False Memory for Words

The developmental literature contains early examples of papers on false memory for words, with articles by Felzen and Anisfeld (1970) and Cramer (1972) being classics in the area. For most of the past 3 decades, however, the study of children's false memories has been dominated by memory for heavily schematized information, such as that which is presented in narrative tasks. Recently, this situation has begun to change, with an increasing number of studies appearing in which children recall or recognize items from word lists. In the past 4 to 5 years, the flow of studies that rely on these more traditional Ebbinghaus-like designs has exceeded those that focus on false memory for schematized information. There are at least two reasons for the move toward such designs. First, developmental investigators have become increasingly concerned with formulating and testing theoretical hypotheses about the causes of false memories. Traditional word-list methodologies make it possible to achieve the high levels of experimental control that are necessary to precisely discriminate among contrasting hypotheses (similarly, for sentence paradigms; Brainerd, Reyna, & Estrada, 2006; Reyna & Kiernan, 1994). Also, existing mathematical models of false memory are defined over such methodologies, for both word lists and sentences, so that they provide the opportunity to secure uncontaminated estimates of key theoretical processes (Brainerd & Reyna, 2005). The second reason (see Reyna et al., 2007) is the advent of a particular word-list methodology that is simple and easy to use, that does not make excessive demands on the abilities of young children (as some schematic memory tasks do), and that is able to produce very high levels of false memory in adults. That task is the DRM paradigm, which we briefly mentioned earlier. We focus on research in which this paradigm was used to study children's false memory, both because those studies are more numerous than those with any other word-list methodology and, more important, because those studies have produced a rich yield of theoretically incisive findings. Below, we begin with a synopsis of the DRM procedure and key findings with adults. Then, we move on to a consideration of developmental studies.

the drm paradigm

Although intense scientific interest in false memory is a recent phenomenon, this methodology for inducing such errors was devised much earlier—nearly a half-century ago, in fact. In this task, participants study short word lists and perform free recall ("Tell me all the words you can remember, in any order") immediately after the last one is presented. Such word lists can vary in the extent to which the words are related to each other in meaning. The words might be chosen so that they share no meaning at all, or each word might share meaning with a few others (e.g., the list might be composed of 15 words, with 3 words apiece from each of 5 taxonomic categories, such as animals, cities, colors, furniture, and foods). As the degree of shared meaning increases, so does false memory: participants tend to recall unpresented words, called intrusions, that have meanings that are similar to the meanings that have been repeatedly cued by list words. Intrusion rates are not normally very high, and are almost always lower than 5% in young adults (e.g., Bjorklund & Muir, 1988).

Deese (1959) found that intrusion rates can be much higher, however. He devised an extreme version of semantically related lists in which all the list words shared meaning by virtue of their association with a single common word, which was not presented as part of the list. To take a concrete example, suppose that participants study the following word list: song, note, sound, piano, sing, noise, band, horn, art, instrument, symphony. This list strikes one as having a coherent meaning inasmuch as the words seem to revolve around music. Indeed, word association norms from the 1950s (Russell & Jenkins, 1954) show that when college students were given the word music and asked to write the first word that comes to mind, these 12 words were the most frequently written ones. Deese used these same norms to construct a total of 36 different lists, each of which consisted of the 12 most frequently given associates of a different common word (e.g., chair, foot, rough, window). His participants listened to each list, and, immediately after the 12th word was heard, they were told to recall as many words from the list as they could remember. The remarkable finding was that, even though the entire list took only about 20 seconds to present, semantic intrusions occurred at very high rates. Nearly all of the 36 stimulus words (usually called "critical distractors") were recalled at much higher levels than the nominal 5% intrusion rate for unpresented exemplars of categorized lists. Far more strikingly, two-thirds of them were falsely recalled 20% or more of the time. To appreciate just how powerful this illusion is, remember that the list is very short (only 12 words) and that participants have just finished listening to it.

Many years later, Roediger and McDermott (1995) rediscovered Deese's task, which is why it is now called the DRM paradigm. They extended the procedure to old/new recognition. Explicitly, participants listened to a series of Deese's (1959) lists, recalling each list immediately after the last word was heard. After the lists had been recalled, the subject responded to a comprehensive recognition test that covered all of the lists. The test probes consisted of some words from each list (e.g., song, note, sound, piano), the unpresented critical distractor for each list (e.g., music), and some unpresented words that were unrelated to any of the lists (e.g., computer, soccer). They found that the level of false memory was even higher for recognition than for recall. The overall false-alarm rates for the critical distractors were in the range of 70% to 80%. This rather astonishing level of memory falsification has been replicated many times by other researchers (e.g., Payne, Elie, Blackwell, & Neuschatz, 1996).

Single Parenting Becoming the Best Parent For Your Child

Single Parenting Becoming the Best Parent For Your Child

Parenting is a challenging task. As a single parent, how can you juggle work, parenting, and possibly college studies single handedly and still manage to be an ideal parent for your child? Read the 65-page eBook Single Parenting Becoming The Best Parent For Your Child to find out how. Loaded with tips, it can inspire, empower, and instruct you to successfully face the challenges of parenthood.

Get My Free Ebook


Post a comment