Creating False Memories of Words with or without Recognition of List Items: Evidence for Nonconscious Processes

Subjects exposed to lists of semantically related words falsely remember nonstudied words that are associated with the list items (e.g., Deese, 1959; Roediger & McDermott, 1995). To determine if subjects would demonstrate this false memory effect if they were unable to recognize the list items, we presented lists of semantically related words with or without a concurrent memory load at rates of 2 s, 250 ms, or 20 ms per word (Experiment 1, between-subjects design) and 2 s or 20 ms per word (Experiment 2, within-subjects design). We found that the subjects falsely recognized semantically related nonstudied words in all conditions, even when they were unable to discriminate studied words from unrelated nonstudied words. Recognition of list items was unnecessary for the occurrence of the false memory effect. This finding suggests that this memory illusion can be based on the nonconscious activation of semantic concepts during list presentation.     

Dissociated developmental trajectories for semantic and phonological false memories

False recognition following presentation of semantically related and phonologically related word lists was evaluated in 8-, 11-, and 13-year-olds. Children heard lists of words that were either semantic (e.g., bed, rest, wake ...) or phonological associates (e.g., pole, bowl, hole ...) of a critical unpresented word (e.g., sleep, roll), respectively. A semantic false memory was defined as false recognition of a semantically related but unpresented word. A phonological false memory was defined as false recognition of a phonologically related but unpresented word. False memories in the two tasks showed opposite developmental trends, increasing with age for semantic relatedness and decreasing with age for phonological relatedness.     

Dissociated developmental trajectories for semantic and phonological false memories

False recognition following presentation of semantically related and phonologically related word lists was evaluated in 8-, 11-, and 13-year-olds. Children heard lists of words that were either semantic (e.g., bed, rest, wake …) or phonological associates (e.g., pole, bowl, hole …) of a critical unpresented word (e.g., sleep, roll), respectively. A semantic false memory was defined as false recognition of a semantically related but unpresented word. A phonological false memory was defined as false recognition of a phonologically related but unpresented word. False memories in the two tasks showed opposite developmental trends, increasing with age for semantic relatedness and decreasing with age for phonological relatedness.     

False recall and false recognition induced by presentation of associated words: Effects of retention interval and level of processing

The effects of retention interval and level of processing on false recall and false recognition of associates were examined. False recall and false recognition were induced by presenting subjects with words closely associated with a non-studied word. Both level of processing and retention interval affected false recall (Experiment 1) and false recognition (Experiment 2) in the same direction with which they affected accurate recall and accurate recognition. That is, semantically processed lists exhibited higher levels of later false recall and false recognition than did superficially processed lists. Furthermore, a decline in false recall and false recognition occurred across retention intervals of 0, 2, and 7 days. However, the decline in false recall and false recognition was less pronounced than the decline in accurate recall and accurate recognition. Results are consistent with source monitoring and fuzzy trace explanations of false recall and false recognition.     

Presentation modality and mode of recall in verbal false memory

False memories were investigated for aurally and visually presented lists of semantically associated words. In Experiment 1, false written recall of critical intrusions was reliably lower following visual presentation compared with aural presentation. This presentation modality effect was attributed to the use of orthographic features during written recall to edit critical intrusions from visually presented lists. As predicted by this hypothesis, the modality effect was eliminated when the mode of recall was spoken rather than written. In Experiment 2, the modality effect in written recall was again replicated and then eliminated with an orienting task that ensured orthographic encoding even of aurally presented words. Thus, the modality effect appears to depend on using orthographic information to distinguish true from false verbal memories.     

The importance of material-processing interactions in inducing false memories

Deep encoding, relative to shallow encoding, has been shown to increase the probability of false memories in the Deese/Roediger-McDermott (DRM) paradigm (Thapar & McDermott, 2001; Toglia, Neuschatz, & Goodwin, 1999). In two experiments, we showed important limitations on the generalizability of this phenomenon; these limitations are clearly predicted by existing theories regarding the mechanisms underlying such false memories (e.g., Roediger, Watson, McDermott, & Gallo, 2001). Specifically, asking subjects to attend to phonological relations among lists of phonologically associated words (e.g., weep, steep, etc.) increased the likelihood of false recall (Experiment 1) and false recognition (Experiment 2) of a related, nonpresented associate (e.g., sleep), relative to a condition in which subjects attended to meaningful relations among the words. These findings occurred along with a replication of prior findings (i.e., a semantic encoding task, relative to a phonological encoding task, enhanced the likelihood of false memory arising from a list of semantically associated words), and they place important constraints on theoretical explanations of false memory.     

