Prequestions do not enhance the benefits of retrieval in a STEM classroom

Abstract

Answering questions before a learning episode—“prequestions”—can enhance memory for that information. A number of studies have explored this effect in the laboratory; however, few studies have examined prequestions in a classroom setting. In the current study, the effects of prequestions were examined in an undergraduate course in chemical engineering. At the start of several class meetings, students were provided with a prequestion to answer about the upcoming lesson, and then were asked to provide ratings of confidence in their answers, familiarity with the content in the prequestion, and how much of the assigned reading they had completed. At the end of class, students were given the same question again (postquestion), along with a different question from the same lesson (new question). On a quiz at the end of each week, students were given the postquestions and new questions again, in addition to never-before-seen questions (quiz-only questions) from the same lessons. Performance on questions at the end of class revealed no difference in performance for postquestions vs. new questions. Although weekly quiz performance revealed an effect of retrieval practice—superior memory for material tested at the end of class (postquestions and new questions) compared to material not tested (quiz-only questions)—there was no difference in weekly quiz performance on postquestions vs. new questions. These results suggest that retrieval practice is beneficial to learning in the classroom. However, prequestions do not appear to enhance learning, nor to enhance the effects of retrieval practice.Electronic supplementary materialThe online version of this article (doi:10.1186/s41235-017-0078-z) contains supplementary material, which is available to authorized users.

Learning from errors: students’ and instructors’ practices, attitudes, and beliefs

“…Although more research on pretesting is needed to fully establish its pedagogical potential (including to address the degree to which guesses need to be somewhat informed (e.g., Kang et al, 2011), the role of associative strength between pretested cues and targets (e.g., Grimaldi & Karpicke, 2012;Knight et al, 2012;cf. Metcalfe & Huelser, 2020), as well as the finding that high confidence errors followed by feedback yields more learning (Butterfield & Metcalfe, 2001), which is also known as hypercorrection), and not all studies have shown benefits of pretesting (e.g., Geller et al, 2017), this body of research suggests that students often stand to benefit from taking pretests or attempting practice questions before new course content is presented . Such pretesting could occur before relevant readings, lectures, or discussion sections are completed, during which the correct answers could be learned.…”

“…Random samples could be used to reduce any effects of selection bias. Potential moderating influences of academic achievement level (e.g., Geller et al, 2017) and academic mindset (e.g., Rattan et al, 2015) could also be investigated. Notably, our data relied entirely on self-report measures that asked respondents to make judgments on issues and topics that were, in some cases, fairly abstract; independent verification of learning behaviours where feasible (e.g., Blasiman et al, 2017) could be used to test the accuracy and validity of those measures.…”

“…Relatedly, further research on the efficacy of pretesting and productive failure is needed before either technique can be endorsed for widespread use. Such research might occur in authentic educational environments (e.g., Geller et al, 2017), and for the case of pretesting, address the specificity of learning that has repeatedly been observed in some experiments (e.g., James & Storm, 2019;cf. Pan, Lovelett, et al, 2019), the role of surprise (Butterfield & Metcalfe, 2001), the finding that generating errors that are semantically related to the correct answers yields more potent learning (Cyr & Anderson, 2018;Zawadzka & Hanczakowski, 2019), various types of pretest and criterial test questions (e.g., St.…”

Selective effects of errorful generation on recognition memory: the role of motivation and surprise

“…Some experiments have found that pretesting improves subsequent retrieval of the pretested material, but does not benefit other, non-pretested information (Carpenter, Rahman, & Perkins, 2018;Pressley, Tanenbaum, McDaniel, & Wood, 1990;Richland et al, 2009;Toftness, Carpenter, Lauber, & Mickes, 2018). Other studies, by contrast, have found that pretesting improves memory for both pretested and nonpretested information (Carpenter & Toftness, 2017), or that pretesting does not significantly benefit retrieval at all (Geller et al, 2017). Indeed, some experiments have even found that testing can impair the learning of new information when it is presented alongside the tested material (Davis & Chan, 2015;Finn & Roediger, 2013).…”

Pan, Sana, Schmitt, and Bjork (2020) Pretesting reduces mind wandering during online lectures

“…For instance, Carpenter and Toftness (2017;see also Bjork et al, 2013) noted that pretesting might serve as a metacognitive "reality check"-that is, highlighting the gaps in one's knowledge and facilitating a search for the relevant information during subsequent study. Another possibility is that pretesting may stimulate curiosity, which in turn improves attention for the information that follows (Geller et al, 2017; see also Metcalfe & Finn, 2011). Although these accounts focus on how pretesting benefits memory, they also imply that pretests can substantially influence attention during learning-and more broadly, raise the possibility that attentional changes may contribute to the pretesting effect itself.…”