Academic journal article Memory & Cognition

Disentangling Encoding versus Retrieval Explanations of the Bizarreness Effect: Implications for Distinctiveness

Academic journal article Memory & Cognition

Disentangling Encoding versus Retrieval Explanations of the Bizarreness Effect: Implications for Distinctiveness

Article excerpt

Recall effects attributed to distinctiveness have been explained by both encoding and retrieval accounts. Resolution of this theoretical controversy has been clouded because the typical methodology confounds the encoding and retrieval contexts. Using bizarre and common sentences as materials, we introduce a paradigm that decouples the nature of the encoding context (mixed vs. unmixed lists of items) from the retrieval set (mixed vs. unmixed retrieval sets). Experiment 1 presented unmixed lists for study, and Experiment 2 presented mixed lists for study. In both experiments, significant bizarreness effects were obtained in free recall when the retrieval set intermixed items but not when the retrieval set consisted of only one item type. Also, Experiment 1, using a repeated testing procedure, did not reveal evidence for more extensive encoding of bizarre sentences than of common sentences. The results support the idea that retrieval dynamics primarily mediate the bizarreness effect, and perhaps more generally, distinctiveness effects.

One of the most persistent and galvanizing issues in memory research and theory has been the relative contributions of encoding versus retrieval processes to retention. The issue is sharply drawn in the debate concerning the locus of distinctiveness effects in memory. A commonly accepted claim, both as an empirical description and sometimes as an explanatory construct for empirical findings (Hunt & Lamb, 2001; Schmidt, 1991), is that distinctiveness enhances memory. There is far less agreement, however, about whether encoding processes or retrieval processes are responsible for so-called distinctiveness effects. Certainly, part of the debate turns on the complexity of the issue and the possible contribution of either encoding or retrieval processes across variations in materials, procedures, participants, and so on (see Schmidt, 1991, for a review). We suggest though that part of the persistence of this theoretical debate hinges on a problematic aspect of the research methodology that characterizes the work often related to distinctiveness. In this article, we first summarize the theoretical debate and identify the methodological problem. We then report two experiments designed to remedy the methodological problem, with the goal of disentangling encoding influences from retrieval influences in distinctiveness effects.

We embed the development of our ideas primarily in the context of the effects of bizarre stimulus materials on free recall, an effect that has been closely linked to distinctiveness processes (Einstein & McDaniel, 1987; McDaniel, DeLosh, & Merritt, 2000; McDaniel & Einstein, 1986; McDaniel, Einstein, DeLosh, May, & Brady, 1995; Waddill & McDaniel, 1998). In a popular paradigm, bizarre materials are constructed by presenting common items in bizarre relations to one another, either through pictorial or sentential materials (e.g., The MAID licked AMMONIA off the TABLE). Parallel common materials use the same items related in a usual fashion (e.g., The MAID spilled AMMONIA on the TABLE). The standard finding is that when the bizarre and common materials are intermixed in the acquisition list, bizarre items are better recalled than common items. By contrast, when the bizarre and common items are presented in unmixed lists, there is no free recall advantage of bizarre items (see Einstein & McDaniel, 1987, for a review, and more recently McDaniel et al., 1995). This effect of list composition with regard to distinctiveness-like effects is not limited to materials that are specifically bizarre, but also extends to other presentations that might be associated with distinctive features. For example, fragmented items (that are generated) generally show better free recall than intact items (that are read) in mixed but not unmixed lists (Hirshman & Bjork, 1988; Slamecka & Katsaiti, 1987), humorous sentences are better recalled than nonhumorous sentences in mixed but not unmixed lists (Schmidt, 1994), and orthographically distinct items are better recalled than orthographically regular items in mixed but not unmixed lists (Hunt & Elliott, 1980). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.