You are on page 1of 13

Copyright The British Psychological Society

Reproduction in any form (including the internet) is prohibited without prior permission from the Society

British Journal of Psychology (2007), 98, 113 q 2007 The British Psychological Society

The British Psychological Society


www.bpsjournals.co.uk

Does collaborative remembering reduce false memories?


Masanobu Takahashi*
University of the Sacred Heart, Japan
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 DeeseRoediger-McDermott paradigm. Participants were rst asked to study and recall ve 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.

Some recent studies of memory have concentrated on social aspects, and have involved two people talking of shared experiences (Edwards & Middleton, 1986). Collaborative remembering refers to this type of remembering (see for reviews, Clark & Stephenson, 1989; Hartwick, Sheppard, & Davis, 1982; Weldon, 2000). When we consider the overall correct performance, many studies show that groups of two or more always recall more than an individual (Basden, Basden, Bryner, & Thomas, 1997; Meudell, Hitch, & Boyle, 1995; Meudell, Hitch, & Kirby, 1992; Weldon & Bellinger, 1997). However, recent researchers prefer to compare the performance of collaborative with that of nominal groups to allow for a statistical comparison (Basden et al., 1997; Weldon & Bellinger, 1997). Performance of nominal groups (the sum of the recall of each individual) is determined by pooling the non-redundant recalls of individual working alone. Thus, the performance of nominal groups represents the level of productivity one would expect if group interaction neither facilitated nor inhibited
*Correspondence should be addressed to Masanobu Takahashi, Department of Psychology, University of the Sacred Heart, Hiroo 4-chome 3-1, Shibuya-ku, Tokyo 1508938, Japan (e-mail: mtakahas@u-sacred-heart.ac.jp).
DOI:10.1348/000712606X101628

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

2 Masanobu Takahashi

group productivity. When the performance of nominal and actual groups are compared, the most interesting result is that collaborative groups recalled less than nominal groups. Weldon and Bellinger referred to this effect as collaborative inhibition. Typically, collaborative inhibition occurs with groups of three or more (Basden et al., 1997; Basden, Basden, & Henry, 2000; Weldon & Bellinger, 1997), but can occur with pairs (Andersson & Ro nnberg, 1995; Finlay, Hitch, & Meudell, 2000; Takahashi & Saito, 2004), although nominal pairs sometimes exhibit no difference from actual pairs (Basden et al., 2000). One promising explanation for collaborative inhibition is related to a retrieval inhibition process similar to the process that underlies the inhibitory effect of part-set cueing (Basden et al., 1997; Finlay et al., 2000). The part-set cueing effect refers to the phenomenon that giving a subset of the list items as recall cues inhibits recall of the remaining items (see for reviews, Nickerson, 1984; Roediger & Neeley, 1982). One largely accepted interpretation that is, the retrieval-strategy disruption interpretation posits that a participants preferred retrieval strategy can be disrupted by giving a subset of the list items during recall (Basden & Basden, 1995). Since participants in a collaborative group interact with each other, it is likely that one person could provide a subset of the target items to the other members in the group. In other words, it is possible to assume that there could involve similar retrieval-strategy inhibition under the part-set cueing and collaborative inhibition (Basden et al., 1997; Finlay et al., 2000). However, when we consider errors, previous research is not so straightforward. Although early researchers found that groups are more accurate than individuals (Perlmutter & de Montmollin, 1952; Yuker, 1955), recent researchers have shown that mean intrusion errors for collaborative groups are greater than those for individuals (Basden et al., 1997; Basden, Basden, Thomas, & Souphasith, 1998). The present experiments were primarily interested in whether, during recall, collaborative groups produce more, or less, intrusion errors than individuals. To compare collaborative and individual error rates, we require high levels of intrusion errors such as those in the Deese-Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger & McDermott, 1995). In the DRM paradigm, individual participants learn lists of words that are each semantically associated with a critical non-presented word. The typical nding is remarkable levels of false recall and false recognition (see for reviews, Roediger, Balota, & Watson, 2001; Roediger, McDermott, & Robinson, 1998). The present experiments used this DRM paradigm and the collaborative and individual intrusion errors were compared. To the best of my knowledge, only one study (Basden et al., 1998) has compared collaborative (groups of three people) and nominal (three people tested individually) groups using the DRM paradigm. Although they observed collaborative inhibition of studied words, the recall of critical non-presented words was equivalent for both kinds of groups. However, Basden et al. used an atypical turn-taking recall procedure in which each participant in the collaborative groups had to recall one word per turn. As Basden et al. discussed, it seems reasonable to assume that during such a turn-taking procedure participants in collaborative groups would feel more obligation to contribute and would have a lower threshold for response production than during a standard recall procedure in which each person could talk at will. In order to approximate as closely as possible a natural social setting and to expand the generality of their results, the present experiments used a standard free-recall procedure. Under such a free for all recall procedure, people in collaborative groups might be afraid to make a mistake and so adopt a conservative recall criterion to output words. As a result, collaborative groups should produce less false recalls than individuals. This reasoning is supported by the ndings of

