Self-Reported Perceptions of Online Learning and Observed Online Learning Behaviors: To What Extent Are They Consistent?

Feifei Han

doi:10.18357/otessac.2021.1.1.48

Abstract

This study investigates to what extent there is an association between students’ self-reported perceptions of online learning and observed online learning behaviors recorded by the learning analytic data. The participants were 319 undergraduates studying an engineering course in an Australian university. Data analyses were conducted using cluster analyses, Hidden Markov Model, one-way ANOVAs, and a cross-tabulation. The relations between students’ self-reported perceptions and their academic learning outcome show that those with positive perceptions tended to have higher scores. The relations between observational online learning behaviors and their academic learning outcome demonstrate that students with most learning sessions achieved more highly. The cross-tabulation finds a significant association between the cluster membership generated by by the self-reported perceptions and observational online learning behaviors. Amongst students who had most study sessions characterized by high percentages of reading and formative states and low percentage of summative states, the proportion of those with positive perceptions (40.2%) was significantly higher than those with negative perceptions (20.0%). Of students who had the least study sessions represented by moderate reading and summative states, and low formative states, the proportion of students with positive perceptions (3.0%) was significantly lower than the proportion of students having negative perceptions (8.7%).

Highlights

Learning analytics in higher education is used for a number of purposes, such as issues of attrition (Dawson, Jovanovic, Gašević, & Pardo, 2017), social presence (Joksimović, Gašević, Kovanović, Riecke, & Hatala, 2015), learning design (Tempelaar, Rienties, & Giesbers, 2015), and education policy (Ferguson et al, 2016)
While some learning analytic studies adopt bottom-up approaches, which are predominantly guided by empirical evidence separated from educational theories; others argue for top-down approaches, which sound theoretically orientated frameworks to guide the analyses of online analytic data (Gašević, Dawson, & Siemens, 2015)
There is a dearth of research, which combines measures of analytics and the intent and experience of the students underpinning them (Han & Ellis, 2017)

Summary

Introduction

Learning analytics in higher education is used for a number of purposes, such as issues of attrition (Dawson, Jovanovic, Gašević, & Pardo, 2017), social presence (Joksimović, Gašević, Kovanović, Riecke, & Hatala, 2015), learning design (Tempelaar, Rienties, & Giesbers, 2015), and education policy (Ferguson et al, 2016). To improve our understanding of student online learning experiences, it is important to investigate how the two approaches are associated with each other, and the extent to which the results derived from the two approaches are triangulated. This study uses top-down (i.e., students’ self-reported perceptions of their online learning experience) and bottom-up approaches (i.e., the patterns of students’ online learning behaviors) separately to investigate the relations between either students’ self-reported and observational online learning experiences and their academic learning outcomes. It examines the association of the patterns of online learning generated from the two. Three research questions guided the present study: 1. What are the relations between students’ self-reported perceptions of online learning environment and their academic learning outcomes?

Methods

Results

Conclusion