The Psychological Record , 2001 , 51 , 207-221

DIRECTED FORGETTING IN EXPLICIT AND IMPLICIT MEMORY: THE ROLE OF ENCODING AND RETRIEVAL MECHANISMS

DAVID E. FLECK University of Cincinnati

DANIEL B. BERCH National Institutes of Health

PAULA K. SHEAR and STEPHEN M. STRAKOWSKI University of Cincinnati

Despite the fact that directed-forgetting effects have been attributed to either retrieval inhibition or selective encoding, there has been no compelling evidence to suggest that either mechanism regulates performance in both implicit and explicit memory. Therefore, in two experiments we sought (a) to determine whether directed forgetting influences tests of implicit (lexical decision) and explicit (recognition) memory and (b) to examine the relative contributions of the encoding and retrieval mechanisms thought to mediate directed forgetting by having participants perform an external interference task (i.e. , sequential finger tapping) at either encoding or retrieval. In Experiment 1, directed-forgetting effects were demonstrated by better performance on remember-cued than on forget-cued words for both lexical decision and recognition. In Experiment 2, external interference disrupted directed forgetting in lexical decision when it occurred at retrieval and in recognition at encoding. These results demonstrate that although directed forgetting occurs on both implicit and explicit tests, it may be independently regulated by differential retrieval on the former and selective encoding on the latter. The discussion focuses on the differential excitation of remember- and forget-cued word representations, as inhibitory processing seemed not to account for directed-forgetting effects in either implicit or explicit memory.

Although cognitive psychologists have long been interested in how people remember information deemed relevant, the ability to adequately forget irrelevant information is also important for optimal memory function. Forgetting prevents irrelevant information from intruding on memory, updates memory, and facilitates performance (Harnishfeger, 1995). Moreover, it mayor may not require willful intent (Johnson, 1994). For example, to glean relevant information during verbal comprehension, comprehenders will engage in intentional forgetting by selectively ignoring irrelevant information (Zacks & Hasher, 1994) and unintentional

This article was supported in part by a University Research Council Award from the University of Cincinnati to David E. Fleck.

Address correspondence and requests for reprints to David E. Fleck, Department of Psychiatry, University of Cincinnati, P. O. Box 670559, Cincinnati, OH 45267-0559. Phone: (513) 558-4940; Fax: (513) 558-3399. (E-mail : FleckDE@email.uc.edu).

208 FLECK ET AL.

forgetting through the suppression of ambiguous word meanings (Gernsbacher, Varner, & Faust, 1990). The ability to forget is adaptive in that (a) information known to be irrelevant is not encoded into memory and (b) irrelevant information does not interfere with the retrieval of relevant information already stored in memory.

One approach to studying forgetting ability is the directed-forgetting paradigm (Bjork, 1970). Directed-forgetting research explores the ability to comply with an overt instruction to forget specific information. In the word-by-word variant of the directed-forgetting procedure, a list of words is presented one word at a time and each is followed by a cue either to remember or forget. 1 This type of stimulus presentation ensures that both remember-cued (R-cued) words and forget-cued (F-cued) words are initially perceived prior to memory testing. Directed forgetting is demonstrated by poorer memory performance for F-cued words, as compared with R-cued words, on either implicit or explicit memory tests.

A distinction between implicit and explicit memory effects is based on the type of test administered at the conclusion of the directed-forgetting study list. Implicit tests assess long-term memory that does not necessarily require intentional recollection or awareness of previously studied information (Roediger, Weldon , Stadler, & Riegler, 1992). In directed forgetting, implicit effects are thought to occur through an automatic reduction in the activation of F-cued item representations, without motivation or conscious recollection of the study list (MacLeod, 1989). In this case, participants are instructed to perform a task that is apparently unrelated to the study list, such as word-stem completion (Lehman, McKinley-Pace, Wilson , Siavsky, & Woodson, 1997), word­fragment completion (Basden, Basden~ & Gargano, 1993; MacLeod, 1989), or repetition priming in lexical decision (MacLeod, 1989).

In contrast, explicit tests assess long-term memory that necessarily requires intentional recollection and awareness of previously studied information (Roediger, Weldon, & Challis, 1989). In directed forgetting, explicit effects are thought to arise through a motivated attempt to reduce the probability of retrieving F-cued items in the future (Johnson, 1994). In this case, participants are instructed to recollect as many R- and F-cued items as possible, regardless of the initial cue associated with each item at study. Directed forgetting in explicit memory has been demonstrated for tests of free recall (Geiselman & Bagheri, 1985; Paller, 1990; Welzel & Hunt, 1977; Woodward & Bjork, 1971), cued recall (Paller, 1990), and recognition (Basden et al. 1993; MacLeod, 1989).