Episodic generation can cause semantic forgetting: Retrieval-induced forgetting of false memories

In two experiments, we tested whether false recognition and false recall were prone to retrieval-induced forgetting, using the retrieval practice paradigm (Anderson, Bjork, & Bjork, 1994). Participants encoded lists of cue-target word pairs associated with a nonpresented, critical theme word and then engaged in retrieval practice for half of the word pairs from half of the lists. As expected, unpracticed targets from practiced lists were recognized (Experiment 1) and recalled (Experiment 2) less well than those from unpracticed lists. In addition, false recognition and false recall of critical items associated with practiced lists was lower than false recognition and false recall of items associated with unpracticed lists. We argue that false memories are prone to inhibitory mechanisms engendered by the retrieval practice paradigm. The results are consistent with the claim that semantically activated critical themes interfere with the episodic retrieval of list words and that inhibition decreases the activation level of these interfering memory representations during retrieval practice.     

False memories of emotional and neutral words

This study used the Deese-Roediger-McDermott paradigm to investigate the direction and the extent to which emotional valence in semantic word lists influences the formation of false memories (FM). The experimental paradigm consisted of 1) a study phase (learning of neutral and negative lists of words semantically associated to a non-presented critical lure (CL), 2) a free recall phase, and 3) a recognition phase. Participants had to indicate whether the displayed item was "new" (new item or non-studied CL) or "old" (studied list item). CL associated with negative word lists elicited significantly more FM than CL associated with neutral word lists. This finding is in contrast to previous work showing that emotional words elicit fewer FM than neutral words. The results of our study also suggest that valence is capable of influencing emotional memory in terms of encoding and retrieval processes.     

Does collaborative remembering reduce false memories?

Collaborative remembering refers to recall by groups rather than by an individual. Three experiments investigated whether, relative to individual remembering, collaborative remembering decreased correct recall and false recall using the Deese‐Roediger‐McDermott paradigm. Participants were first asked to study and recall five lists of 15 words that were each semantically associated with a critical non‐presented word. Half the participants recalled the words by themselves, while the remaining half were assigned to pairs and collaboratively recalled the words. In Experiment 1, pairs produced the same number of false or correct words as individuals who were tested alone. In Experiment 2, the interpersonal closeness of the groups was also manipulated: friends and pairs who were not friends were assigned to the collaborative groups. Both friends and non‐friends produced fewer false or correct words than individuals. Experiment 3, in which the performance of the individuals and non‐friend pairs were compared using a recall test of the same 75 words as the previous experiments, replicated the results of Experiment 2. These results are discussed in terms of the retrieval‐strategy disruption.     

Opposing effects of phonological similarity on item and order memory of words and nonwords in the serial recall task

This study shows that the classical phonological similarity effect (PSE) in immediate serial recall is critically affected by the lexicality of list items, the type of phonological similarity involved, and the scoring procedure. PSE was present in the serial recall score when phonologically distinct words were compared to words that share the middle vowel and end consonant (rhyming lists). PSE was absent in the serial recall score when phonologically distinct words were compared to words that share the initial and final consonants (consonant frame lists). There was a reversal of PSE in serial recall of nonwords when comparing distinct lists to both types of similar lists. Recall accuracy on the other hand was higher for distinct lists regardless of lexicality. Item errors dominated in relation to order errors in the case of nonwords, whereas order errors dominated in relation to item errors in the case of words. Furthermore, order errors were more common for phonologically similar lists, whereas item errors were more common for phonologically distinct lists. This may be the result of intra‐list and inter‐list interference, respectively. The dominance of the former error type may cause a classical PSE, whereas the dominance of the latter error type may cause a reversal of PSE. Finally, an item identification task yielded no evidence of an association between intra‐list interference and discriminability of items in a list.     