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

Ross, Spencer, Linardatos, Lam, and Perunovic (2004). They found that collaboration reduced memory errors with older couples and familiar memory tasks (e.g. remembering items from a shopping list), although they replicated previous ndings of collaborative inhibition. On the basis of signal detection analyses, they concluded that collaboration induced individuals to be more conservative in the memory tasks. Spreading activation has been proposed as one explanation for the creation of false memories (Roediger & McDermott, 1995; Roediger et al., 1998, 2001). According to the spreading activation account, the critical non-presented word becomes activated as an implicit associate response to its associates in the study list. Given the strength of spreading activation is a function of the total number of activated list words, the greater number of the associate words in the list should increase the probability of false remembering. This prediction is supported by several investigators (e.g. Robinson & Roediger, 1997). Some researchers argue that false memories may be partially created during tests (Marsh, McDermott, & Roediger, 2004; Roediger & McDermott, 1995). Basden et al. (1998) hypothesized that because members of collaborative groups hear one anothers output, the number of associate words to which they are exposed should be greater than for members of nominal groups. Thus, as suggested by the research conducted by Robinson and Roediger (1997), the greater number of different associate words encountered during collaborative recall should increase false recall. However, they found that collaboration did not increase recall of critical non-presented words although it inhibited recall of presented words. However, some researchers have demonstrated an inhibitory effect of part-set cueing for the critical non-presented words, as well as for the presented words in the DRM paradigm (Ba uml & Kuhbandner, 2003; Kimball & Bjork, 2002; Reysen & Nairne, 2002). Based on the assumption that a similar retrieval-strategy disruption underlies both collaborative inhibition and part-set cueing inhibition (Basden et al., 1997; Finlay et al., 2000), and on the basis of the previously discussed Ross et al. (2004) study, I hypothesize that collaborative pairs would reduce both correct and false recalls relative to nominal pairs.

EXPERIMENT 1 Method
Participants Participants were 132 female undergraduate students at the University of the Sacred Heart. Since there is some evidence of a gender difference in communication and conversation (Anderson & Leaper, 1998) and to avoid an effect of mixed-gender groups, only females participated in the experiment. Their ages ranged from 18 to 38 years. They were classmates who had been in the same class for at least 3 months. They were given a partial course credit or volunteered to participate. Design The experimental design was a mixed 2 2 factorial ANOVA with the groups (nominal or collaborative pairs) as a between-participants variable and word types (correct or false recalls) as a within-participant variable.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

4 Masanobu Takahashi

Materials All ve lists were selected from the materials used by Miyaji and Yama (2002) and had similar structures to those of Roediger and McDermott (1995). Each list was composed of 15 words related to one critical non-presented word. The order of the ve-list presentations was consistent. Each list was recorded on audio tape by a male voice at the rate of one word per second. An auditory cue recall was also recorded after the last word of each list so that the participants had to write down the words immediately.

Procedure The experiment consisted of two successive sessions. In the rst session, participants were tested in groups of 2 to 10 people. They were told to remember the lists of words which would be presented by an audio tape player and write the last word rst and then to recall the remainder in any order. After each list the participants were given 2 minutes for an immediate written free recall. In the second session, after the fth immediate free recall test, participants were given another unexpected written free recall test. Each participant was assigned either to one of the 33 pairs or as one of the 66 individuals who recalled the material by themselves. To avoid environmental context-dependent effects (cf. Smith, Glenberg, & Bjork, 1978), they were escorted by the experimenter to separate small rooms to recall the words either individually or collaboratively. Participants in the individual recall groups were asked to write down all the words they could remember. Each pair of participants was given a sheet of paper, and one person was asked to volunteer to be the scribe. They were required to recall the material collaboratively, and to complete one recall sheet. No time limit was required to complete the nal free recall test which typically took less than 20 minutes.

