Author

Melinda Maher

Date of Graduation

2000

Document Type

Dissertation/Thesis

Abstract

Although temperament is generally defined as an individual’s stable behavioral style, comparisons of temperament ratings derived from different sources typically yield only low to moderate correlations. The present study investigated three potential influences on toddler temperament ratings: the context of the observation, the raters’ amount of experience with the child, and rater characteristics. Thirty toddlers (14 to 36 months old) were included in the study. For each toddler, temperament ratings were provided by his/her mother, father, two day care teachers, a home observer, and a day care observer. Toddler temperament was assessed using the TTS, EAS, and a global temperament measure. In addition, all raters answered self-report measures of depression, anxiety, temperament, and mood. First, context was found to be related to toddler temperament ratings. More significant correlations (p < .01) were found between pairs of raters who observed the toddlers in the same context (mother-father, teacher-teacher, parent-home observer, and teacher-day care observer) in comparison to pairs of raters who observed the toddlers in different contexts (parent-teacher, parent-day care observer, and teacher-home observer). In addition, t-tests indicated several significant differences (p < .01) in temperament ratings derived from the two contexts. These findings indicated that some temperament dimensions may display variability in different environments. Secondly, the raters’ amount of experience with the toddlers was found to be related to temperament ratings. More significant correlations (p < .01) were found between pairs of raters with similarly high amounts of experience with the toddlers (mother-father and teacher-teacher) in comparison to pairs of raters with disparate amounts of experience (parent-home observer and teacher-day care observer). In addition, observers rated the toddlers more positively than the parents or teachers. These findings indicated that observers’ limited experience with the toddlers may lead to incomplete or unreliable ratings of temperament. Finally, a series of standard multiple regressions indicated that raters’ self-reported characteristics were significantly related to some toddler temperament dimensions (p < .05). Additionally, mothers’ and teachers’ characteristics had a more significant influence on their toddler temperament ratings in comparison to fathers’ and observers’ characteristics. However, rater characteristics were not found to significantly alter agreement levels between raters.

Share

COinS