The exact nature of the processes necessary for directed forgetting on implicit and explicit tests is unclear, but inhibition of F-cued items at

1The word-method procedure can be contrasted with list-method directed forgetting in which the cueing may occur between many blocks of words, or half way between two lists of words (e.g., Harnishfeger & Pope, 1996; Macleod, 1975; Zacks, Radvansky, & Hasher, 1996). The present article primarily reviews word-method directed forgetting. The reader is referred to Bjork (1989) , Golding and Macleod (1998), and Johnson (1994) for comprehensive reviews of both list- and word-method directed forgetting.

DIRECTED FORGETTING 209

the time of retrieval is one possibility. Retrieval inhibition may account for forgetting through an automatic process that suppresses the previous activation of irrelevant mental representations (e.g., Harnishfeger, 1995), restricting retrieval from long-term memory. Although directed-forgetting effects on explicit memory tests have been attributed to retrieval inhibition (Bjork, 1989; Geiselman & Bagheri, 1985; Harnishfeger & Pope, 1996; Macleod, 1989), there has been comparatively little investigation of the role of retrieval inhibition in implicit tests. Macleod (1989) demonstrated directed-forgetting effects on two implicit memory tests, repetition priming in lexical decision and word-fragment completion , concluding t~at F-cued words are inhibited at the time of retrieval. Basden et al. (1993) contend that on word-fragment completion the effect occurs when participants become aware that fragments can be completed with words from the study list and adopt an explicit retrieval strategy. Thus, implicit tests may be functionally transformed into explicit tests.

An alternative account of directed forgetting is by the suppression of F-cued words during selective encoding. In the first demonstrations of selective encoding, words were maintained equally well until an F cue selectively prevented further processing, attenuating recall of F-cued words as compared with R-cued words (Woodward & Bjork, 1971). Subsequent research revealed that enhancement was necessary because giving participants a greater opportunity to rehearse words increased the recall of R-cued information relative to F-cued information, which remained unchanged (Wetzel, 1975; Wetzel & Hunt, 1977). Hasher and Zacks (1988) have since developed a framework that involves attention suppression, and have applied the process to explain directed­forgetting effects on explicit memory tests (Le., recall and recognition; Zacks & Hasher, 1994; Zacks, Radvansky, & Hasher, 1996). Attention suppression is a plausible explanation for directed forgetting in implicit memory as well, if participants do, in fact, adopt explicit retrieval strategies on implicit tests. The basic assumption is that forgetting is a controlled process that operates by suppressing the previous activation of irrelevant thoughts, restricting their access to working memory and conscious awareness.

Experiment 1

With the above considerations in mind, Experiment 1 was conducted to examine whether directed forgetting influences tests of both implicit (Le. , lexical decision) and explicit (Le., recognition) memory. Experiment 1 was designed as the first attempt to replicate Macleod's (1989) directed-forgetting effect for repetition priming in lexical decision, a reaction time (RT) test of implicit memory that reduces the likelihood of explicit retrieval. We reasoned that because lexical decisions require a speeded and nongenerative response, they are made without explicit retrieval. Even if subjects become aware that many words on a lexical­decision task were on the study list, an explicit retrieval strategy should

210 FLECK ET AL.

not benefit performance on rapid word/non-word decisions. In all likelihood, the additional time required to apply such a strategy would diminish performance. We predicted a directed-forgetting effect in implicit memory using the RT measure of repetition priming in lexical decision.

Experiment 1 was also intended to examine whether directed forgetting, during a recognition test, would affect RT and accuracy in explicit memory. We felt that a RT measure of recognition would be more comparable to the RT measure obtained for lexical decision, and it may be more sensitive to the processing deficiencies we intended to induce by external interference in Experiment 2. We predicted directed-forgetting effects in explicit memory using both RT and accuracy measures of recognition.

Method Participants. Forty-two undergraduates, 28 men and 14 women, from

the University of Cincinnati volunteered to take part in Experiment 1. All of the subjects were recruited from the Psychology Department's subject pool and received credit toward a course requirement.

Stimuli. Stimuli were presented via a microcomputer in white upper­case letters centered on a black background using the Micro Experimental Laboratory software (Schneider, 1988). A set of 64 words was selected from a.normative list of word meanings (Ferraro & Kellas, 1990) to comprise two separate directed-forgetting study lists of 32 words each. The words in each list were equated for frequency, degree of lexical ambiguity, and length. An R cue or an F cue was assigned to each word in equal numbers (i.e. , 16 of each type per two lists) and then reversed to form four study lists that were counterbalanced across sessions.

