The structure of temperament in college -aged and middle -aged adults and a test of the goodness -of-fit model

Susan Ruth Waddington, Fordham University

Abstract

The purpose of this investigation was to extend the study of temperament beyond young adulthood to middle-aged adults. To advance measurement development efforts, the factor structure of the Revised Dimensions of Temperament Survey (DOTS-R) was investigated in 397 college-aged and 297 middle-aged adults. The dimensions derived from the factor analysis for the middle-aged adults were used to assess predictions based on the goodness-of-fit model of person-context relations and the contrasting acontextual personological model. Scarr and McCartney's (1983) developmental theory of environmental effects proposing a correspondence of the individual and context as development proceeds was incorporated into an elaboration of a model of fit within the social context of the marital dyad. Spouse correspondence of temperament was assessed by calculating a similarity index over the set of temperament dimensions measured by the DOTS-R. Similarity indices were compared for 82 actual and randomly created dyads to assess whether the extent of spouse correspondence of temperament was greater than that expected by chance, a prediction derived from Scarr and McCartney's theory. For actual couples, to test predictions based on the goodness-of-fit model, similarity indices were correlated with the psychosocial outcome measures of self-esteem, anxiety, depression, and marital adjustment. Multiple correlations between individuals' temperament dimensions and psychosocial outcome measures were also computed to test predictions based on the personological model.^ Results found that for both the college-aged and middle-aged samples the factor structure of the DOTS-R was best represented by a 9-factor model. Similarity indices computed for actual and random couples were not significantly different as hypothesized. For actual couples, spouse similarity of temperament was significantly correlated with one outcome measure, self-esteem, and individuals' temperament dimensions were correlated with all four outcome measures, self-esteem, anxiety, depression, and marital adjustment. These findings provided stronger support for the personological compared to the goodness-of-fit model. It was suggested that measurement development efforts continue along with efforts to specify and test predictions based on a goodness-of-fit model appropriate to adulthood. ^

Subject Area

Psychology, Developmental|Psychology, Personality|Sociology, Individual and Family Studies

Recommended Citation

Susan Ruth Waddington, "The structure of temperament in college -aged and middle -aged adults and a test of the goodness -of-fit model" (January 1, 1996). ETD Collection for Fordham University. Paper AAI9613863.
http://fordham.bepress.com/dissertations/AAI9613863

Share

COinS