Measuring mathematics interest and affect: An item response theory evaluation of the Self Description Questionnaire I (SDQI)
Year: 2014
Type: dissertation
Abstract: In response to today’s high-tech global economy, there has been an increasing need for professionals in the Science, Technology, Engineering, and Mathematics (STEM) fields. To enhance students’ STEM performance and career aspirations, researchers have suggested capturing their mathematics interest as early as elementary school. Research on academic interest, however, has suffered from various conceptual, theoretical, and measurement issues. In the literature, there are mixed findings regarding gender, ethnic, and age differences in mathematics interest which has entailed reexamination of the existing measures of interest. The purpose of this dissertation study was to evaluate the psychometric properties of the Self Description Questionnaire I (SDQI) with regard to children’s mathematics interest based on the item response theory (IRT) framework. \nThe data for this study were drawn from the Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K) with a sample of 14,631 children. These children were assessed in third- and fifth-grade using the ECLS-K adapted SDQI mathematics subscale, which consists of eight items with ordered response options ranging from 1 to 4. The IRT-based evaluation suggests that the SDQI items have sufficient discrimination, but insufficient item thresholds in assessing third- and fifth-graders’ mathematics interest. Specifically, the IRT parameter estimates indicate that 4-point scale of the SDQI items does not function much better than a dichotomous scale, particularly among third-graders. Thus, it is recommended that the current SDQI be reconsidered for its age appropriateness. In addition, affective items, which involve enjoyment, liking, and positive emotions, are found to be better indicators of interest than cognitive items that involve perceived competence, ability beliefs, or self-efficacy. \nIn terms of measurement bias across gender, ethnic, and age groups, item-level bias emerged in most of the items as assessed by the IRT-based Wald Test of differential item functioning. In general, it was found that boys and White children tend to endorse higher-order options of cognitive, perceived competence items to represent their mathematics interest, while girls and ethnic minority children tend to endorse higher-order options of affective items. This affective-cognitive distinction of measurement bias indicates that children may interpret and respond to self-report interest measures differently as a function of their demographic background, resulting in artificial item-level group differences as detected by the SDQI, although the scale score can still be trusted in such group comparisons. Finally, results of the item parameter drift analysis indicate that children become more sensitive to the response anchors as they grow older, which reaffirms that the current SDQI items with the 4-point scale performs poorly among young children. Taken together, findings of this dissertation study provide insights into the conceptualization and operationalization of mathematics interest and affect, as well as practical implications for future use, and development of self-report measures in this field.
Author: Tianlan Wei
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
SDG: Quality education
Open Access status: green
OpenAlex ID: W2787962898