In addition to the 32 study-list words, two separate sets of 30 foil words were selected to comprise each memory test; one set for lexical decision and the other for recognition. However, both memory tests were comprised of the same set of target stimul i, which consisted of R- and F­cued words from the corresponding study list. A single set of 60 pseudowords (i.e. , pronounceable nonwords) was also selected for the lexical-decision task.

Procedure. Participants were tested in groups of 7 to 9 in a small computer laboratory. Initially, participants were assigned to computers and instructed that they would be shown a list of words, some associated with a prompt to remember and others associated with a prompt to forget. They were informed that they would subsequently be asked to recollect only "remember" words. Study-list trials began with an orienting prompt (******) lasting for 1 s, followed by a 0.5-s blank screen, then a word lasting 2 s, a remember (RRRRRR) or forget (FFFFFF) cue lasting 3 s, and another 0.5-s blank screen constituting the intertrial interval. Two buffer words were added to the beginning and end of the study list to reduce the effect of word order. The buffer words were always R-cued, and with this exception , all study-list stimuli were presented in a different random order to each participant.

Following the administration of the study list, the participants

DIRECTED FORGETTING 211

performed a lexical-decision task for which they were instructed to make word/nonword decisions, via a key press, as quickly and accurately as possible. They were also informed that the task was meant only to fill time prior to the memory test that would follow, to minimize explicit retrieval. The lexical-decision task consisted of 122 experimental stimuli (16 old R­cued words, 16 old F-cued words, 30 new foils, and 60 pseudowords). A trial began with a 1-s orienting prompt (******), followed by a 0.25-s blank screen, a word or pseudoword lasting until a key-press response, and a 1-s blank screen constituting the intertrial interval. The first four trials were considered practice, and the responses to these items (two words and two pseudowords) were not analyzed. Except for practice items, all experimental stimuli were presented to each participant in a different random order.

A recognition-memory test followed the lexical-decision task. Participants were instructed to decide whether each word was on the original study list, regardless of the initial instruction cue associated with each word. They were also instructed to respond with a key press as quickly and accurately as possible. Unlike typical recognition instructions that emphasize accuracy only, speed was also stressed so that reaction times could be used as an index of processing efficiency. The 62 words on the recognition test (16 old R-cued words, 16 old F-cued words, and 30 new foils) were presented in a different random order to each participant. The sequence of events for trials was the same as in the lexical-decision task.

lexical decision always preceded recognition. A test-type/order of presentation confound was not expected, based on Macleod's (1989) finding that order of presentation had no effect on either lexical decision or recognition performance. Therefore we tested lexical decision first, disguised as a filler task, to (a) reduce explicit retrieval strategies during the implicit lexical-decision task and (b) prevent greater repetition priming effects during lexical decision from additional exposure to target words during recognition testing (Scarborough , Gerard, & Cortese, 1979).

Results Lexical decision. The omnibus test for lexical decision consisted of a

one-way (stimulus type: R cued , F cued, new words) analysis of variance (ANOVA) conducted on RT. Pseudoword means were not entered because they were not pertinent to the predictions; however, the mean RT for pseudowords was 745 ms.

Stimulus type significantly influenced RT in lexical decision, F(2 , 82) = 20.77, MSE = 1689.15, P < .001 . Remember-cued words were verified the most rapidly (M= 616 ms, SO= 84 ms), followed by F-cued words (M = 640 ms, SO = 88 ms), and then by new words (M = 657 ms, SO = 83 ms). As predicted, planned t tests using the Bonferroni correction procedure revealed a significant directed-forgetting effect (R-cued < F­cued word RT), ~41) = 3.11 , P < .01. Moreover, significant positive repetition-priming effects were obtained for both R-cued words (R-cued <

212 FLECK ET AL.

new word RT), ~41) = 6.75, P < .001, and F-cued words (F-cued < new word RT) , ~41) = 3.15, P < .01 .

Recognition. The omnibus tests for recognition consisted of two one­way ANOVAs, one conducted on RT and one on proportion correct. New­word data were not analyzed because the data of interest were hit rates for R- and F-cued words (see Macleod, 1989). The RT and proportion­correct means for new words were 811 and .77, respectively.

Stimulus type significantly influenced both RT, F(1 , 41) = 9.88, MSE = 8798.38, P < .001 , and proportion correct, F(1 , 41) = 28.67, MSE = 197.35, P < .001, in recognition. Remember-cued words were recognized more rapidly (M = 719 ms, SO = 128 ms) and accurately (M = .78, SO = .19) than F-cued words (M = 765 ms, SO = 138 and M = .67, SO = .21, respectively). Therefore, not only was a directed-forgetting effect achieved with both RT and accuracy measures, but there was no evidence of a speed/accuracy trade-off.

