered hypotheses) and other (environmental mastery, life purpose, positive relations, environmental mastery) subscales. The autonomy effect was driven by the lowest quantile, suggesting individuals with the lowest personal autonomy experience the strongest link between feelings of autonomy and stability. Although we are limited in our ability to infer causality, these results invite the possibility that self-representation stability may help to support well-being overall, rather than being restricted to just self-related functions. These findings mirror previous work showing that individuals with higher subjective self-concept clarity had greater autonomy and environmental mastery (Diehl & Hay, 2011).

Autobiographical processes in daily life predict stability in young adults

We next explored cognitive mechanisms supporting self-representation stability. We found no support for our hypothesis that in-lab personal semantics would predict stability. While self-judgments rely on semantic information (Grilli & Verfaellie, 2015; Klein & Loftus, 1993), their recall in a separate task did not relate to self-representation stability. One interpretation of this result is methodological: the Autobiographical Interview captures a broad range of personal semantic details (e.g., “I grew up in Seattle”, “I am a vegetarian”), which may not always overlap with abstract trait-based knowledge that anchors stable self-representations. It is possible that a more trait-specific measure within autobiographical memory may have yielded different results. Alternatively, this result may suggest that recalling semantic details in the lab may simply be less relevant to self-representation stability than meaning-making and reflection that occur in daily life. In daily life, personal semantic knowledge is accessed and updated in socially embedded and perhaps more trait-relevant situations, rather than a decontextualized lab environment. Future research could employ naturalistic measures that may better assess personal semantic knowledge in daily life to understand this distinction better.

Our hypotheses about episodic memory were partially supported. As expected, in-lab episodic autobiographical memory was unrelated to self-representation stability, suggesting that trait-like episodic memory ability may not maintain self-concept. We also predicted that younger adults’ episodic recall in daily life would support self-representation stability, but found no such relationship. Instead, younger adults’ broader autobiographical thoughts (encompassing episodic, semantic, future, and present-oriented self-related thoughts) were positively related to stability. Episodic retrieval in daily life may be briefer, more fragmented, less voluntary, and more contextually cued, rather than memories retrieved deliberately in laboratory settings (Rasmussen & Berntsen, 2011). This may limit its contribution to broader identity processes unless the memory is elaborated or connected to other self-relevant thoughts. Among younger adults, self-relevant reflection, rather than episodic recall alone, may support self-representation stability by integrating experiences into the developing self. Accordingly, frequent episodic memory evocation may reflect an earlier phase of self-concept development.

Overall, the fact that these findings emerged in younger, but not older adults, extends previous research showing that individuals with lower self-concept clarity (i.e., young adults) use autobiographical memory to maintain subjective self-continuity (Bluck & Alea, 2009). It also aligns with evidence that threatened self-concept clarity prompts autobiographical recall, increasing self-continuity (Jiang et al., 2020). Our results offer converging evidence using an objective measure of self-stability, supplementing previous work that focused on the phenomenological experience of continuity.

Additionally, the involvement of only daily-life autobiographical processes in implicit stability illustrates an important distinction between in-lab and naturalistic measures. While laboratory measures capture purposeful, top-down retrieval, EMA captures spontaneous, involuntary, and privately experienced memories triggered by context or immediate situations (Berntsen, 2020; Craik, 2022). Indeed, we observed in related work that naturalistic memories do not appear to share the same age differences as observed in the lab. While younger adults often demonstrate higher episodic specificity than older adults in laboratory settings (Simpson et al., 2023), this age difference is either absent (McVeigh et al., 2025) or reversed (Puig Rivera et al., 2025) across age groups when assessed naturalistically. This discrepant pattern appears attributable to older adults’ significantly elevated episodic specificity in real-world settings (McVeigh et al., 2025). Here, we show that daily-life autobiographical processes that support self-stability differ by age group, extend beyond episodic memory, and are not reflected in laboratory-based in-lab measures. More research is needed to explore further these relationships, including the content and function of daily autobiographical thought on sense of self across the lifespan.

