The Effects of Quality Self-Explanations on Robust Understanding of the Control of Variables Strategy

In previous studies we compared the effectiveness of three pedagogical approaches on the acquisition and robust understanding of the control of variables strategy (CVS). In our first study, Sao Pedro, Gobert, Heffernan & Beck (2009), we showed that at immediate posttest, two direct learning conditions, with and without explanation, significantly outperformed a discovery learning condition in remediating multiply confounded experiments. In our second study, Sao Pedro, Gobert & Raziuddin (2010), we retested students six months later and found that performance at this skill dropped for those in the direct-no explain condition. However, the direct+explain condition students maintained their performance. Given the efficacy of the direct+explain condition, we investigated in this work which instructional components of this condition predicted acquisition, transfer, and long-term retention of CVS. We found that the quality of self-explanations generated while critiquing setups during practice tasks was the most predictive of robust CVS understanding, above and beyond other typical predictors such as pretest score and time-on-task.