A published review of 117 empirical studies assessing the impact of critical thinking instruction on critical thinking (CT) skills and dispositions confirms my suspicions: some instruction has a powerful positive effect and some instruction has no effect at all. Overall, these results support the view that, on balance, there is a measurable positive effect. And there are tentative indications of what kind of instruction fosters the most marked improvements in critical thinking.

In their paper “Instructional Interventions Affecting Critical Thinking Skills and Dispositions: A Stage 1 Meta-Analysis,” (Review of Educational Research 78:4, December 2008), Philip Abrami and six Canadian co-authors attempt to derive measurements of the effect of pedagogical “interventions” intended to improve critical thinking.

In their meta-analysis, Abrami et al. used Ennis’ taxonomy to divide the studies into four categories by intervention approach: “general,” “infusion,” “immersion” and “mixed,” characterized as follows:

  • General – CT is taught as a separate subject, with no specific subject-matter focus
  • Infusion – CT is taught by means of adopting a critical perspective on the content of a different subject; CT principles are explicitly articulated and deeply contextualized in parallel with the learning of the substantive content of the subject
  • Immersion – like infusion, but in which CT principles are not explicitly articulated
  • Mixed – a combination of general and either infusion or immersion

Despite large variances in the effects of the interventions that made comparisons of their statistical means dubious in several respects, the study supports at least three significant inferences from the data.

One conclusion is that when instructors receive special training on preparation for teaching CT skills, the effect of their interventions were the greatest. Better outcomes “can be achieved through active, purposeful training and teacher support.”

Another conclusion is that collaborative learning activities, if present, do contribute to more positive outcomes, but this effect is quite weak.

Last, but not least, the results of the Abrami study are relevant to the ongoing debate about the transferability of critical thinking education between the generalists on the one hand, and the specifists, on the other. One conclusion implied by the study was that, among the four types of intervention, a “mixed” type produces a more statistically significant effect than other approaches! In other words, “developing CT skills separately and then applying them to course content explicitly works best.”  Based on this result, the generalists and specifists could each proclaim a “victory” for their respective sides of the debate.

The generalist view is that critical thinking instruction need not be delivered in the context of subject-matter instruction to be effective. For example, generalists assert that fallacies of reasoning can be communicated in a way that presumes only common knowledge, and that a grasp of the differences between sound and fallacious reasoning, once acquired, can be applied in a variety of contexts.

The specifist view is that since forms of reasoning and argumentation are subject-matter specific, logical competence in one field does not reliably transfer to other fields. Critical thinking, if it is taught at all, must be taught by means of the adoption of a critical perspective within the context of subject matter-oriented instruction.

Based on Abrami et al., the generalists can affirm that adding general critical thinking training to a subject-matter focused curriculum can improve results. Similarly, the specifists can maintain that general critical thinking training is only really productive if the opportunity exists to apply those skills within the context of a subject-matter rich learning experience, where the real learning takes place.

Despite the preliminary nature of these results, they are pretty convincing that we now need constructive investigation rather than more critique. The generalists and specifists should no longer be focused on the whether the other side’s preferred style of intervention works, but how and under what circumstances their own preferred style can be rendered most effective.

(H/T to John Trent on LinkedIn for the reference.)