Assessing self-representation stability with ICC

Lastly, our research supports using double-entry intraclass correlations to measure the stability of objective self-representation. Unlike rank-order stability or mean-level change, ICC offers a holistic, ipsative view of consistency by capturing both profile shape and elevation. Previous studies show that ipsative personality stability better predicts identity continuity than rank-order stability (Dunkel & Worsley, 2016), because self-appraisals of personality traits are not entirely independent, and changes in one domain may be linked to shifts in others. Rank-order stability is too insensitive in such cases, because correlations stay high if the profile’s shape remains intact even when overall levels change. Likewise, mean-level change only reflects a profile’s average level, without considering shape. In contrast, ICC detects both elevation and shape differences, providing a more conservative and precise measure (McCrae, 2008). McCrae’s (2008) study compared various profile agreement indices, finding that double-entry ICC was the most accurate, surpassing rank-order stability and other options. Therefore, ICC offers a more comprehensive metric of stability, capturing how traits are arranged as a whole.

Limitations

Our sample had two primary limits to generalizability. First, we included more older adults than younger adults. Second, our sample was not representative of the United States or Arizona, with more women and non-Hispanic Caucasian participants than population estimates. The older adult group was also more homogeneous than the younger adult group, with a higher proportion of Caucasian participants (98% compared to 76% in the younger adult group). Lack of diversity and representativeness remain key issues in psychology (Roberts & Mortenson, 2023). Future research would benefit from more representative sampling and better demographic matching between age groups.

Additionally, we recognize that self-representation stability may have been more precisely assessed with additional measurements (e.g., three timepoints). However, our design makes the observed findings especially compelling, as they emerged even with two time-points across a relatively short interval of 2 weeks. Personality measures typically show changes across much longer intervals (months to years), and our use of a short interval was intended to isolate consistency in self-appraisals from more substantial changes in personality associated with developmental change or late-life cognitive decline. Because personality is unlikely to shift meaningfully within two weeks, we propose that any observed inconsistencies more clearly reflect the construct of interest: self-representation stability.

Similarly, we note that we investigated only one component of self-concept (personality) while self-representations include many additional elements outside our study’s scope. Yet, our choice of metric is validated by the fact that we replicated previous findings associated with holistic self-concept stability.

Conclusion

In this study, older adults aged 60-90 demonstrated higher self-representation stability relative to young adults aged 18-35, and higher stability was positively related to well-being. Younger adults’ everyday autobiographical thought tendencies supported greater self-representation stability, potentially reflecting developmentally appropriate self-updating, but in-lab memory remained unrelated across age groups. These findings shed light on an important older adult advantage and provide insight into the role of autobiographical processes in supporting younger adults’ stable self-representation maintenance.

Supplementary Material

1

Acknowledgments

We acknowledge members of our research group who contributed to the coordination of this work, including Cait Cegavske, Chris Griffith, Alexis Garcia, and Vannia Puig-Rivera. We thank Eric Andrews for his invaluable contribution in developing the Mind Window app. Additionally, we thank the team of undergraduate research assistants in the Human Memory Lab who assisted in the manual scoring of Autobiographical Interview data.

Funding

This research was supported by National Institutes of Aging grants (R01AG068098 & R56AG068098) awarded to Jessica Andrews-Hanna and Matthew Grilli, as well as the Arizona Department of Health Services/Arizona Alzheimer’s Consortium.

Footnotes

Conflicts of interest

All authors declare no conflicts of interest.

AI Use

Artificial intelligence, with OpenAI’s ChatGPT, was used for syntax troubleshooting in R. Otherwise, artificial intelligence was not used in the creation of this article.

Data availability

This study was preregistered on Open Science Framework (https://osf.io/8vzj2/?view_only=afad3f7ac1394bf2b79c92b7838e5116) on March 11, 2025, before data analysis, which began on March 13, 2025. Data collection occurred before preregistration to investigate a different research question. Data related to the present research question were not examined until after preregistration. Deviations to preregistration are reported in Supplementary Table 1 (we report one deviation related to sample size and one related to an added descriptive variable). All study materials, including hypotheses, data, and analysis plan/scripts, are publicly available via the preregistration link.