Discussion Experiment 1 demonstrated directed-forgetting effects for both

implicit and explicit memory tests. Although directed-forgetting effects are common on recognition tests (Johnson, 1994; Macleod, 1975), these results are the first to replicate a directed-forgetting effect in lexical decision, an implicit memory task that uses RT as the dependent measure (Macleod, 1989). Because lexical decision requires a speeded and nongenerative response, an explicit retrieval explanation of directed forgetting seems tenuous (Basden et aI., 1993; lehman et aI., 1997). Even if participants were aware that test words were on the study list, an explicit retrieval strategy should not benefit lexical-decision performance because (a) response speed would preclude its use and (b) conscious recollection of the study list is not necessary for proper lexical-decision performance. Therefore, although explicit retrieval strategies are involved in recognition, they are not likely responsible for the directed-forgetting effect in lexical decision.

In replicating Macleod (1989), Experiment 1 rules out the fundamental concern that the directed-forgetting effect in lexical decision can be established through mental operations identical to those in recognition. However, two additional questions pertinent to directed­forgetting theory are left unanswered. First, what are the underlying mechanisms responsible for the differential processing of R- and F-cued words in implicit and explicit memory, and are they distinct? Second, are inhibitory concepts such as retrieval inhibition (Bjork, 1989) and attention suppression (Zacks & Hasher, 1994) necessary to explain directed­forgetting performance?

Experiment 2

Experiment 2 was designed to examined the relative contribution of retrieval and encoding mechanisms thought to mediate directed

DIRECTED FORGETTING 213

forgetting by having participants perform an external interference task (i.e., sequential finger tapping) at either encoding or retrieval. The rationale behind Experiment 2 was that interference would disrupt directed-forgetting performance, but only during the stage at which primary task processing was ongoing. We reasoned that, because there is no compelling evidence to suggest that a single mechanism mediates directed forgetting in both memory systems, the mechanisms might indeed be distinct. A directed-forgetting procedure was employed that required participants to perform an external interference task during the study list and/or memory tests. For conditions in which there was no interference, directed-forgetting effects were predicted for both lexical decision and recognition, as demonstrated in Experiment 1. We reasoned that if retrieval inhibition mediates directed-forgetting performance on either task, then the directed-forgetting effect should be attenuated by interference at retrieval. If selective encoding is involved, then the directed-forgetting effect should be attenuated by interference at encoding. Finally, if directed forgetting depends on efficient processing by both mechanisms, then interference at both encoding and retrieval would be expected to disrupt the directed-forgetting effect.

Experiment 2 was also designed to examine the importance of inhibitory influences on directed-forgetting performance. We reasoned that if retrieval inhibition mediates directed forgetting in lexical decision (Macleod, 1989), then decision times on F-cued words should be slower than decision times on new words in the absence of interference (i.e., negative priming) .2 However, this ~ffect should be attenuated by external interference at retrieval. In addition, R-cued words should be faster than new words in the absence of interference (i.e., positive priming) and this effect too should be reduced by interference at retrieval.

Alternatively, if attention suppression mediates directed forgetting in recognition (Zacks & Hasher, 1994), then external interference at encoding should benefit performance on F-cued words, relative to a no interference condition , by diverting attentional resources from the process of suppressing irrelevant information. In addition, performance on R-cued words should be diminished by interference during encoding relative to a no interference condition.

Method Participants. Ninety-three undergraduates at the University of Cincinnati,

who had not participated in Experiment 1 , were recruited from the Psychology Department's subject pool and received course credit for participating in Experiment 2. Thirteen participants were later excluded, 1 who was physically unable to comply with the task demands, 1 who was hearing impaired, and 11 who did not perform the extemalinterferencetasktocriterion.This brought the number of participants to 80 (34 men, 46 women).

2This was not the case in Experiment 1, although the positive priming of F-cued words was small enough in magnitude, compared to that for R-cued words, that we felt it was important to assess negative priming in Experiment 2.

214 FLECK ET AL.

Stimuli. The computer apparatus and stimulus lists were the same as in Experiment 1.

Procedure. The procedure for Experiment 2 was similar to Experiment 1 except for the addition of an external interference task used to divert attention from the primary task. Sequential finger tapping was employed as the interference task for two reasons. First, it requires the executive functions of sequencing and planning that consume attention and interfere with other strategic processing (Moscovitch, 1994). Second, it competes for the same manual response mechanism as the key press required for the primary task, thereby interfering with automatic processes of retrieval (Posner & Snyder, 1975).

