Measuring the learning effects of inquiry-based teaching with computer simulations

Abstract

1 Purpose Ample scientific research has been focused on the learning effects of computer simulations in science education. However, most research has ignored the pedagogical context in doing so (authors, 2012). The purpose of this study is to investigate different approaches to teaching with computer simulations, and in doing so, taking the pedagogical context into account. 2 Theoretical framework The difference between the two simulation-supported teaching approaches that we investigated, is that in one of the conditions the teacher taught in a way that was prepared by the teacher him-/herself, and in the other condition the teacher taught with the same computer simulations following a Peer Instruction approach (Crouch, Watkins, Fagen, & Mazur, 2007). 3 Method This experiment has been performed in the classes taught by three teachers. Each teacher taught physics to parallel classes, whereby each class was taught in a different way with computer simulations: in one of the classes Newtonian mechanics was taught in way that was prepared by the teacher, and in the other class a Peer Instruction approach was followed, supported by an electronic voting system. In both conditions teachers used the same computer simulations. The lesson series of both conditions were designed as follows: in the first lesson the students conducted a pretest, in the second and third lesson the experiment was conducted, in the fourth lesson the students performed a posttest, and in the fifth lesson –one month later– a delayed posttest was conducted. At the pretest, the posttest, and the delayed posttest data was collected by having the students fill in the Force Concept Inventory, a questionnaire measuring conceptual insight in Newtonian mechanics (Hestenes, Wells, & Swackhamer, 1992). Learning effects were investigated by performing paired-samples t-tests. 4 Results Paired-samples t-tests show that when teachers prepared the lessons themselves, scores on the FCI do not increase between the pretest (M = 12.24, SD = 4.01) and the posttest (M = 12.88, SD = 4.00): t(41) = -1.65, p = .11; or the pretest (M = 12.23, SD = 4.03) and the delayed posttest (M = 13.20, SD = 3.74): t(39) = -2.01, p = .05. However, in case teaching with computer simulations is supported by our Peer Instruction approach, FCI-scores increase between the pretest (M = 11.96, SD = 4.30) and the posttest (M = 13.93, SD = 4.49): t(56) = -4.28, p < .01; as well as between the pretest (M = 11.85, SD = 4.35) and the delayed posttest (M = 14.46, SD = 4.57): t(53) = -5.35, p < .01.