References

  1. Addis DR, Roberts RP, & Schacter DL (2011). Age-related neural changes in autobiographical remembering and imagining. Neuropsychologia, 49(13), 3656–3669. 10.1016/j.neuropsychologia.2011.09.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Andrews E, Grilli M, Bethard S, Mehl M, Allen J, Arch J, & Andrews-Hanna J (2024). Patterns of Everyday Thought as Correlates of Psychological Well-Being and Possible Targets for Personalized Mental Health Treatment. OSF. 10.31234/osf.io/9vwkt [DOI] [Google Scholar]
  3. Andrews-Hanna JR, & Grilli MD (2021). Mapping the imaginative mind: Charting new paths forward. Current Directions in Psychological Science, 30(1), 82–89. 10.1177/0963721420980753 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baumeister RF (1998). The self. In The handbook of social psychology, Vols. 1-2, 4th ed (pp. 680–740). McGraw-Hill. [Google Scholar]
  5. Berntsen D (2020). Involuntary autobiographical memories and their relation to other forms of spontaneous thoughts. Philosophical Transactions of the Royal Society B: Biological Sciences, 376(1817), 20190693. 10.1098/rstb.2019.0693 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bleidorn W, Schwaba T, Zheng A, Hopwood CJ, Sosa SS, Roberts BW, & Briley DA (2022). Personality stability and change: A meta-analysis of longitudinal studies. Psychological Bulletin, 148(7–8), 588–619. 10.1037/bul0000365 [DOI] [PubMed] [Google Scholar]
  7. Bluck S, & Alea N (2009). Thinking and talking about the past: Why remember? Applied Cognitive Psychology, 23(8), 1089–1104. 10.1002/acp.1612 [DOI] [Google Scholar]
  8. Brandtstädter J, & Greve W (1994). The Aging Self: Stabilizing and Protective Processes. Developmental Review, 14(1), 52–80. 10.1006/drev.1994.1003 [DOI] [Google Scholar]
  9. Burns RA (2020). Age-Related Differences in the Factor Structure of Multiple Wellbeing Indicators in a Large Multinational European Survey. Journal of Happiness Studies, 21(1), 37–52. 10.1007/s10902-019-00077-y [DOI] [Google Scholar]
  10. Butzer B, & Kuiper NA (2006). Relationships between the frequency of social comparisons and self-concept clarity, intolerance of uncertainty, anxiety, and depression. Personality and Individual Differences, 41(1), 167–176. 10.1016/j.paid.2005.12.017 [DOI] [Google Scholar]
  11. Carstensen LL, & Mikels JA (2005). At the Intersection of Emotion and Cognition: Aging and the Positivity Effect. Current Directions in Psychological Science, 14(3), 117–121. 10.1111/j.0963-7214.2005.00348.x [DOI] [Google Scholar]
  12. Chessell ZJ, Rathbone CJ, Souchay C, Charlesworth L, & Moulin CJA (2014). Autobiographical Memory, Past and Future Events, and Self-images in Younger and Older Adults. Self and Identity, 13(4), 380–397. 10.1080/15298868.2013.836132 [DOI] [Google Scholar]
  13. Conway MA (2005). Memory and the self. Journal of Memory and Language, 53(4), 594–628. 10.1016/j.jml.2005.08.005 [DOI] [Google Scholar]
  14. Craik FIM (2022). Reducing age-related Memory Deficits: The Roles of Environmental Support and self-initiated Processing Activities. Experimental Aging Research, 48(5), 401–427. 10.1080/0361073X.2022.2084660 [DOI] [PubMed] [Google Scholar]
  15. Cuervo-Lombard C, Raucher-Chéné D, Linden M. V. d., & Voltzenlogel V (2021). Characteristics of Self-Defining Memories in Middle-Aged and Older Adults. Current Aging Science, 14(1), 39–45. 10.2174/1874609813666201006142514 [DOI] [PubMed] [Google Scholar]
  16. Devitt AL, Roberts R, Metson A, Tippett LJ, & Addis DR (2023). Neural substrates of specific and general autobiographical memory retrieval in younger and older adults. Neuropsychologia, 108754. 10.1016/j.neuropsychologia.2023.108754 [DOI] [PubMed] [Google Scholar]