Ten groups averaging 8 participants each (range 6 to 10) were randomly assigned to one of five external interference conditions (Le., two groups and 16 participants per interference condition) and tested in a small computer laboratory. Thus, all participants within each group were assigned to the same interference condition. The five external interference conditions included sequential finger tapping at (a) encoding, (b) retrieval in lexical decision, (c) retrieval in recognition, (d) both encoding and retrieval , or (e) not at all. Note that separate retrieval interference conditions for lexical decision and recognition resulted in four interference conditions per memory task.

During those portions of the procedure that required finger tapping, the investigator and an assistant monitored the accuracy with which participants tapped. Each participant was monitored for sequence and timing accuracy on two consecutive tapping sequences at several discrete intervals, and each monitored interval was scored on a pass/fail basis. Those individuals who did not meet an 80% accuracy criterion were dropped from subsequent analyses.

Before being presented with the study list, participants practiced tapping with the fingers of their dominant hands in the following order: index, ring, middle, small. They were required to perform to the beat of a metronome, at a rate of 90 taps per minute. After 3 minutes of practice they were administered a lexical-decision task with 60 trials. This task was used for fu rther practice and introduced participants to the dual-task procedure. None of the stimuli from the practice trials were used in the experimental trials. Upon completing the practice portion, participants were administered a study list, a lexical-decision task, and a recognition test as described in Experiment 1.

Results Lexical decision. The omnibus test for lexical decision consisted of a

4 (interference condition: none, continuous, encoding, retrieval) x 3 (stimulus type: R cued, F cued, new words) mixed ANOVA conducted on RT. Pseudoword means were not entered because they were not pertinent to the predictions. Mean RT for pseudowords was 740, 755, 893, and 990 ms for the no interference, encoding, retrieval, and continuous interference condition , respectively. In order to examine

DIRECTED FORGETTING 215

directly whether external interference differentially influenced directed forgetting at encoding and retrieval , a 2 (encoding vs. retrieval interference conditions) x 2 (R-cued vs. F-cued words) mixed ANDVA was also conducted.

Table 1 presents mean RT for the lexical-decision task. Main effects of interference condition , F(3, 60) = 9.02, MSE = 54524.46, P < .001 , and stimulus type, F(2 , 120) = 15.76, MSE = 1928.68, P < .001 , were qualified by an interaction between the two factors, F(6, 120) = 2.32, MSE = 1928.68, P < .05. The direct encoding/retrieval comparison revealed a main effect of interference condition , F(1 , 30) = 4.84, MSE = 45318.72, P < .05, that was qualified by an interaction between interference condition and stimulus type, F(1, 30) = 4.53, MSE = 1883.05, P < .05.

Planned t tests using a nonterminating Holm sequential Bonferroni procedure (Seaman, Levin, & Serlin, 1991) revealed directed-forgetting effects (R-cued < F-cued word RT) in the no interference condition, ~15) = 2.70, P < .02, and encoding interference condition , ~15) = 2.66, P < .02. In other words, external interference disrupted directed forgetting in lexical decision when it occurred at retrieval , implicating a retrieval mechanism. Additionally, both R- and F-cued words were positively primed (R-cued and F-cued < new word RT) in the no interference condition , ~15) = 5.17, P < .001 , and ~15) = 2.84, P < .02, respectively. However, negative priming (F-cued > new word RT) did not occur in either the no interference condition or retrieval interference condition , ps > .05, possibly indicating that inhibition does not underlie the retrieval mechanism responsible for directed forgetting in lexical decision.

Table 1

Mean (and Standard Deviation) Reaction Time in Milliseconds for Lexical Decision as a Function of Interference Condition and Stimulus Type

Stimulus Type Condition R cued F cued New

No interference Encoding Retrieval Continuous

627 (72) 643 (80) 783 (208) 802 (188)

649 (61) 678 (104) 772 (185) 856 (201)

672 (66) 680 (93) 825 (221) 839 (202)

Note. R cued = old remember-cued words; F cued = old forget-cued words; New = foil words.

Recognition. The omnibus tests for recognition consisted of 4 x 2 mixed ANOVAs conducted on both RT and proportion correct. Unlike lexical decision, new-word data were not analyzed because the data of interest were hit rates. The RT (and proportion correct) means for new words were 672 ms (.89), 680 ms (.91), 825 ms (.87), and 839 ms (.88) , for the no interference, encoding, retrieval , and continuous interference condition, respectively. An additional two ANOVAs were conducted, one for RT and the other for proportion correct, that focused on the encoding and retrieval interference conditions, as in lexical decision.

