Recently the process-product paradigm for research on teaching has undergone revision. Notable among updates to this paradigm are what Doyle (1980) calls the "mediating process paradigm," or what Winne (1981) refers to as the "cognitive mediational paradigm." Both views postulate that teachers do not directly influence student product variables, such as achievement. Rather, teachers influence students by causing them to think and behave in particular ways during teaching. These mediating events, in turn, may lead to changes in outcome variables. Hence, the effects of teaching on learning may be mediated by students' behaviors and cognitive processing during instruction. The cognitive mediational paradigm opens up to question the match between ways researchers hypothesize that learners