  17. Diehl M, & Hay EL (2011). Self-Concept Differentiation and Self-Concept Clarity across Adulthood: Associations with Age and Psychological Well-Being. The International Journal of Aging and Human Development, 73(2), 125–152. 10.2190/AG.73.2.b [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Dunkel CS, & Worsley SK (2016). Does identity continuity promote personality stability? Journal of Research in Personality, 65, 11–15. 10.1016/j.jrp.2016.09.001 [DOI] [Google Scholar]
  19. Duval C, Desgranges B, de La Sayette V, Belliard S, Eustache F, & Piolino P (2012). What happens to personal identity when semantic knowledge degrades? A study of the self and autobiographical memory in semantic dementia. Neuropsychologia, 50(2), 254–265. 10.1016/j.neuropsychologia.2011.11.019 [DOI] [PubMed] [Google Scholar]
  20. Erikson EH (1963). Childhood and Society (Vol. 2). Norton. https://brill.com/display/book/edcoll/9783657768387/B9783657768387_s048.xml [Google Scholar]
  21. Fivush R, & Grysman A (2023). Accuracy and reconstruction in autobiographical memory: (Re)consolidating neuroscience and sociocultural developmental approaches. WIREs Cognitive Science, 14(3), e1620. 10.1002/wcs.1620 [DOI] [PubMed] [Google Scholar]
  22. Fleeson W, & Gallagher P (2009). The implications of Big Five standing for the distribution of trait manifestation in behavior: Fifteen experience-sampling studies and a meta-analysis. Journal of Personality and Social Psychology, 97(6), 1097–1114. 10.1037/a0016786 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Fuentes A, & Desrocher M (2012). Autobiographical Memory in Emerging Adulthood: Relationship with Self-Concept Clarity. Journal of Adult Development, 19(1), 28–39. 10.1007/s10804-011-9131-1 [DOI] [Google Scholar]
  24. Grilli MD (2017). The association of personal semantic memory to identity representations: Insight into higher-order networks of autobiographical contents. Memory, 25(10), 1435–1443. 10.1080/09658211.2017.1315137 [DOI] [PubMed] [Google Scholar]
  25. Grilli MD, & Sheldon S (2022). Autobiographical event memory and aging: Older adults get the gist. Trends in Cognitive Sciences, 26(12), 1079–1089. 10.1016/j.tics.2022.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Grilli MD, & Verfaellie M (2015). Supporting the self-concept with memory: Insight from amnesia. Social Cognitive and Affective Neuroscience, 10(12), 1684–1692. 10.1093/scan/nsv056 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Jiang T, Chen Z, & Sedikides C (2020). Self-concept clarity lays the foundation for self-continuity: The restorative function of autobiographical memory. Journal of Personality and Social Psychology, 119(4), 945–959. 10.1037/pspp0000259 [DOI] [PubMed] [Google Scholar]
  28. Klein SB, & Lax ML (2010). The unanticipated resilience of trait self-knowledge in the face of neural damage. Memory, 18(8), 918–948. 10.1080/09658211.2010.524651 [DOI] [PubMed] [Google Scholar]
  29. Klein SB, & Loftus J (1993). Behavioral Experience and Trait Judgnents about the Self. Personality and Social Psychology Bulletin, 19(6), 740–745. 10.1177/0146167293196009 [DOI] [Google Scholar]
  30. Koenker R (2005). Quantile Regression. Cambridge University Press. 10.1017/CBO9780511754098 [DOI] [Google Scholar]
  31. Koenker R, & Bassett G (1978). Regression Quantiles. Econometrica, 46(1), 33–50. 10.2307/1913643 [DOI] [Google Scholar]
  32. Lampraki C, Spini D, & Jopp DS (2023). Life Span Development of Self-Continuity in late Adulthood: The Impact of Chronological Age, Loss of Partner and Aversive Childhood Events. Self and Identity, 22(4), 669–688. 10.1080/15298868.2022.2157869 [DOI] [Google Scholar]