Table 2 displays mean RT and proportion correct data for the

216 FLECK ET AL.

recognition test. For the RT data, there were main effects of interference condition, F(3, 60) = 5.34, MSE = 92851.55, P < .005, and stimulus type, F(1 , 60) = 22.48, MSE = 6757.76, P < .001, as well as a nonsignificant trend toward an interaction of the two factors, F(3, 60) = 2.40, MSE = 6757.76 , P = .076. The direct encoding/retrieval comparison revealed main effects of interference condition, F(1, 30) = 8.09, MSE = 92969.07, P < .01, and stimulus type, F(1, 30) = 9.57, MSE = 8039.85, P < .005, and a marginally significant interaction between the two factors, F(1, 30) = '4.04, MSE = 8039.85, P < .054.

Planned t tests revealed directed-forgetting effects (R-cued < F-cued RT) in the no interference condition, ~15) = 3.70, P < .005, and retrieval interference condition, ~15) = 3.05, P < .01. In other words, external interference disrupted directed forgetting in recognition when it occurred at encoding, implicating an encoding mechanism. Additionally, R-cued words were not recognized more rapidly, and F-cued words were not recognized more slowly, in the no interference condition as compared with the encoding interference condition, ps > .05.

The overall analysis of proportion correct revealed a main effect of stimulus type, F(1, 60) = 56.29, MSE = 138.79, P < .001 , and in the direct encoding/retrieval comparison , stimulus type was again significant, F(1 , 30) = 24.51 , MSE = 163.37, P < .001 . Stimulus type did not interact with interference condition in either analysis.

Planned t tests demonstrated that directed-forgetting effects (R-cued > F-cued word accuracy) occurred in the no interference, continuous, encoding, and retrieval conditions, ~15) = 4.43, P < .001, ~15) = 3.94, P < .002, ~15) = 5.40, P < .001, and ~15) = 2.69, P < .02, respectively. Therefore, despite the differential influence of external interference on RT, it had no differential influence on levels of recognition accuracy. Recognition accuracy was greater in the no interference condition than ir.l the encoding interference condition for both R- and F-cued words, ~30) =

Table 2

Mean (and Standard Deviation) Reaction Time in Milliseconds and Mean Proportion Correct for Recognition as a Function of Interference Condition and Stimulus Type

Stimulus Type Condition R cued F cued

No interference Reaction time 721 (115) 821 (165) Proportion correct .89 (.16) .75 (.20)

Encoding Reaction time 787 (140) 812 (129) Proportion correct .76 (.15) .57 (.17)

Retrieval Reaction time 959 (252) 1074 (320) Proportion correct .85 (.19) .72 (.23)

Continuous Reaction time 962 (288) 999 (271) Proportion correct .82 (.18) .65 (.16)

Note. R cued;;:: remember-cued words; F cued = forget-cued words.

DIRECTED FORGETTING 217

2.32, P < ~05 , and ~30) = 2.69, P < .05, respectively, possibly indicating that attention suppression does not underlie the encoding mechanism responsible for directed forgetting in lexical decision.

Discussion In lexical decision, external interference at encoding and retrieval had

different influences on directed-forgetting processing speed. There was a significant directed-forgetting effect for tapping at encoding that was attenuated for tapping at retrieval. Moreover, continuous tapping interfered with the differential activation of R- and F-cued words found in the no interference condition, resulting in slow performance.3 These findings are consistent with the idea that directed forgetting in implicit memory is mediated by a retrieval mechanism, because external interference disrupted processing at retrieval in lexical decision.

Evidence against the view that retrieval inhibition underlies directed forgetting in implicit memory is marshaled from the finding that interference only disrupted positive priming of R-cued words at retrieval. The fact that F-cued words were not negatively primed in the no interference condition, but instead positively primed (consistent with Experiment 1), suggests that retrieval inhibition of irrelevant information may not be a necessary component of directed forgetting in implicit memory. Instead, directed forgetting may result from the differential excitation of R- and F-cued word representations at retrieval.

Although lexical decision and recognition performance was not directly compared for the present study, both tests assessed the same previously studied stimuli using RT. Unlike the pattern of directed-forgetting effects in lexical decision, recognition RT suggested an encoding deficit. In recognition, directed-forgetting speed was again significant in the no interference condition, but not in the continuous interference condition. However, the differential influence of interference on directed forgetting at encoding and retrieval was the reverse of that found for lexical decision. There was a significant directed-forgetting effect when tapping occurred at retrieval that was eliminated when tapping occurred at encoding. These results are consistent with the assumption that directed forgetting in recognition memory is mediated by a selective encoding mechanism, because the differential processing of relevant and irrelevant information was disrupted by interference at encoding in recognition.

