Continuing the Conversation About Face-to-Face, Online, and Blended Learning a Meta-Analysis of Empirical Literature 2006-2017

This paper serves two purposes. First, it statistically compares learning outcomes of face-to-face, online and blended learning instruction. Then it looks at the instructional practices that are associated with effective blended and online learning. A meta-analysis of 30 studies, with 3,687 participants, resulted in 36 effect sizes. The contrasts comprised of 21 studies that looked at blended learning vs. face-to-face education, eight contrasts that looked at online learning vs. face-to-face education, and seven contrasts that looked at blended vs. online learning. The observed summary effect size was a small effect size of 0.397 with Confidence Interval (CI) of 0.242 to 0.553, with a Z-total of 5.000, and significant p <0.000. This implies that the overall effect size in treatment condition (blended or online learning) is 0.397 standard deviations higher than the control condition (face-to-face or online learning). There were eight instructional practices; self-directedness, peer interactions, feedback, multiple learning ways, student orientation, instructor presence, multiple assessments, and accountability. Instructional practices with over 90% reported in studies included: accountability (100%), multiple assessments (92%), feedback (92%), and multiple learning ways (92%). All the instructional practices reported according to Cohen (1988) revealed a small average effect size. This meta-analysis recognizes that blended learning will be replacing face-to-face education and results showed a small significant gain in learning outcomes from a previous meta-analysis (Benard, Borokhovski, Schmid, Tamim, & Abrami, 2014a, 2014b; Means, Toyama, Murphy, & Baki, 2013; Vo, Zhu, & Diep, 2017; Zhao, Lei, Yan, Lai, & Tan, 2005).

ISBN: 9780355384215