Spaced Practice for Conceptual Understanding

I've attacked and studied spaced practice throughout my work and my classes. I've blogged about it many times (see here). I was asked to review a journal article on the effects of spaced practice on understanding concepts and functions. I was not impressed, and it was mostly with the writing capabilities of the authors.

The journal article "Effects of Spaced Versus Massed Training in Function Learning" offers a confusing and difficult to understand look into whether or not spaced learning is as effective in learning concepts as it is with non-linked information. 

McDaniel et al. (2013) describe the rationale for the experiment as a "focus on training of a particular conceptual entity within one training session using repeated training stimuli" (p. 1417). The authors state that their focus is on concept understanding and function learning rather than memorization of verbal paradigms, as has often been studied (McDaniel et al., 2013). Although the authors provide compelling reasons when expressing the need for this study, and they do an excellent job of citing other relevant studies and works and discussing the differences between spaced practice, massed learning, and interleaving and how these will be used in their research, they fail to communicate clearly the experiment design or results in a manner that makes sense for the reader.   

McDaniel et al. (2013) begin their article by describing the principles behind long-term memory development compared to working memory and even include discussions of functional relationships, which they intend to study. McDaniel et al. (2013) provide examples of these functional relationships as predictions of one thing based on the relationship to another, such as "predictions of harvest yields based on the amount of rainfall" (p. 1418) or stock market yields and P/E ratio. The author's discussion of understanding abstractions and concepts are understandable and builds the reader's frame of reference for their research design. The purpose of the experiment is quite valid in that being able to determine if a more effective learning delivery systems like spaced learning can provide the same benefits to function and conceptual understanding would be a worthwhile undertaking. However, McDaniel et al. (2013) become far too ambiguous and quickly lose the reader when discussing the methods and results of their experiments. The description of the experiments could have been more fully understood by providing some examples of test questions and actual results from individual test subjects. Instead, McDaniel et al. (2013) lose the reader in their data and results analysis to the point that their intentions to measure concept understanding are cluttered and confusing. The difficulty of measuring both spaced and massed learning against extrapolation, interpolation, against random verse strategic delivery comprises a multitude of components to analyze that never tend to come together for the reader nor become clear. The authors would have benefitted from showing a table of their overall results to help the reader track the variables and research design. 

Other studies have been completed that focus on similar aspects of spaced practice and do so in an easier to understand manner. For example, articles such as Karpicke's (2016) wherein the author found that "on the concept map assessments, practicing retrieval during learning produced the best performance, even better than studying the material by making concept maps" (p. 5).  Another example is in the work by Vlach and Sandhofer (2013) and their description of experiments with spaced practice for generalization in children. Their results were positive, and they surmised in their General Discussion section that in spaced practice, "perceptual features are likely to be forgotten, whereas the underlying abstract structure is likely to be remembered to a greater degree" (Vlach and Sandhofer, 2013, p. 7). These studies focused on the same research, measuring concept understanding through spaced practice delivery, yet their methods, design, and results are far easier to digest and comprehend for readers.

In terms of their experiment results, McDaniel et al. (2013) continue to be confusing and say in the Practical Applications section that repeated massed learning of random sequences "does not support good learning" (p. 1431), then state "function learning is effective with massed experiences that are strategically (rather than randomly) selected," (p. 1431), but then continue and write that cued presentation of input-output pairs yielded better results for spaced learning.  The reader is left wondering what the experiments were designed to prove and the results.  Compared to both Karpicke (2016) and Vlach and Sandhofer (2013), the ability to relate this information in an easy-to-understand and well-laid-out manner is sorely missing in this journal article. 

References

Karpicke, J. (2016). A powerful way to improve learning and memory; Practicing retrieval enhances long-term, meaningful learning. American Psychological Association. https://www.apa.org/science/about/psa/2016/06/learning-memory

McDaniel, M. A., Fadler, C. L., & Pashler, H. (2013). Effects of spaced versus massed training in function learning. Journal of Experimental Psychology. Learning, Memory, and Cognition, 39(5), 1417–1432. 

Vlach, H. A., & Sandhofer, C. M. (2012). Distributing learning over time: the spacing effect in children’s acquisition and generalization of science concepts: Spacing and generalization. Child Development, 83(4), 1137–1144.