Establishing the Construct Validity of a Developmentally Sensitive Self-Referent Encoding Task
Abstract
The beliefs we hold about ourselves emerge early in life and play a prominent role in theories of cognitive vulnerability to depression. However, understanding how associations between self-beliefs and adjustment evolve over time requires valid and developmentally sensitive measurement tools for assessing self-beliefs in youths. The Self-Referent Encoding Task (SRET), a behavioural task assessing biases in processing self-referent information, is one of the most widely used experimental paradigms for evaluating self-views in the context of cognitive vulnerability to depression. However, few studies of its psychometric properties have been conducted. In the absence of this critical measurement work, the validity of the SRET for assessing self-views across early development remains unclear. As the first in-depth psychometric analysis of the SRET, my dissertation addresses these measurement concerns with the goal of advancing the literature on self-views and their associations with depression in youth. Specifically, I examined the factor structure and item functioning of SRET items (Study 1), the measurement invariance of SRET endorsement scores between boys and girls and across development (Study 2), and the predictive validity of SRET processing scores for depressive symptoms, as well as reciprocal associations between depressive symptoms and self-schemas (Study 3). Results from my first study support the two-factor structure of the SRET, with items loading onto positive and negative factors. In addition, I identify standardized and developmentally sensitive SRET items for assessing positive and negative youth self-concepts. Findings from my second study indicate that SRET endorsement scores show moderate to high stability during late childhood and adolescence, as well as measurement invariance over time and across sex. Lastly, findings from my third study support the predictive validity of positive and negative SRET processing scores for depressive symptoms across late childhood to mid-adolescence, but also highlight the role of depressive symptoms on negative self-schemas. Taken as a whole, my dissertation provides conditional support for the validity and utility of SRET endorsement and processing scores for studying self-views and their associations with depressive symptoms in youth. I discuss implications for the developmental dynamics of youth self-views and their associations with depressive symptoms across late childhood and adolescence.