By Caitelin McCoy, Kenyon ’17
Christian Kandler, Anna E. Kornadt, Birk Hagemeyer, and Franz J. Neyer completed a longitudinal twin study about personality development in old age. On December 29, 2014 the Journal of Personality and Social Psychology published this report for the public.
“Personality is commonly defined as the sum or product of all characteristics that describe individual differences in thoughts, feelings, strivings, and behaviors that are relatively enduring across situations and over time within a certain reference population, such as age or culture” (Kandler, Zimmermann, & McAdams, 2014). This study focused on the Big Five Taxonomy (neuroticism v. emotional stability, extraversion v. introversion, openness to experience, agreeableness, and conscientiousness) which captures traits as they are portrayed in any possible cultural setting.
Two related concepts in this study are construct personality focus on specific cognitive and affective features of psychological individuality. Two constructs in this study are locus of control or perceived control (the extent to which individuals believe that they can control events they experience). Affect intensity is the typical strength of individuals’ emotional responsiveness in their everyday life. Perceived control reflects a continuum from externality (life and personality is determined on the environment) to internality (life and personality is under one’s own control). The two constructs, perceived control and affect intensity, are distinct from one another, but also show links to the Big Five traits. (Perceived control is negatively linked to neuroticism, and positively linked to extraversion and conscientiousness while affect intensity positively associated with neuroticism, extraversion, and agreeableness).
Kandler et al. reviewed empirical findings on normative age-trends in personality traits and individual differences over time. Age trends reflect average change on a personality traits over time, but individuals differ in their level of change. This may affect their rank-order and/or lead to an increase or decrease of individual differences over time, without changing the individuals’ rank-order.
Roberts and colleagues (2006) found that the average level of neuroticism decreased in young adult hood when compared to older adults, and agreeableness and conscientiousness increased with age. This pattern of mean-level trends showing decreases in neuroticism and increases in agreeableness and conscientiousness has been labeled social maturation because the skills necessary for social integration and mastering tasks are being increased. In a study of adults aged 69-72 by Mottus, Johnson, and Deary (2012) they found almost no significant mean-level changes for Big Five traits, but reported that an older person would show a decrease in extraversion, intellect (openness), agreeableness, and conscientiousness, while, in older women, there was an increase in neuroticism.
The current longitudinal twin study investigated the patterns and sources of continuity and change in personality traits in old age ranging from 64-89 years old over two measurement occasions five years apart. They studied the relations between personality development and change in psychological well-being.
There were five specific aims in this study. First they examined mean-level trends in personality and well-being. Reverse trends from older compared to younger aged persons was expected. The first hypothesis (Hypothesis 1a) was an increase in neuroticism and decreases in extraversion, openness, agreeableness, conscientiousness, and perceived control. Ideally the valence of emotionality should change and the average level of emotional reactivity and well-being should stay the same. Hypothesis 1b: because of a balance between losses of emotion regulation ability and enhanced expertise in management strategies of affect regulation in older adults. Second, the refined maturity principle for older ages suggests that increases in neuroticism and decreases in the other traits act to compensate for loss of perceived control. It was expected that perceived control prospectively predicts Big Five personality traits and that change in these traits are associated with change in perceived control. Thirdly, the degree of rank-order continuity and individual differences in change in personality traits and well-being was investigated. Hypothesis 3 states, in line with the plasticity principle or personality development that individual ranks in personality traits changes beyond random fluctuations due to measurement of error. Fourth they disentangled genetic from environmental components of rank-order continuity and change in personality traits and well-being. Hypothesis 4 states that rank-order continuity in all personality traits should be due to genetic and environmental factors. It is in conjunction with this hypothesis that twins were used in the study; to ensure genetic continuity. Finally, they investigated the interrelation between personality traits and psychological well-being in old age. It was expected that individual differences in well-being are associated with individual differences in levels and changes in neuroticism, agreeableness, conscientiousness, and perceived control (Hypothesis 5).