  33. Levine B, Svoboda E, Hay JF, Winocur G, & Moscovitch M (2002). Aging and autobiographical memory: Dissociating episodic from semantic retrieval. Psychology and Aging, 17(4), 677–689. 10.1037/0882-7974.17.4.677 [DOI] [PubMed] [Google Scholar]
  34. Liao H-W, & Bluck S (2023). Recalling self-disruptive events and maintaining self-continuity in adulthood. Psychology and Aging, 38(1), 17–29. 10.1037/pag0000719 [DOI] [PubMed] [Google Scholar]
  35. Lind M, Bluck S, & McAdams DP (2021). More Vulnerable? The Life Story Approach Highlights Older People’s Potential for Strength During the Pandemic. The Journals of Gerontology: Series B, 76(2), e45–e48. 10.1093/geronb/gbaa105 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Markus H, & Wurf E (1987). The dynamic self-concept: A social psychological perspective. Annual Review of Psychology, 38, 299–337. 10.1146/annurev.ps.38.020187.001503 [DOI] [Google Scholar]
  37. McAdams DP (2013). The Psychological Self as Actor, Agent, and Author. Perspectives on Psychological Science, 8(3), 272–295. 10.1177/1745691612464657 [DOI] [PubMed] [Google Scholar]
  38. McCrae RR (2008). A Note on Some Measures of Profile Agreement. Journal of Personality Assessment, 90(2), 105–109. 10.1080/00223890701845104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McVeigh K, Deffner A, Hernandez D, Sbarra D, Andrews-Hanna J, Mehl M, & Grilli M (2025). In Here and Out There: Evidence That Age-Related Differences in Memory Specificity are Attenuated in Natural Social Conversations. OSF. 10.31234/osf.io/sja6t_v2 [DOI] [Google Scholar]
  40. Prebble SC, Addis DR, & Tippett LJ (2013). Autobiographical memory and sense of self. Psychological Bulletin, 139(4), 815–840. 10.1037/a0030146 [DOI] [PubMed] [Google Scholar]
  41. Puig Rivera VA, Andrews ES, Cervantes LJ, Freveletti D, Huentelman M, Grilli MD, & Andrews-Hanna JR (2025, June). Sharper With Age? Enhanced Episodic Specificity and Socioemotional Content in Older Adults’ Everyday Autobiographical Thoughts. 10.17605/OSF.IO/ZV2AX [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Puig Rivera V, Grilli M, & Andrews-Hanna J (2025). An Examination of Age-Related Differences in Autobiographical Thinking using an Experience Sampling Approach: Insights from the Mind Window App. 10.17605/OSF.IO/ZV2AX [DOI] [Google Scholar]
  43. Rasmussen AS, & Berntsen D (2011). The unpredictable past: Spontaneous autobiographical memories outnumber autobiographical memories retrieved strategically. Consciousness and Cognition, 20(4), 1842–1846. 10.1016/j.concog.2011.07.010 [DOI] [PubMed] [Google Scholar]
  44. Rathbone CJ, Moulin CJA, Conway MA, & Holmes EA (2012a). Autobiographical memory and the self. In Cognitive psychology, 2nd ed (pp. 546–576). Oxford University Press. [Google Scholar]
  45. Rathbone CJ, Moulin CJA, Conway MA, & Holmes EA (2012b). Autobiographical memory and the self. In Cognitive psychology, 2nd ed (pp. 546–576). Oxford University Press. [Google Scholar]
  46. Richman SB, Pond RS, Dewall CN, Kumashiro M, Slotter EB, & Luchies LB (2016). An Unclear Self Leads to Poor Mental Health: Self-Concept Confusion Mediates the Association of Loneliness with Depression. Journal of Social and Clinical Psychology, 35(7), 525–550. 10.1521/jscp.2016.35.7.525 [DOI] [Google Scholar]
  47. Ritchie TD, Sedikides C, Wildschut T, Arndt J, & Gidron Y (2011). Self-concept Clarity Mediates the Relation between Stress and Subjective Well-being. Self and Identity, 10(4), 493–508. 10.1080/15298868.2010.493066 [DOI] [Google Scholar]
