Most biologists argue that intelligent design (ID) theory, which has recently been judged a pretext for advancing religious belief (see www.pamd.uscourts.gov/kitzmiller/kitzmiller_342.pdf), has no place in the classroom. Yet Steven D. Verhey proposes in “The Effect of Engaging Prior Learning on Student Attitudes toward Creationism and Evolution” (BioScience 55: 996–1003) that allowing ID to be critically discussed in the college classroom may facilitate appreciation of evolution. Verhey tested this proposal by contrasting sections of introductory biology that included or did not include creationist information. He reports that adding creationist materials to the curriculum did enhance attitude change. In an accompanying editorial, “How Can We Help Students Really Understand Evolution?”(BioScience 55: 923), Craig E. Nelson lauds these results as “powerful ev-idence”that introducing ID into the classroom led to “extensive change toward more scientifically viable views.”
I applaud Verhey's attempt to provide empirical evidence in support of this novel approach. Unfortunately, his study does not provide a sound basis for Nelson's enthusiastic endorsement. The design of the study is flawed, as the two groups differed not only in whether ID materials were provided, but also in which text chapters and supplementary proevolution texts were assigned. Moreover, students were asked to state their initial attitudes toward creationism only at the end of term, when their recollections may have been influenced by the experience of taking part in the study.
The most serious problem is with data analysis. While Verhey's primary finding is that attitude change was greater for the group exposed to creationist information, his measure inexplicably includes change away from evolution. Thus this measure cannot be used to determine the effectiveness of this approach in instilling proevolution change, the goal of this study. Verhey did provide data that allow this critical issue to be examined. He reported that 9 students out of 38 in the creationism-plus-evolution group shifted their attitudes toward evolution, compared with 5 students out of 28 in the evolution-alone group. However, I have been informed that the 5-students datum is in error, and the correct number is 2. Using the correct value, the groups do not differ significantly by Fisher's exact test (p = 0.10, two-tailed). A two-tailed test is appropriate here, given that a plausible alternative outcome is that the teaching of ID may instead shift attitudes toward creationism. Nevertheless, even with a one-tailed test, the results remain nonsignificant (p = 0.07). Moreover, 3 students in the creationism-plus-evolution group shifted toward creationism; none in the evolution-alone group did. Consequently, Verhey's report does not provide evidence supporting the introduction of ID into the college classroom as a pedagogical technique to promote acceptance of evolution.