Evidence against the view that attention suppression underlies the selective encoding of R- and F-cued words in recognition derives from the finding that encoding interference did not benefit performance on F-cued words relative to the no interference condition. If a suppression process was involved, encoding interference might be expected to release attentional resources from the process of suppressing irrelevant

3The magnitude of the directed-forgetting effect is actually larger in the continuous interference condition than in the no interference condition, but the effect is not significant (see Table 1). This effect may simply reflect increased ,variability or added difficulty leading to greater

218 FLECK ET AL.

information as predicted (e.g., Wetzel, 1975; Wetzel & Hunt, 1977), thereby improving the recognition of F-cued words. Instead, recognition accuracy for R- and F-cued words was diminished by interference relative to the no interference condition , as would be expected if the differential enhancement of relevant and irrelevant information were involved in directed forgetting.

General Discussion

Experiment 1 represents the first replication of a directed-forgetting effect for RT in implicit memory, namely repetition priming in lexical decision (Macleod, 1989). The results of Experiment 1 also serve as one of the first demonstrations of a directed-forgetting effect for RT in recognition , although the effect has been well documented for accuracy measures in explicit memory (Johnson, 1994; Macleod, 1975). Moreover, despite the influence of interference on recognition speed in Experiment 2, the directed-forgetting accuracy effect for recognition was not disrupted in any condition.

These results demonstrate that RT measures of directed forgetting are more sensitive to interference effects than are accuracy measures and may represent a methodological advance. This knowledge may prove useful to investigators who are concerned with how certain functional and structural disorders, such as obsessive compulsive disorder (Wilhelm, McNally, Baer, & Florin, 1996), posttraumatic stress disorder (McNally, Clancy, Metzger, lasko, & Pitman, 1998), and epilepsy (Fleck, Berch, Shear, Schefft, Privitera, & Strakowski, 1999), disrupt directed-forgetting performance. The practice of reporting RT along with accuracy may further an understanding of forgetting processes in recognition, especially considering the differential performance on RT in the present study.

The present results also shed light on a number of potential interpretations of directed-forgetting effects. Previous investigations have yielded conflicting results regarding the extent to which word method directed-forgetting influences explicit and implicit memory tests. On the one hand Paller (1990) only obtained a directed-forgetting effect for an explicit test while, on the other hand, Macleod (1989) obtained effects for both explicit and implicit tests. Moreover, directed-forgetting effects, or lack thereof, have generally been interpreted as resulting from either encoding (Basden ef aI., 1993; Paller, 1990) or retrieval mechanisms (Bjork, 1989; Macleod, 1989) in both implicit and explicit memory. The present results not only demonstrate that word method directed-forgetting influences both implicit and explicit long-term memory, but suggest further that directed forgetting in lexical decision is most strongly reflective of automatic regulation at retrieval , while directed forgetting in recognition RT may be most influenced by intentional regulation at encoding. Thus, directed forgetting may be more accurately viewed as an ability that relies on distinct processing mechanisms in different memory systems, at least

DIRECTED FORGETTING 219

when considering the tests employed herein. Previous accounts of directed forgetting in implicit memory based on selective encoding may stem from the use of implicit tasks that can be performed by an explicit retrieval strategy (i.e., word-stem and word-fragment completion tasks). However, true tests of implicit memory that can not be completed by explicit retrieval , such as speeded repetition priming in lexical decision, may be dependent on differential retrieval.

A second interpretation with respect to directed-forgetting results has been that inhibitory processes account for the relatively poor performance on F-cued trials relative to R-cued trials. Although R- and F-cued information seemed to undergo differential processing in the present study, concepts of suppression (Zacks & Hasher, 1994) and inhibition (Bjork, 1989; Macleod, 1989) were not necessary to explain the results. If attention suppression were involved, then one might expect finger tapping to interfere with its functioning, thereby increasing performance on F-cued words relative to similar stimuli in the no interference condition. Moreover, if retrieval inhibition were involved one might expect slower reaction times on F-cued words relative to new, foil words, because of the inhibitory influence of negative priming. These patterns were not detected, which leads us to conclude that the concept of differential excitation, albeit at different stages of information processing, may be sufficient to explain directed-forgetting effects on lexical decision and recognition memory tests.

Further research is needed to determine more fully how different tests are affected by directed-forgetting manipulations and under what conditions. It could be argued that results similar to those in Experiment 2 might be obtained by comparing two explicit or two implicit tests. Replication using a variety of explicit and implicit measures seems warranted; however, traditional measures may not be fully adequate. For example, in directed-forgetting paradigms, free recall is performed at near floor levels when explicit memory is disrupted by either a secondary task in healthy volunteers (Fleck & Berch, 1996) or brain damage in neurological patients (Fleck et aI. , 1999). In addition, nongenerative RT measures other than lexical decision would need to be developed to test implicit memory, because traditional word-stem and word-fragment completion measures do not reduce the potential for explicit retrieval.

