After reading several articles today I see that when I was studying linear algebra and calculus, I really should have taken a course in statistics. Wei, Peng, and Chou who co-authored “Can more interactivity improve learning achievement in an online course? Effects of college students’ perception and actual use of a course-management system on their learning achievement” are well cited scholars, however, I do question some of their methods.
In the abstract it states “This study aims to investigate how interactivity influence learners’ use of interactive functions in the course-management system (CMS) and their online learning performance.” Anecdotal information was gathered in the form of a questionnaire at the final exam. The researchers then applied statistics to the information. Although it is possible to do this I question how it makes the anecdotal information more accessible or meaningful. W. Trochim has some interesting information on conclusion validity in his book, Research Methods Knowledge Base.
In the Implications and recommendations section Wei, Peng, and Chou say “students’ online discussion scores were directly influenced by students’ self-reported frequency of logging into the e3-system”. This implies that self-reporting a login has a direct effect on the discussion scores. Perhaps that should read correlation but again I don’t have that stats course. (yet) I think that the direct influence comes from the actual logging into the CMS.
A more interesting question might have been to compare the correlation of the actual interactions of the student with the CMS to their online-learning performance, then compare the correlation of the self-reported interactions of the student with the CMS to their online-learning performance and then contrast those two results. That would be another study.