Engaging in Post-Refutation Retrieval Practice Increases the Durability of Conceptual Change

• In Event: Reducing the Impact of Misconceptions

Abstract

The purpose of this study was to investigate the long-term efficacy of the refutation method for promoting conceptual change and the use of retrieval practice as a post-refutation activity to increase its durability. The framework used to develop the study and interpret its findings was drawn from theories about conceptual change (Chi, 2008; Kendeou & O’Brien, 2014). Although the efficacy of the refutation method is well supported (Guzzetti et al. 1993; Tippett, 2010), there is evidence that people revert back to their misconceptions over time (Butler, Fazio, & Marsh, 2011). The present study investigated the potential efficacy of combining the refutation method with retrieval practice. Practicing the retrieval of information from memory improves long-term retention and transfer of learning to new contexts (Butler, 2010; Roediger & Butler, 2011). In addition, recent research shows that transfer can be improved through introducing variability into the process of retrieval practice by requiring the application of knowledge to different practice problems (Butler, Black-Maier, Raley, & Marsh, 2017). 
The study was conducted over two sequential semesters in an undergraduate online introductory psychology course (n = 273). At the start of each semester, students completed a belief scale that included 18 myths about psychology. During the semester, they viewed brief videos in class (one per class session), each of which targeted a single myth using the refutation method. The main experimental manipulation consisted of the learning activity that followed the videos: repeated retrieval practice with the same question, repeated retrieval practice with different questions, or no further exposure (i.e. refutation only). Learning activity was manipulated within-students, but between-materials. For myths assigned to the retrieval practice conditions, students engaged in spaced retrieval practice via three multiple-choice application questions – one given immediately after the video and the others given after the videos in the next two classes. After the Spring semester, students from the Fall and Spring courses again filled out the belief scale and then answered a set of application questions about the correct information that refutes the myths. The timing of this post-course survey represented the other experimental manipulation – either a six-month or one-month retention interval.
The average pre-course belief rating for the myths was 4.3 on a seven-point scale (1=totally false; 7=totally true) and ratings did not differ by condition (all Fs<1). Students held misconceptions (rating of 5+) for 50% of the items. On myths for which students initially held misconceptions, both types of retrieval practice (variable: -2.5; same: -2.4) led to a significantly greater decrease in belief ratings relative to receiving the refutation only (-1.9)[F(2,526)=12.326, p<.001]. The change in ratings was also significantly greater after one month (-2.5) than six months (-2.1). Both types of retrieval practice (variable: 73%; same: 70%) led to a significantly greater performance on the post-course application questions relative to receiving the refutation only (63%)[F(2,484)=8.381, p<.001], but there was no difference between retention intervals. Overall, these findings show that the refutation method produces durable benefits, but engaging in retrieval practice after the refutation can enhance the efficacy of the intervention.

Authors

• Andrew C. Butler, Washington University in St. Louis

• Cynthia Alarcon, The University of Texas - Austin