  48. Roberts SO, & Mortenson E (2023). Challenging the White = Neutral Framework in Psychology. Perspectives on Psychological Science, 18(3), 597–606. 10.1177/17456916221077117 [DOI] [PubMed] [Google Scholar]
  49. Robinson MD, & Clore GL (2002). Episodic and semantic knowledge in emotional self-report: Evidence for two judgment processes. Journal of Personality and Social Psychology, 83(1), 198–215. 10.1037/0022-3514.83.1.198 [DOI] [PubMed] [Google Scholar]
  50. Rutt JL, & Löckenhoff CE (2016). From past to future: Temporal self-continuity across the life span. Psychology and Aging, 31(6), 631–639. 10.1037/pag0000090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Ryff CD (1989). In the eye of the beholder: Views of psychological well-being among middle-aged and older adults. Psychology and Aging, 4(2), 195–210. 10.1037/0882-7974.4.2.195 [DOI] [PubMed] [Google Scholar]
  52. Sedikides C (2021). Self-Construction, Self-Protection, and Self-Enhancement: A Homeostatic Model of Identity Protection. Psychological Inquiry, 32(4), 197–221. 10.1080/1047840X.2021.2004812 [DOI] [Google Scholar]
  53. Sedikides C, Hong EK, & Wildschut T (2023). Self-Continuity. Annual Review of Psychology, 74(Volume 74, 2023), 333–361. 10.1146/annurev-psych-032420-032236 [DOI] [PubMed] [Google Scholar]
  54. Sedikides C, Wildschut T, Cheung W-Y, Routledge C, Hepper EG, Arndt J, Vail K, Zhou X, Brackstone K, & Vingerhoets AJJM (2016). Nostalgia fosters self-continuity: Uncovering the mechanism (social connectedness) and consequence (eudaimonic well-being). Emotion, 16(4), 524–539. 10.1037/emo0000136 [DOI] [PubMed] [Google Scholar]
  55. Simpson S, Eskandaripour M, & Levine B (2023). Effects of Healthy and Neuropathological Aging on Autobiographical Memory: A Meta-Analysis of Studies Using the Autobiographical Interview. The Journals of Gerontology: Series B, gbad077. 10.1093/geronb/gbad077 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Soto CJ, & John OP (2017). Short and extra-short forms of the Big Five Inventory–2: The BFI-2-S and BFI-2-XS. Journal of Research in Personality, 68, 69–81. 10.1016/j.jrp.2017.02.004 [DOI] [Google Scholar]
  57. Sow F, Dijkstra K, & Janssen SMJ (2023). Developments in the functions of autobiographical memory: An advanced review. WIREs Cognitive Science, 14(3), e1625. 10.1002/wcs.1625 [DOI] [PubMed] [Google Scholar]
  58. Terracciano A, Stephan Y, Luchetti M, & Sutin AR (2018). Cognitive Impairment, Dementia, and Personality Stability Among Older Adults. Assessment, 25(3), 336–347. 10.1177/1073191117691844 [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Waterman AS (1982). Identity development from adolescence to adulthood: An extension of theory and a review of research. Developmental Psychology, 18(3), 341–358. 10.1037/0012-1649.18.3.341 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2133379-supplement-1.pdf (805KB, pdf)

Data Availability Statement

This study was preregistered on Open Science Framework (https://osf.io/8vzj2/?view_only=afad3f7ac1394bf2b79c92b7838e5116) on March 11, 2025, before data analysis, which began on March 13, 2025. Data collection occurred before preregistration to investigate a different research question. Data related to the present research question were not examined until after preregistration. Deviations to preregistration are reported in Supplementary Table 1 (we report one deviation related to sample size and one related to an added descriptive variable). All study materials, including hypotheses, data, and analysis plan/scripts, are publicly available via the preregistration link.

RESOURCES