Opposing effects of phonological similarity on item and order memory of words and nonwords in the serial recall task

This study shows that the classical phonological similarity effect (PSE) in immediate serial recall is critically affected by the lexicality of list items, the type of phonological similarity involved, and the scoring procedure. PSE was present in the serial recall score when phonologically distinct words were compared to words that share the middle vowel and end consonant (rhyming lists). PSE was absent in the serial recall score when phonologically distinct words were compared to words that share the initial and final consonants (consonant frame lists). There was a reversal of PSE in serial recall of nonwords when comparing distinct lists to both types of similar lists. Recall accuracy on the other hand was higher for distinct lists regardless of lexicality. Item errors dominated in relation to order errors in the case of nonwords, whereas order errors dominated in relation to item errors in the case of words. Furthermore, order errors were more common for phonologically similar lists, whereas item errors were more common for phonologically distinct lists. This may be the result of intra-list and inter-list interference, respectively. The dominance of the former error type may cause a classical PSE, whereas the dominance of the latter error type may cause a reversal of PSE. Finally, an item identification task yielded no evidence of an association between intra-list interference and discriminability of items in a list.     

False memories are hard to inhibit: differential effects of directed forgetting on accurate and false recall in the DRM procedure

Directed forgetting research shows that people can inhibit the retrieval of words that they were previously instructed to forget. The present research applied the directed forgetting procedure to the Deese/Roediger and McDermott (DRM) recall task to determine if directed forgetting instructions have similar or different effects on accurate and false memory. After studying lists of semantically related words, some participants were told to forget those lists, whereas other participants were not. All participants were then shown additional lists to remember. Following study, all participants were asked to free recall as many of the studied words as possible, including those they were previously instructed to forget. Directed forgetting instructions inhibited the accurate recall of studied words, but not the false recall of nonstudied critical words, whether measured by a within-participant or between-participants design. Contrary to an implicit activation hypothesis, false memories survived instructions to forget. These findings were reviewed in terms of fuzzy trace theory and the activation/monitoring approach to false memory.     

Beneficial Effect of Pictures on False Memory in the DRMRS Procedure

Using the Deese-Roediger-McDermott-Read-Solso (DRMRS) procedure in which list items are shown that are semantically related to a central theme that is not presented, undergraduates (N?=?61) studied either short or long lists in the form of auditory words alone, pictures alone, or pictures accompanied by auditory words. Correct recall and correct recognition scores were similar in all three conditions. However, false memory (reporting the non-presented central theme) was reduced with pictures alone compared to words alone for the recall task, and for the recognition task with short lists. Confidence was higher for correct recall and correct recognition than for false recall and false recognition, but people who were higher or lower in confidence did not differ on any memory measure. Theoretical and practical implications of these results are discussed.     

Remembering words not presented in lists: Can we avoid creating false memories?

Can subjects avoid creating false memories as outlined in Roediger and McDermott’s (1995) false recognition paradigm if they are forewarned about this memory illusion? We presented subjects with semantically related word lists, followed by a recognition test. The test was composed of studied words, semantically related nonstudied words (critical lures), and unrelated nonstudied words. One group of subjects was uninformed about the false recognition effect, a second group was urged to minimize all false alarms, and a third group was forewarned about falsely recognizing critical lures. Compared with the uninformed and cautious subjects, the forewarned subjects reduced their false alarm rate for critical lures, and they made remember and know judgments equally often for recognized studied words and critical lures. But forewarning did not eliminate the false recognition effect, as these subjects and those in the other groups made numerous false recognitions in this task.     

Revisiting the rise and fall of false recall: Presentation rate effects depend on retention interval

Leading theories of false memory predict that veridical and false recall of lists of semantically associated words can be dissociated by varying the presentation speed during study. Specifically, as presentation rate increases from milliseconds to seconds, veridical recall is predicted to increase monotonically while false recall is predicted to show a rapid rise and then a slow decrease-a pattern shown by McDermott and Watson (2001) in a study using immediate recall tests. In three experiments we tested the generality of the effects of rapid presentation rates on veridical and false memory. In Experiments 1 and 2 participants exhibited high levels of false recall on a delayed recall test, even for very fast stimulus onset asynchronies (SOA)-contrary to predictions from leading theories of false memory. When we switched to an immediate recall test in Experiment 3 we replicated the pattern predicted by the theories and observed by McDermott and Watson. Follow-up analyses further showed that the relative output position of false recalls is not affected by presentation rate, contrary to predictions from fuzzy trace theory. Implications for theories of false memory, including activation monitoring theory and fuzzy trace theory, are discussed.     

