Date of Award
Doctor of Philosophy
Jason C. Chan
Testing can improve retention of tested information (see Roediger & Karpicke, 2006a for review), but it can also impair memory for nontested, related information: an effect termed retrieval-induced forgetting (Anderson, Bjork, & Bjork, 1994). To my knowledge, retrieval-induced forgetting has only been shown in experiments where participants study all exemplars and then perform one retrieval practice phase (see Anderson, 2003; Raaijmakers & Jakab, 2013; Storm & Levy, 2012 for reviews). Researchers have demonstrated that interpolating memory tests during the learning phase can reduce the amount of proactive interference one experiences when one learns new information (Szpunar, McDermott, & Roediger, 2008). In this dissertation, I examined how studying information in multiple blocks rather than one block influenced the buildup of proactive interference. Previous researchers (Szpunar et al., 2008) examined how testing can reduce the buildup of practice interference when all previously studied information was tested. I examined whether testing only a subset of the exemplars in a block would inoculate participants against the buildup of proactive interference. Furthermore, I examined how the presentation order of the Rp- exemplars (i.e., nontested exemplars from the tested categories) relative to the Rp+ (tested) exemplars influenced the magnitude of retrieval-induced forgetting. Retrieval-induced forgetting researchers generally endorse one of two accounts: the inhibition (Anderson, 2003) or blocking account (Raaijmakers & Jakab, 2013). I predicted that if blocking drives retrieval-induced forgetting, I would find retrieval-induced forgetting regardless of presentation order of the Rp- and Rp+ exemplars. However, if inhibition drives retrieval-induced forgetting, I would only find retrieval-induced forgetting in conditions where Rp- exemplars were presented prior to retrieval practice. In three experiments, to-be-learned information was presented either in one block followed by retrieval practice (the cumulative retrieval practice condition), or presented in four blocks. For the latter participants, some were given math instead of retrieval practice between each block (the interim math condition), some participants studied the Rp- exemplars in blocks prior to the presentation and retrieval practice of the Rp+ exemplars (the high competition condition), and the remaining participants studied and received retrieval practice over the Rp+ exemplars before learning the Rp- exemplars (the low competition condition). Then participants either had a 20-minute (Experiments 1 and 3) or 10-minute distractor period (Experiment 2). Finally, participants either were given a category cued recall test (Experiment 1) or a category-plus-stem cued recall test (Experiments 2 and 3). In Experiment 1, participants demonstrated retrieval-induced forgetting in the cumulative retrieval practice, high competition, and low competition conditions. However, in Experiment 2, participants only demonstrated retrieval-induced forgetting in the cumulative retrieval practice and high competition conditions. The results of Experiment 3 were inconsistent with Experiments 1 and 2 and may have been due to chance. The current dissertation provides evidence for both the inhibition and blocking accounts.
Matthew R. Erdman
Erdman, Matthew R., "Can interpolated testing reduce retrieval-induced forgetting?" (2013). Graduate Theses and Dissertations. 16083.