The data in this study were drawn from the Genetic Oriented Life Span Study on Differential Development. (L-Data) The twin study was initiated in 1937 with 90 twin pairs and in the follow-up study between 1994 and 1999 with a focus on later adulthood, only 20 complete pairs could be tested. The sample was extended to include new pairs ranging from 64-85 years old. The upper age group was included to reduce the probability of dropout due to mortality. After 5 years, the participants were retested. At time 1 there were 410 participants, but at time 2 there were 324. The dropout was larger for older people and women, missing at random tests which indicated that the dropout rate could be assumed to be randomly distributed.
Personality traits were measured on the 60-item German Neuroticism-Extraversion-Openness Five-factor inventory and was used to capture Big Five trait scores. Perceived control was measured using a 32 item questionnaire on general competence and control beliefs. Affect Intensity was measured by a 40-item German version of the Affect Intensity Measure. Well-being was measured as satisfaction with momentary states and current living conditions. Negative states and moods were negatively coded.
To estimate mean-level trends and individual differences in trajectories for all of the variables, they ran latent change model analyses. Developmental trajectories for the variables were estimated using the level variable to represent individual differences in the variable that remain stable across assessment waves, whereas the change variable reflects the individual differences in linear change over time. To interpret the latent change as true change each of the variables needed to be correlated.
Rank-order continuity for latent true scores was more than 80% except for affect intensity. Individual differences in personality traits are for the most part stable over time in old age which points to individual differences in change in this period of life. Age was significantly correlated with the level of extraversion, conscientiousness, and perceived control. These correlations indicate that older people perceive lower levels of control and tended to be less extraverted and conscientious. These associations suggest that women score higher in neuroticism and agreeableness, where men score higher in extraversion, conscientiousness, perceived control and well-being. They did not find significant prospective effects from Big Five traits on perceived control except for perceived control on conscientiousness indicating higher levels of perceived control at time 1 and lower levels at time 2. The analysis yielded significant negative change correlations between neuroticism and perceived control and increasing neuroticism was associated with decreasing perceived control. Well-being did not show significant prospective effects on personality traits, the effects of initial personality traits on well-being was largely paralleled with the initial correlations indicating that personality levels affect well-being, but not vice versa. Individuals who showed stronger decreases in neuroticism and increases in perceived control and conscientiousness had a higher probability of ranking high in well-being.
Consistent with Hypothesis 1a, and in line with the findings the results suggest that mean-level change in personality traits in old age drifts in the opposite direction as for younger adults and one explanation for this is that older people become aware of losses in cognitive and physical functionality and have to be more wary of possibly health risks and dangers of everyday life, while being more selective of activities and investments in social relationships. All of this reduces perceived control, increases neuroticism, decreases extraversion and agreeableness, reduces openness, and decreases conscientiousness.
The study provides strong support for different mean-level trends in old age in opposition to younger and middle aged adults that are due to different processes of adaptation in later adulthood. Older individuals significantly differed in their developmental trajectories in all personality traits and well-being because latent modeling analyses, individual differences in change were corrected for random fluctuation between measurement occasions due to error of measurement.
Even though individual differences in personality traits are largely stable over time, personality can change and individuals differ in change throughout lifespan. Genetic as well as environmental sources affect continuity in individual differences and variation on intraindividual change which indicates that multiple sources driver personality development and plasticity in old age.
Kandler, C., Kornadt, A.E., Hagemeyer, B., &Neyer, F.J. (2014, December 29). Patterns and Sources of Personality Development in Old Age. Journal or Personality and Social Psychology.
Mottus, R., Johnson, W., &Dreary, I.J. (2012) Personality traits in old age: Measurement and rank-order stability and some mean-level change. Psychology and Aging, 27, 243-249
Roberts, B.W., &Wood, D. (2006) Personality development in the context of neo-socioanalytic model of personality. In D.K Mroczek, T.D. Little, D.K. Mroczek, & T.D. Little (Eds.) Handbook of personality development (pp. 11-39). Mahwah, NJ: Erlbaum Publishers.