Results
The individual, nominal and collaborative conditions were scored and analysed in a manner similar to Weldon and Bellingers (1997) analyses. That is, the individual recall data were computed for the total number of correct and false recalls for each individual. The nominal pairs in the nal test were formed from random combinations of participants who recalled the words by themselves. The correct and false recalls of the nominal pairs were obtained by combining each individual recall and allowing for any redundant items that appeared in both. Some previous studies (Meudell et al., 1992; 1995) compared the rst and second recall tests to analyse changes in recall status of particular items (i.e. new/forgotten items). However, separate analyses were conducted on data from the rst and second free recall tests for the following reasons. First, the reason for asking the participants for the immediate free recall is because the act of initial recall might enhance later false memories in a nal recall test (Roediger & McDermott, 1995). Second, I cannot make direct comparisons between the immediate and nal tests because I do not know how much time was spent on the nal free recall test. Noncritical intrusions, which were relatively rare, were not analysed because the data of interest were correct and false recalls (i.e. critical non-presented words). All analyses were considered as signicant at the p :05 level or better.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

Immediate test The results from the immediate test were based on individual performance to control for differences between forthcoming pairs. The left column of Table 1 presents mean proportions (the average of the data from 66 individuals) for each word type by each group on ve immediate tests. In the immediate test, 59 and 61% of the critical nonpresented words were incorrectly recalled. These proportions revealed high levels of false recall for the participants who were tested individually, which is similar to the levels of those observed in Roediger and McDermott (1995).
Table 1. Mean proportions of correct and false recalls in immediate and nal tests for individuals, nominal pairs and collaborative pairs (Experiment 1) Immediate test Nominal Correct False .62 (.08) .59 (.23) Collaborative .64 (.08) .61 (.23) Individuals .44 (.12) .58 (.28) Final test Nominal .63 (.11) .79 (.17) Collaborative .65 (.09) .72 (.26)

Note. Standard deviations are in parentheses. Mean proportion in the immediate test are the average of the data from 66 individuals in each condition.

A 2 (groups: nominal or collaborative pairs) 2 (word type: correct or false recalls) mixed analysis of variance (ANOVA) was performed on the recall scores and revealed that there were no main effects of groups and word type, F 1; 130 :68, MSE .03 and F 1; 130 2:00, MSE .03, respectively. There was no signicant interaction between groups and word type, F 1; 130 :003, MSE .03. The two groups of participants were therefore matched on their initial level of recall. Final test The right column of Table 1 presents mean proportions for each word type by each group in the nal test. A 2 2 mixed ANOVA was performed on the recall scores, in a similar manner to the analysis of the immediate test. There was a signicant main effect of word type, F 1; 64 21:18, MSE .03, p , :01, with more false recalls than correct recalls. However, there was no main effect of groups and interaction between groups and word type, F 1; 64 :05, MSE .03 and F 1; 64 1:08, MSE .03, respectively. These results show that the collaborative pairs did not recall any more words than the nominal groups.

Discussion
These results showed that collaborative pairs did not inhibit correct recalls relative to nominal groups. Put another way, these results are a failure in nding collaborative inhibition, although the result is consistent with some previous experiment showing that two-person groups do not recall more than nominal groups (Basden et al., 2000). Similarly, the false recall of critical non-presented words did not differ between either the collaborative or nominal groups. This result is contradictory to the prediction, although it is in accordance with Basden et al. (1998), who indicated that false recall of critical non-presented words was equivalent for collaborative and nominal groups of three people.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

6 Masanobu Takahashi

One explanation for the failure to obtain collaborative inhibition is that in the present experiment the pairs were allowed to vary freely. That is, most of the participants who formed the collaborative groups were not only classmates but had also been very close friends since elementary school. In fact, Andersson and Ro nnberg (1995) found that pairs of friends tend to have less collaborative inhibition than do pairs of people who are not friends. Close friends share memories and know each others retrieval style (cf. Wegner, Erber, & Raymond, 1991). Thus, it seems plausible that close friends might produce more effective retrieval cues for each other than do people who are not friends (Andersson & Ro nnberg, 1997). In the present experiment, given that most of the participants in the collaborative groups were close friends, the effect of cross-cueing might reduce or eliminate collaborative inhibition. The purpose of the second experiment was intended primarily to test the hypothesis of Experiment 1, but with manipulation of the interpersonal closeness of the groups: friends and people who were not friends were assigned to the collaborative groups. Based on the assumption that friends produce more effective retrieval cues for each other than do non-friends, we predicted that more collaborative inhibition for non-friend pairs than for friend pairs would be obtained.

