LONGITIDUNAL ASSESSMENT DESIGNS FOR NON-COGNITIVE CONSTRUCTS: APPLICATIONS USING TIME-SERIES DATA FROM AN EMOTION SCALE FIELD STUDY

This paper proposes a longitudinal assessment approach for the assessment of non-cognitive constructs. Focusing on how to evaluate “state” versus “trait” properties of propensities, a comparative narrative is provided attesting to the utility of employing a longitudinal perspective. Illustrations on how to collect, analyze and interpret time-series data are provided using three separate studies, conducted as a part of a larger longitudinal study, referred to as the Emotion Ruler Field Study, investigating emotional aspects of learning for college students. The first study illustrates how inferences made about stress and tenseness levels of college students may differ when based on self-assessments taken over twelve days versus on those taken only once. The second and the third studies illustrate how inferences made about sleep quality and psychological resilience of college students may evolve when the variances observed over several weeks (within-persons) are taken into account in the analysis of individual differences (between-persons)