OK, here is why I don't frequently report that "this study said this" or "that study said that" - here we have a case where students in an online version of a class did more poorly than students in the in-class version. Fine, no problem. But now we have the study's author emphasizing the significant of the results. "The figures represent about a 10-percent poorer understanding of the material by the online students, Mr. Brown says. 'That is a significant difference,' he says. 'That is not a statistical artifact.'" Well - how does Mr. Brown know that his study is not the one study that is the exception when a statistical method guarantees, say, plus or minus five percent, nineteen times out of twenty? I wish that people who comment on statistics would understand that exceptions are the rule, and that such surveys should be looked at only in the aggregate, not on a case-by-case basis. Sheesh. Good job by the author, though, in placing this survey in that wider context (hey, maybe Chronicle writers are reading the occasional kind critiques in OLDaily).CRLF
Today: 5 Total: 1094 [Direct link] [Share]