EXPERIMENT 2 Method
Participants Participants were 132 undergraduate female students at the University of the Sacred Heart. Their ages ranged from 18 to 24 years. All of them volunteered to participate. Forty-four participants were assigned to the individual recall group: during both tests they recalled all the words while alone. The remaining 88 were assigned to either of the two types of collaborative recall groups: pairs of people who were not friends or who were friends. The pairs who were strangers had never talked or socialized. The pairs of friends were close friends who had socialized for at least 2 years. Design The experimental design was a mixed 3 2 factorial ANOVA with the groups (nominal pairs, non-friends or friends) as a between-participants variable and the word types (correct or false recalls) as a within-participant variable. Materials and procedure The materials and the procedure were identical to that of Experiment 1 except for the pairs, who were either non-friends or friends.

Results
The data were scored and analysed as in Experiment 1.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

Immediate test The left column of Table 2 presents mean proportions for each word type by each group on the immediate test. A 3 (groups: nominal, non-friends or friends) 2 (word type: correct or false recalls) mixed ANOVA was performed on the recall scores and revealed that there were no main effects of groups and word type, F 2; 129 1:34, MSE .03 and F 1; 129 3:62, MSE .03, respectively. There was no signicant interaction between groups and word type, F 2; 129 :72, MSE .03. As in Experiment 1, the three groups of participants were therefore matched on their initial level of recall.

Table 2. Mean proportions of correct and false recalls in immediate and nal tests for individuals, nominal pairs and collaborative groups of two people (Experiment 2) Immediate test Nominal Correct False .63 (.07) .60 (.24) Non-friends .60 (.09) .56 (.24) Friends .59 (.05) .53 (.23) Individuals .48 (.10) .62 (.21) Final test Nominal .69 (.07) .80 (.12) Non-friends .63 (.11) .66 (.19) Friends .64 (.09) .71 (.17)

Note. Standard deviations are in parentheses. Mean proportion in the immediate test are the average of the data from 44 individuals in each condition.

Final test The right column of Table 2 presents mean proportions for each word type by each group in the nal test. The recall scores were analysed with a 3 2 mixed ANOVA in a similar manner to the analysis of the immediate test. There were signicant main effects of groups and word types, F 2; 63 6:06, MSE .02, p , :01 and F 1; 63 8:22, MSE .02, p , :01, respectively. Of particular interest is that there was no signicant interaction between groups and word type, F 2; 63 1:01, MSE .02. Tukeys honestly signicant difference (HSD) post hoc comparisons indicated that nominal pairs recalled more correct or false recalls than the other two collaborative pairs, which did not differ reliably. Consequently, the results showed collaborative inhibition, thus extending them to demonstrate that a close relationship between two people does not affect the inhibition.

Discussion
In contrast to the results of Experiment 1, collaborative groups recalled less correct recalls than nominal groups. The results therefore conrmed that collaborative inhibition emerges in associatively related list learning. These results t with previous research (Andersson & Ro nnberg, 1995; Basden et al., 1998; Finlay et al., 2000; Weldon & Bellinger, 1997). More importantly, collaborative pairs reduced false recalls relative to nominal pairs as predicted. Importantly, the manipulation of the interpersonal closeness of the groups had no effect on the collaborative inhibition, contradictory to the prediction. However, Andersson and Ro nnberg (1995) found that friends tended to suffer less from collaborative inhibition than did non-friends in the recall of complex materials such as stories. In contrast, groups of friends were no better than non-friends with word recall, which is the material used in the present experiments. Thus, the results are consistent with Andersson and Ro nnbergs ndings.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

