Libby, K. and Neisser, U. (2001) Structure and strategy in the associative false memory paradigm. Memory, 9, 145-163.

The DRM paradigm (Deese, 1959; Roediger & McDermott, 1995) has received a great deal of attention for its false memory effect. A DRM list contains thematically related words, which tend to elicit false memories. Consider the following example: BED, REST, AWAKE, TIRED, DREAM…etc All words relate to a keyword. In this instance, the keyword is SLEEP. It has been found that participants have strong false memory effect for the word SLEEP when it is presented on a recognition test. Many interpretations for this effect have been formulated. First, consider the association perspective. According to the association perspective, false memories for the key word transpire because words presented on the list repeatedly cue it. So, words like BED, REST, AWAKE, TIRED, DREAM cue the key word SLEEP during their presentation. The total associative strength of the list combine to predict the likelihood of false recall for the key word. Another interpretation to consider is based on gist (Brainerd & Reyna, 1998). When participants use a gist representation of the list, the key word is likely to be recalled (e.g. BED, REST, AWAKE, TIRED, DREAM cue the gist representation for SLEEP because each item relates to a SLEEP theme).

The purpose of the present paper is to demonstrate that the structure of the DRM list influences participants’ willingness to use a gist-based strategy, which should affect false recall rates. Another aspect important to the authors’ purpose pertains to varying memorial strategies. Thus, three experiments were designed to emphasize this variation. It is predicted that recall should be reduced in situations when gist-based strategies are discouraged. Each experiment is designed to take the focus away from the gist structure of the lists. Shifting participants focus away from gist should significantly reduce false recall of key words.

Method

Experiment 1

An alternative memory strategy focusing upon working memory and rehearsal should be used. Participants repeated DRM words to themselves. The repetition is more effective for short lists versus long lists – as long as no interference is present. There are four conditions: list length (short or long) was crossed with interference (present or absent).

Participants Seventy-two Cornell undergraduates participated in exchange for extra course credit.

Materials Stimuli were based on the middle 18 lists from Roediger and McDermott’s (1995) Appendix. Word lists consisted of 15 words in conditions long-list/distraction and long-list/rehearsal. Participants in short-list condition heard six lists of either length 6, 7, or 8.

Procedure Participants were tested in groups of 1 to 6. Each list participants listened to was followed by a 30 s interval. During the interval, participants either counted or rehearsed; then participants were signaled to recall the list. Those in the short-list condition were given the full amount of recall time of those in the long-list condition. Participants in the long-list/distraction condition were told to remember some lists of words and to do some arithmetic problems. Participants were also told to complete the test with care and to not guess. Those in the long-list/rehearsal condition were informed that this was a memory experiment and that they would be asked to remember some lists of words. Similar procedures were followed in the short-list/distraction and short-list/rehearsal conditions. However, lists varied in length and the duration for recall was lengthened.

Results

False recall of key words: False recall of key words occurred often in the long-list/distraction and long-list/rehearsal conditions. The rate of falsely recalling the key word in the short-list/rehearsal condition is significantly lower than in the short-list/distraction condition (See Table 1).

False recall of other non-presented words: There were no significant interactions involving the non-presented item type factor. There is a significant main effect of length, a significant interaction of length and activity, and a marginally significant main effect of non-studied list type. List length had an effect upon false recall of other non-presented words.

Veridical recall of list words The rate of veridical recall for the short-list/rehearsal condition is significantly higher than in the short-list/distraction condition; and also higher than in the 2 long-list conditions.

**Short lists and a sufficient amount of practice time helped participants to falsely recall the key words significantly less often than those in other conditions (i.e. short-lists/distraction, long-list/distraction and no distraction). Shorter lists maximize verbatim memory strategy, thus reducing the reliance on gist. Results are consistent with previous research which shows that increasing the distinctive features of presented words reduces false memories in the DRM paradigm.

Lists that offer a different associative organization that are not highly suggestive of the key word may reduce false memory. Participants are asked to guess the ‘secret’ word contained in the meaning of the list. Guessing the key word as the secret word should eliminate false recall of the key word. Rates of false recall for other non-presented words should not be influenced.

Participants. Thirty-seven Cornell undergraduates were given extra course credit for their participation.

Materials. The same lists used in Experiment 1 were used. Participants were assigned to the long-list/game and short-list/game condition.

