Cloudworks is no longer accepting new user registrations, and will be closing down on 24th June 2019. We hope to make a read-only archive of the site available soon after.
Psychometric Analyses of the Expectancy-Value-Cost Scale in Advanced Nanotechnology MOOCs
Cloud created by:
15 May 2017
Learner motivation is a primary driver of behavior in MOOCs or any other educational settings. In our previously studied MOOCs, nearly all learners have exhibited strong intrinsic motivation, as they generally were specifically interested in learning the course content itself, independent of external benefits. While demographics are frequently collected to better understand learners, an alternative measure of motivation is necessary to adequately discern between individual learners. One potentially promising measure of motivation for MOOC learners is the Expectancy-Value-Cost (EVC) Scale, which is intended to measure learners’ beliefs about their abilities to be successful in a course, the value of that course, and the costs of achieving success in that course. Previous applications of the EVC instrument have been conducted with children and adolescents in face-to-face learning contexts, but these constructs may be particularly relevant for MOOC learners with more pressing needs and constraints. We examined the psychometric properties of the EVC scale in the older, more diverse population of our advanced technological MOOCs (n = 407). Specifically, we studied learners in two MOOCs offered through the edX platform: Principles of Electronic Biosensors and Nanophotonic Modeling. This research applied several psychometric analyses, including exploratory and confirmatory factor analyses, invariance analysis, and item response theory, using learner responses to EVC items on the pre-course surveys. Exploratory and confirmatory factor analysis results indicate that the factor structure found in previously studied populations for the EVC instrument is maintained for our MOOC learners. At the same time, measurement invariance analyses suggest that the interpretation of each factor may vary slightly across different subpopulations. Further, the item response theory analyses suggest that some items provide stronger information about the EVC scale for MOOC learners than other items. Our results are expected to guide modifications to this scale in order to improve weak or inconsistent item functioning for use in future MOOCs.