8 Masanobu Takahashi

One motivational explanation of collaborative inhibition is that of social loang, or that people may not work as hard when they are a member of a group as they would , Williams, & Harkins, 1979). It is well-known that friends, when working alone (Latane unlike people who are unfamiliar with each other, may inhibit social loang (e.g. Latane & Darley, 1970). If this is so, a different pattern would emerge between pairs of friends and pairs who are not friends. The results of Experiment 2 do not support these proposals (see also, Weldon, Blair, & Huebsch, 2000). Why was collaborative inhibition found in Experiment 2, but not in Experiment 1 despite the same procedure as in Experiment 2? Although it is not so clear why the null results emerged in Experiment 1, I might attribute the results to differences of retrievalstrategy disruption. To determine whether collaboration disrupted mutual retrieval strategies, Basden et al. (1998) calculated the adjusted ration of clustering (ARC) scores (Roenker, Thompson, & Brown, 1971) as an indicator of subjective organization. This measure ranges from 2 1.00 to 1.00, with 1.00 representing perfect organization and zero indicating no tendency beyond the chance to recall in an organized fashion. They found less organization in collaborative than individual recall, and also obtained collaborative inhibition. However, I could not use their techniques for assessing organization because the participants output order in the nal test was not recorded in the two experiments. However, since each participants output orders in the ve immediate tests were obtained, I used the relative order (RO) measures (Asch & Ebenholz, 1962), which express the number of pairs of items appearing in the same relative order between the initial presentation and the recall protocol as a proportion of the total number of pairs in the recall. A RO measure of .50 represents random recall, and a measure of 1.00 reects perfect preservation of relative order. I rst calculated the ve RO measures for each participant, and then computed the sum of ve absolute values of the differences between the two people in each pair, who had worked together (collaborative pair) or been formed from random combination (nominal pair) in the nal test. Given that participants might maintain a stable organization of the materials from the immediate to the nal recall, they should make use of it for optimal retrieval. Thus, the larger this value is, the more group members organizations are assumed to be dissimilar and susceptible to disruption by the others output. In Experiment 1, the values were lower for collaborative pairs (M :70) than for nominal pairs (M :86). However, this difference was not signicant, t 64 1:77, SE .09. In contrast, in Experiment 2, non-friend (M :80) and friend (M :81) groups had slightly greater values than the nominal group (M :78), although there was no signicant difference among the three groups, F 2; 63 :06, MSE .09. Thus, in Experiment 1, less retrieval-strategy disruption might result from partners with more similar organizations due to chance factor, thereby resulting in no evidence for collaborative inhibition. Consistent with this speculation, some researchers have found that experimental manipulations designed to increase the similarity of group members retrieval strategies eliminated the collaborative inhibition (Basden et al., 1997, 2000; Finlay et al., 2000). There was one limitation in the present experiments. In both Experiments 1 and 2, prior individual testing might make source memory errors more frequent because the participants must keep track of what was presented and what was not, and what they were able to recall on the individual tests. Thus, prior individual testing might contribute to false memory. Although some experiments on collaborative memory research have included prior individual testing and some have not, none that have measured false memory used prior individual tests. One reason that Basden et al. (1998)

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

did not nd a signicant difference might be due to whether there was prior individual testing. Therefore, the purpose of Experiment 3 was to determine whether the basic ndings of Experiment 2 would generalize to a situation in which all individual and collaborative pairs were tested only for nal recall. In Experiment 3, people who were not friends were assigned to the collaborative group and subjective organization was examined using RO and ARC measures.

EXPERIMENT 3 Method
Participants Participants were 80 female undergraduate students, 75 from the University of the Sacred Heart and the remaining 5 from another university. Their ages ranged from 18 to 22 years. All of them volunteered to participate. Of the participants 40 were assigned to the individual recall group and the remaining 40 to the collaborative recall group. The pairs who were strangers had never talked or socialized.

Design, materials and procedure The design, materials and procedure were identical to that of Experiment 1 except that there was only one recall test of the 75 words.

Results
The data were scored and analysed as in previous experiments.