Revisiting the rise and fall of false recall: Presentation rate effects depend on retention interval

Leading theories of false memory predict that veridical and false recall of lists of semantically associated words can be dissociated by varying the presentation speed during study. Specifically, as presentation rate increases from milliseconds to seconds, veridical recall is predicted to increase monotonically while false recall is predicted to show a rapid rise and then a slow decrease—a pattern shown by McDermott and Watson (2001) in a study using immediate recall tests. In three experiments we tested the generality of the effects of rapid presentation rates on veridical and false memory. In Experiments 1 and 2 participants exhibited high levels of false recall on a delayed recall test, even for very fast stimulus onset asynchronies (SOA)—contrary to predictions from leading theories of false memory. When we switched to an immediate recall test in Experiment 3 we replicated the pattern predicted by the theories and observed by McDermott and Watson. Follow-up analyses further showed that the relative output position of false recalls is not affected by presentation rate, contrary to predictions from fuzzy trace theory. Implications for theories of false memory, including activation monitoring theory and fuzzy trace theory, are discussed.     

False recognition following study of semantically related lists presented in jumbled word form

Three experiments explored a jumbled word effect in false recognition. Lists of theme-related items were presented in word or nonword form. Results indicated that critical lures semantically related to studied items were falsely recognised regardless of whether they were presented as words or nonwords. High false recognition rates to either SLEEP or SELEP following study of an appropriate theme list of items in nonword form should only occur if nonwords are recoded at study. With study conditions conducive to recoding, jumbled words induced false recognitions based on semantic associations among their respective base words. Disrupting a recoding process by creating “difficult” letter rearrangements for jumbled words (Experiment 2) appeared to eliminate the false recognition effect. In Experiment 3, presentation durations ranged from 110 ms to 880 ms. Although there was little evidence of a semantic false recognition effect at the fastest presentation rate, the brief durations appeared to be effective in eliminating the effect when items were studied in nonword form. These results appear to be consistent with an encoding activation/retrieval monitoring model.     

False recognition following study of semantically related lists presented in jumbled word form

Three experiments explored a jumbled word effect in false recognition. Lists of theme-related items were presented in word or nonword form. Results indicated that critical lures semantically related to studied items were falsely recognised regardless of whether they were presented as words or nonwords. High false recognition rates to either SLEEP or SELEP following study of an appropriate theme list of items in nonword form should only occur if nonwords are recoded at study. With study conditions conducive to recoding, jumbled words induced false recognitions based on semantic associations among their respective base words. Disrupting a recoding process by creating "difficult" letter rearrangements for jumbled words (Experiment 2) appeared to eliminate the false recognition effect. In Experiment 3, presentation durations ranged from 110 ms to 880 ms. Although there was little evidence of a semantic false recognition effect at the fastest presentation rate, the brief durations appeared to be effective in eliminating the effect when items were studied in nonword form. These results appear to be consistent with an encoding activation/retrieval monitoring model.     

False memories in children and adults: age,distinctiveness, and subjective experience

This study investigated developmental trends associated with the Deese/Roediger-McDermott false-memory effect, the role of distinctive information in false-memory formation, and participants' subjective experience of true and false memories. Children (5- and 7-year-olds) and adults studied lists of semantically associated words. Half of the participants studied words alone, and half studied words accompanied by pictures. There were significant age differences in recall (5-year-olds evinced more false memories than did adults) but not in recognition of critical lures. Distinctive information reduced false memory for all age groups. Younger children provided with distinctive information, and older children and adults regardless of whether they viewed distinctive information, expressed higher levels of confidence in true than in false memories. Source attributions did not significantly differ between true and false memories. Implications for theories of false memory and memory development are discussed.     

Advantages and disadvantages of phonological similarity in serial recall and serial recognition of nonwords

The phonological similarity effect (PSE) was studied with lists of nonwords in one task of serial recall and one task of serial recognition. PSE was critically affected by the scoring procedure and the type of phonological similarity involved, and the effect diverged in several ways from the findings of previous studies on words. PSE was absent in serial recall, regardless of scoring procedure, when phonologically similar items that shared the midvowel were compared with phonologically distinct items. PSE was reversed when serial recall and item recall scores of rhyme items and consonant frame items were compared with distinct items, but it was present in the position accuracy score of rhyme lists. In serial recognition, PSE was absent when rhyme lists were compared with distinct lists. Recognition was better for consonant frame lists than for rhyme lists, and there was a marginally significant reversal of PSE when consonant frame lists were compared with distinct lists. In the view of Fallon, Groves, and Tehan's (1999) study and the present study, rhyming improves item recall and serial recall but diminishes position accuracy, regardless of lexicality. But consonant frame lists with differing midvowels have higher item recall, serial recall, and position accuracy scores than do rhyme lists.