1 Introduction Enhancing memory is among the most important tasks for both educators and students. It is important for researchers of memory, then, to examine the types of memory that can be enhanced, the degree to which those types of memory are enhanced, and the strategies that best aid in information retrieval. One strategy which shows some promise is testing, or more accurately, retrieval practice. This strategy has been studied by many researchers in the last 30 years or so (Karpicke, 2017, ). Studies have found that retrieval practice significantly enhances memory relative to restudying, and this phenomenon is called “the testing effect” or “retrieval practice effect” (RPE) (Roediger & Butler, 2010 ; Rowland, 2014 ). Research has found the RPE to be quite robust for both verbal (Karpicke, 2017 ), visuospatial (Carpenter & Pashler, 2007 ; Brinegar et al., 2016 ; Bufe & Aslan, 2018 ; Carpenter & Kelly, 2012 ), and even action (Kubik et al., 2018 ; Hotta et al., 2016 ) memory, and can persist over time (Bertilsson et al., 2021; Pashler et al., 2007 ; Kornell et al., 2011 ; Yiğit et al., 2014 ). In this paper, I will focus mainly on verbal and visuospatial memory.
The easiest way to study the testing effect is to teach participants word lists or word pairs. It should be no surprise, then, that most of the studies to date, (save for several notable exceptions), have studied the RPE for lists of words and word pairs, rather than for higher-level or more complex content (for examples, see Bertilsson et al. (2021, ) and Su et al. (2020, )). Some of the exceptions to the simple “list” studies on the testing effect, including studies by Greving and Richter (2018, ) and Jensen et al. (2020, ), have shown some testing effects, but with interesting and complicated results. The complications in these results may be, in part, due to differences in course content difficulty and methodology. For instance, Greving and Richter (2018, ) taught basic cognitive psychology class content over multiple classroom sessions, with either testing (multiple-choice or short-answer) or restudy occurring right after each session, and found a robust testing effect (only for the short-answer testing subgroup) for content retention; on the other hand, Jensen et al. (2020, ) also found a testing effect for biology students with three multiple-choice unit exams and either a closed- or an open-book final, but did not find enhanced content retention. Interestingly, Yiğit et al. (2014 ) found a several month long testing effect for classroom psychology learning, regardless of the testing format. Because of the difficulty in standardizing testing effect methodologies for studies of higher-level information, it is difficult to compare results across these studies. As a suggestion for future research, meta-analyses of this topic should carefully consider the methodologies. Therefore, much of the focus has been on methodologies which are easy to standardize and consistently replicate, such as word lists. Additional and careful research in this area is sorely needed.
There have been numerous explanations for the testing effect, though evidence supporting each of them has been inconsistent and mixed, such that only limited conclusions may be drawn from each explanation, and multiple explanations seem plausible. In the next section, I will outline and discuss the theories and hypotheses that may be relevant to the SEC theory and briefly examine evidence for and against each of them. Note that I will not discuss all the theories on the testing effect, only those with relevance to the SEC theory. Then, I will discuss the different factors that do and do not affect the RPE and also conditions that are necessary for its occurrence. Then I will develop my hypotheses about the testing effect for medium-level cognitive processing (e.g., short stories) and discuss the aspects I plan to investigate.