Procedure. Groups of 1 to 6 participants were tested together. All were given 18 lists. The procedure is similar to Experiment 1. Participants recorded their responses in a booklet of 18 response sheets. Practice trials were given.

Results

False recall of key words Trails in which participants correctly guessed the key word as the secret word are referred to as C-trails. If the participant failed to guess the key word, then the trails are referred to as X-trails. Participants of both conditions rarely falsely recalled the key word on C-trails. However, participants falsely recalled the key word more frequently on X-trails (See Table 2).

False recall of other non-presented words There was a significant effect of list length. The main effect of trial type on false recall of other non-presented words is not significant. This result supports the prediction that the effect of guessing the key word as the secret word would be different for false recall of the key word than for false recall of other non-presented words.

Veridical recall Participants recalled a larger proportion of the list words on C-trials than on X-trials. The effect of list-length on veridical recall does not differ between Experiments 1 and 2, regardless of whether C- or X-trials are considered.

** Experiment 2 demonstrated that DRM lists presented in a context emphasizing a higher-order structure that excludes the key word can almost eliminate false recall of the key word, even with long word lists. C- and X-trials from both conditions of Experiment 2 strengthen the claim that the game reduced false recall of the key word due to an effect of higher-order structure on memory strategies.

The DRM lists of this experiment were designed to compete with gist processes. Participants were to pay special attention to focus words during encoding. They were asked to count and report the number of focus words on the list, remember those focus words in the order in which they were presented, and remember as many of the other list words as possible. It was predicted that key-focus participants should produce fewer false reports of the key word than control participants.

Participants. Thirty-six Cornell undergraduates participated for extra course credit.

Materials. Fifteen or Roediger and McDermott’s (1995) lists include at least one word that starts with the same letter as the key word. Ten of these lists were modified and used in Experiment 3A. The DRM lists were 9, 10, or 11 words long.

Procedure. Groups of 1 to 6 participants were tested together. The lists were presented by audiotape. Those in the key-focus condition counted the amount of words that began with the focus letter (for each list). They were warned not to guess and were given a practice trail. Participants in the control condition were told that the experiment pertained to memory.

Results

False recall of key words Participants falsely recalled the key word infrequently in the key-focus condition as compared to the control condition (See Table 3).

Veridical recall The key-focus condition had fewer veridical recalls than the control condition (See Table 3). This is contrary to results found in Experiments 1 and 2. The conditions in these experiments with the lowest rates of false recall of key word also had the highest rates of veridical recall.

** The focus-letter task greatly reduced the rate of falsely recalling the key word, as predicted. The key-focus instructions reduced veridical recall of the list words (which might reflect an interference effect). The results raise the question that it could be possible that both recall rates could decrease as a result of increased task difficulty. Experiment 3 provides a direct empirical test for this question.

Focus lists from Experiment 3A were modified so that each list contained 2 alternative sets of focus words. There were 2 focus conditions: given word letters (key-focus) or other letters (other-focus) as focus letters. There was also a control condition.

Participants. Fifty-five Cornell undergraduates participated for extra course credit.

Materials. Lists from Experiment 3A were modified (see above).

Procedure. Participants were tested together in groups of 1 to 8. Those in the control condition were given the same type of directions and answer booklets as the control group in Experiment 3A. Instructions for focus participants in Experiment 3A were given to key-focus and other-focus conditions. The focus in letters given to participants differed (from those given in Exp. 3A).

Results

False recall of key words False recall for the key word was lower in the key-focus condition than in the other 2 conditions. Rates in the other 2 conditions were similar. The rate of falsely recalling the key word in the key-focus condition and that in the other 2 conditions was significant (See Table 3).

Veridical recall List words veridically recalled were the same as the key-focus and other-focus conditions (See Table 3). The rate of veridical recall in both focus conditions and in the control condition was significant.

** The results of Experiment 3B contradict the hypothesis. False recall of key words was reduced only when participants focused on different letters. Veridical recall was reduced regardless of focus letter.

The experiments presented in the present paper demonstrate that reductions in false recall occur because the structure of the experimental situations encouraged participants to use strategies other than just depending on the gist of the list. False recall was shown to be greatly reduced on trails where participants guessed the secret word. List length was not a factor that affected this reduction. It was also shown that increased saliency of individual list items, which reduces false memory, can be manipulated in different ways and still maintain the reduction effect. The context of the tasks seemed to greatly affect participant’s memories of the lists. Many different theories have been offered as to why the results of the present experiment were obtained.