Final test Table 3 presents mean proportions for each word type by each group in the nal test. The recall scores were analysed with a 2 2 mixed ANOVA in a similar manner to the analysis of the nal test in Experiment 1. There were signicant main effects of groups and word types, F 1; 38 8:95, MSE .03, p , :01 and F 1; 38 127:60, MSE .02, p , :01, respectively. Of particular interest is that there was no signicant interaction between groups and word type, F 1; 38 :96, MSE .02. These results replicated the basic nding of Experiment 2 in the situation where no prior individual testing was conducted. Thus, the results indicate that nominal pairs had more correct or false recalls than the collaborative pairs.
Table 3. Mean proportions of correct and false recalls in nal test for individuals, nominal pairs and collaborative pairs (Experiment 3) Individual Correct False .23 (.07) .51 (.25) Nominal .37 (.07) .73 (.19) Collaborative .29 (.04) .59 (.20)

Note. Standard deviations are in parentheses.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

10 Masanobu Takahashi

Subjective organization To examine whether collaboration disrupted the retrieval strategies of the two individuals in a group, I calculated subjective organization as both RO and ARC measures for the nal recall test. The average of each organizational measure for the two people in a nominal group served as that groups corresponding organizational measure. RO measures were signicantly lower for the collaborative groups (M :51) than for the nominal groups (M :56), t 38 2:28, SE .02, p , :05. However, ARC measures showed no signicant difference between the nominal groups (M :52) and the collaborative groups (M :53), t 38 :30, SE .05. Perhaps the RO measures might be more sensitive than the ARC measures in the present experiment. Since the words were presented in an auditory format and at a relatively fast rate, the participants might have relied more on serial information for subjective organization. Thus, on the basis of the RO measure, collaborative inhibition was associated with disrupted organization.

Discussion
The present experiment replicated the results of Experiment 2 and extended them to a situation where there was no prior individual testing. The results show that collaborative groups recalled fewer correct or false items than individual groups. Clearly, the pattern of the results is consistent with the prediction but is inconsistent with the ndings of Basden et al. (1998). Therefore, I rule out the possibility that Basden et al.s failure to nd a signicant difference was due to an inuence of prior individual testing.

GENERAL DISCUSSION
The results of the three experiments demonstrate that collaboration affects correct and false recalls in the same ways. Furthermore, collaborative inhibition was apparent in both Experiments 2 and 3. These results are nicely consistent with previous research (Andersson & Ro nnberg, 1995; Basden et al., 1998; Finlay et al., 2000; Weldon & Bellinger, 1997). Similarly, I obtained the inhibition of false recalls in a collaborative remembering situation, in accordance with the prediction. Basden et al. (1998) argued, on the basis of the spreading activation account, that collaborative groups produced more false memories than nominal groups because the number of associates to which collaborative groups are exposed should be greater than for members of nominal groups. However, the results of Experiments 2 and 3 provide strong evidence against the predicted pattern of the spreading activation account. As noted in the introductory section, the spreading activation account is not supported by the ndings of an inhibitory effect of part-set cueing, either for the critical nonpresented words or for the presented words in the DRM paradigm (Ba uml & Kuhbandner, 2003; Kimball & Bjork, 2002; Reysen & Nairne, 2002). Therefore, both the present data, and the ndings obtained from research on the inhibitory effect of part-set cuing, support the view that spreading activation does not play a crucial role in the creation of false memories during a test. Why then did collaborative remembering reduce false recalls in the present experiment? In their second experiment, Basden et al. (1998) showed that with categorized lists, recall of high-taxonomic frequency, critical non-presented words

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

11

was greater for collaborative groups than for nominal groups. This discrepancy may be mainly due to the methodological difference of the recall procedures. That is, Basden et al. used a turn-taking recall procedure in which each participant in the collaborative group had to recall one word per turn. In contrast, the present experiments used a free-recall procedure in which each person could talk at will. As Basden et al. discussed, group pressure to participate is presumably heightened under a turn-taking recall procedure. Perhaps such strong group pressure might force members of collaborative groups to lower their criterion for producing items, thereby producing more errors. In contrast, as noted in the introductory section, in a free-recall procedure the members in the collaborative groups might be afraid to make a mistake and thus adopt a conservative recall criterion to produce words (cf. Vollrath, Sheppard, Hinsz, & Davis, 1989). Thus, I do not rule out that social and motivational factors such as group pressure may inuence false memory in collaborative remembering. In a related sense, one possible cognitive explanation for the observed results is that collaboration might affect a reality-monitoring process, which refers to discrimination between memories of internally generated information and memories of externally derived information (Johnson & Raye, 1981; Johnson, Hashtroudi, & Lindsay, 1993). According to the reality monitoring account, it is assumed that participants falsely recall the critical non-presented words because of a failure to discriminate between those words and externally presented list words. Although speculative, it can be argued that people in the collaborative pairs might have to set up an optimal retrieval-strategy whenever it is disrupted by the other members recall. This set-up process could induce participants to focus more attention on their memories and thus evaluate them more carefully. As a result, participants would adopt a strict monitoring criterion. Such a criterion shift would result in a decrease of false memories in collaborative remembering. However, as previously mentioned, if the participants felt group pressure in a situation such as a turn-taking recall procedure (Basden et al., 1998), they might adopt a looser monitoring criterion, thereby producing more false memories. Of course, this explanation is quite speculative but may be an interesting direction for future research. In sum, we do need both cognitive and social factors to understand fully the phenomenon of false memories in collaborative remembering.

