Environmental Facilitators and Barriers to Student Persistence in Online Courses: Reliability and Validity of New Scales

Abstract This study aimed at building a reliable and valid scale for environmental factors related to student persistence in online courses, particularly relevant for adults or lifelong learners. Drawing on the social integration and external attribution scales and subscales of Kember et al. as a starting point, data collected in Canadian universities were randomly split into two samples. The first sample (n 1 = 385) was used to explore the data set through principal component and reliability analyses. These confirmed a two-factor environmental scale composed of encouragements (factor 1) and time-events items (factor 2), as well as a two-factor persistence scale that included potential dropout (factor 3) and cost-benefit items (factor 4). All factors showed a very good internal consistency. The second sample (n 2 = 381) was used to confirm the structural validity of environmental and persistence scales through confirmatory factor analyses and to compare this new structure to the subscales of Kember et al. While the latter resulted in an insufficient model fit, the new environmental and persistence scales yielded a very good model fit with strong goodness-of-fit indices and statistics. These results confirmed the structural validity of the new scales, which can be trusted for use in further empirical studies related to online student persistence. The new scales can also be used by practitioners to detect at-risk students early in a semester, allowing to offer them specific individual support to foster student persistence in online courses.