Acknowledgements
This research was partly supported by a Grant-in-Aid for Japanese Scientic Research (No.10610141), from the Ministry of Education, Science and Culture. Portions of this research were presented at the 3rd International Conference of Memory, Valencia, July 2001. I am grateful to Henry Roediger and Maryanne Garry for reading and commenting on an earlier draft of this paper. I thank David Payne, Babara Basden and Jan Andersson for their insightful comments and suggestions. I also thank Yuki Kunisada for her assistance with data collections and analyses.

References
Anderson, K. J., & Leaper, C. (1998). Meta-analyses of gender effects on conversational interruption: Who, what, when, where, and how. Sex Roles, 39, 225252. Andersson, J., & Ro nnberg, J. (1995). Recall suffers from collaboration: Joint recall effects of friendship and task complexity. Applied Cognitive Psychology, 9, 199211.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

12 Masanobu Takahashi Andersson, J., & Ro nnberg, J. (1997). Cued memory collaboration: Effects of friendship and type of retrieval cue. European Journal of Cognitive Psychology, 9, 273287. Asch, S. E., & Ebenholtz, S. M. (1962). The process of free recall: Evidence for non-associative factors in acquisition and retention. Journal of Psychology, 54, 331. Basden, B. H., Basden, D. R., Bryner, S., & Thomas, R. L., III. (1997). A comparison of group and individual remembering: Does collaboration disrupt retrieval strategies? Journal of Experimental Psychology: Learning Memory, and Cognition, 23, 11761189. Basden, B. H., Basden, D. R., & Henry, S. (2000). Costs and benets of collaborative remembering. Applied Cognitive Psychology, 14, 497507. Basden, B. H., Basden, D. R., Thomas, R. L., III, & Souphasith, S. (1998). Memory distortion in group recall. Current Psychology, 16, 225246. Basden, D. R., & Basden, B. H. (1995). Some tests of the strategy disruption interpretation of partlist cuing inhibition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 16561669. Ba uml, K.-L., & Kuhbandner, C. (2003). Retrieval-induced forgetting and part-list cuing in associatively structured lists. Memory and Cognition, 31, 11881197. Clark, N. K., & Stephenson, G. M. (1989). Group remembering. In P. B. Pauls (Ed.), Psychology of group inuence (2nd ed., pp. 357391). Hillsdale, NJ: Erlbaum. Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58, 1722. Edwards, D., & Middleton, D. (1986). Joint remembering: Constructing an account of shared experience through conversational discourse. Discourse Processes, 9, 423459. Finlay, F., Hitch, G. J., & Meudell, P. R. (2000). Mutual inhibition in collaborative recall: Evidence for a retrieval-based account. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 15561567. Hartwick, J., Sheppard, B. H., & Davis, J. H. (1982). Group remembering: Research and implications. In R. A. Guzzo (Ed.), Improving group decision making in organizations (pp. 4172). New York: Academic Press. Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitoring. Psychological Bulletin, 114, 328. Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Review, 88, 6785. Kimball, D. R., & Bjork, R. A. (2002). Inuences of intentional and unintentional forgetting on false memories. Journal of Experimental Psychology: General, 131, 116130. , B., & Darley, J. M. (1970). The unresponsive bystander: Why doesnt he help? New York: Latane Appleton-Century-Crofts. , B., Williams, K., & Harkins, S. (1979). Many hands make light the work: The causes and Latane consequences of social loang. Journal of Personality and Social Psychology, 37, 822832. Marsh, E. J., McDermott, K. B., & Roediger, H. L., III. (2004). Does test-induced priming play a role in the creation of false memories? Memory, 12, 4455. Meudell, P. R., Hitch, G. J., & Boyle, M. M. (1995). Collaboration in recall: Do pairs of people crosscue each other to produce new memories? Quarterly Journal of Experimental Psychology, 48A, 141152. Meudell, P. R., Hitch, G. J., & Kirby, P. (1992). Are two heads better than one? Experimental investigations of the social facilitation of memory. Applied Cognitive Psychology, 6, 525543. Miyaji, Y., & Yama, H. (2002). Making Japanese list which induce false memory at high probability for the DRM paradigm. Japanese Journal of Psychonomic Science, 21, 2126. (In Japanese with English summary). Nickerson, R. S. (1984). Retrieval inhibition from part-set cuing: A persisting enigma in memory research. Memory and Cognition, 12, 531552. Perlmutter, H. V., & de Montmollin, G. (1952). Group learning of nonsense syllables. Journal of Abnormal and Social Psychology, 47, 762769. Reysen, M. B., & Nairne, J. S. (2002). Part-set cuing of false memories. Psychonomic Bulletin and Review, 9, 389393.

Copyright The British Psychological Society


Reproduction in any form (including the internet) is prohibited without prior permission from the Society

Collaborative remembering and false memories

13

Robinson, K. J., & Roediger, H. L., III. (1997). Associative processes in false recall and false recognition. Psychological Science, 8, 231237. Roediger, H. L., III., Balota, D. A., & Watson, J. M. (2001). Spreading activation and arousal of false memories. In H. L. Roediger III., J. S. Nairne, I. Neath, & A. M. Surprenant (Eds.), The nature of remembering: Essays in honor of Robert G. Crowder (pp. 95115). Washington, DC: American Psychological Association. Roediger, H. L., III., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803814. Roediger, H. L., III., McDermott, K. B., & Robinson, K. J. (1998). The role of associative processes in creating false memories. In M. A. Conway, S. E. Gathercole, & C. Cornoldi (Eds.), Theories of memory (Vol. 2, pp. 187245). Hove, UK: Psychology Press. Roediger, H. L., III., & Neely, J. H. (1982). Retrieval blocks in episodic and semantic memory. Canadian Journal of Psychology, 36, 213242. Roenker, D. L., Thompson, C. P., & Brown, S. C. (1971). Comparison of measures for the estimation of clustering in free recall. Psychological Bulletin, 76, 4548. Ross, M., Spencer, S. J., Linardatos, L., Lam, K. C. H., & Perunovic, M. (2004). Going shopping and identifying landmarks: Does collaboration improve older peoples memory? Applied Cognitive Psychology, 18, 683696. Smith, S. M., Glenberg, A. M., & Bjork, R. A. (1978). Environmental context and human memory. Memory and Cognition, 6, 342353. Takahashi, M., & Saito, S. (2004). Does test delay eliminate collaborative inhibition? Memory, 12, 722731. Vollrath, D. A., Sheppard, B. H., Hinsz, V. B., & Davis, J. H. (1989). Memory performance by decision-making groups and individuals. Organizational Behavior and Human Decision Processes, 43, 289300. Wegner, D. M., Erber, R., & Raymond, P. (1991). Transactive memory in close relationships. Journal of Personality and Social Psychology, 61, 923929. Weldon, M. S. (2000). Remembering as a social process. In D. L. Medin (Ed.), The psychology of learning and motivation: Advance in research and theory (Vol. 40, pp. 67120). San Diego, CA: Academic Press. Weldon, M. S., & Bellinger, K. D. (1997). Collective memory: Collaborative and individual processes in remembering. Journal of Experimental Psychology: Learning, Memory, and Cognition, 23, 11601175. Weldon, M. S., Blair, C., & Huebsch, P. D. (2000). Group remembering: Does social loang underlie collaborative inhibition? Journal of Experimental Psychology: Learning Memory, and Cognition, 26, 15681577. Yuker, H. E. (1955). Group atmosphere and memory. Journal of Abnormal and Social Psychology, 51, 1723. Received 17 January 2004; revised version received 21 January 2006

You might also like