References

BASDEN, B. H., BASDEN, D. R., & GARGANO, G. J. (1993). Directed forgetting in implicit and explicit memory tests: A comparison of methods. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 603-616.

BJORK, R. A. (1970). Positive forgetting: The non-interference of items intentionally forgotten. Journal of Verbal Learning and Verbal Behavior, 9, 255-268.

220 FLECK ET AL.

BJORK, R. A. (1989). Retrieval inhibition as an adaptive mechanism in human memory. In H. L. Roediger, III & F. I. M. Craik (Eds.), Varieties of memory and consciousness: Essays in honour of Endel Tulving (pp. 309-330). Hillsdale, NJ : Lawrence Erlbaum.

EYSENCK, M. w. , & CALVO, M. G. (1992). Anxiety and performance: The processing efficiency theory. Cognition and Emotion, 6, 409-434.

FERRARO, F. R., & KELLAS, G. (1990). Normative data for number of word meanings. Behavior Research Methods, Instruments, & Computers, 22, 491-498.

FLECK, D. E. , & BERCH, D. B. (1996). Differential time course of cognitive inhibition in implicit and explicit memory. Poster presented at the 37th Annual Meeting of the Psychonomic Society, Chicago.

FLECK, D. E., BERCH, D. B. , SHEAR, P. K., SCHEFFT, B. K., PRIVITERA, M. D., & STRAKOWSKI, S. M. (1999). Directed forgetting deficits in patients with temporal lobe epilepsy: An information processing perspective. Journal of the International Neuropsychological Society, 5, 543-549.

GEISELMAN, R. E. , & BAGHERI, B. (1985). Repetition effects in directed forgetting: Evidence for retrieval inhibition. Memory and Cognition, 13, 57-62.

GERNSBACHER, M. A., VARNER, K. R., & FAUST, M. (1990). Investigating differences in general comprehension skill. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 430-445.

GOLDING, J. M. , & MACLEOD, C. M. (Eds.) (1998) . Intentional forgetting: Interdisciplinary approaches. Mahwah, NJ: Lawrence Erlbaum Associates .

HARNISHFEGER, K. K. (1995). The development of cognitive inhibition: Theories, definitions, and research evidence. In F. N. Dempster & C. J. Brainerd (Eds.) , Interference and inhibition in cognition (pp. 175-204). San Diego, CA: Academic Press.

HARNISH FEGER, K. K., & POPE, R. S. (1996) . Intending to forget: The development of cognitive inhibition in directed forgetting. Journal of Experimental Child Psychology, 62, 292-315.

HASHER, L. , & ZACKS, R. T. (1988). Working memory, comprehension, and aging: A review and a new view. In G. H. Bower (Ed.), The psychology of learning and motivation (Vol. 22, pp. 193-225). San Diego, CA: Academic Press.

JOHNSON , H. M. (1994). Processes of successful intentional forgetting . Psychological Bulletin, 116, 274-292.

LEHMAN, E. B. , MCKINLEY-PACE, M. J., WILSON, J. A., SLAVSKY, M. D., & WOODSON, M. E. (1997) . Direct and indirect measures of intentional forgetting in chi ldren and adults: Evidence for retrieval inhibition and reinstatement. Journal of Experimental Child Psychology, 64, 295-316.

MACLEOD, C. M. (1975). Long term recognition and recall following directed forgetting. Journal of Experimental Psychology: Learning, Memory, and Cognition, 1, 271-279.

MACLEOD, C. M. (1989). Directed forgetting affects both direct and indirect tests of memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 13-21.

MCNALLY, R. J., CLANCY, S. A. , METZGER, L. J., LASKO, N. B., & PITMAN, R. K. (1998). Directed forgetting of trauma cues in adult survivors of childhood sexual abuse with and without posttraumatic stress disorder. Journal of Abnormal Psychology, 107, 596-601.

MOSCOVITCH, M. (1994). Cognitive resources and dual-task interference effects at retrieval in normal people: The role of the frontal lobes and medial temporal cortex. Neuropsychology, 8, 524-534.

DIRECTED FORGETTING 221

PALLER, K. A. (1990). Recall and stem-completion priming have different electrophysiological correlates and are modified differentially by directed forgetting . Journal of Experimental Psychology: Learning, Memory, and Cognition, 6, 1021-1032.

POSNER, M. I. , & SNYDER, C. R. R. (1975). Facilitation and inhibition in the processing of signals. In P. M. A. Rabbitt & S. Dornic (Eds.), Attention and Performance V (pp. 669-682) . New York: Academic Press.

ROEDIGER, H. L. , III , WELDON , M. S., & CHALLIS, B. H. (1989). Explaining dissociations between implicit and explicit measures of retention : A processing account. In H. L. Roediger III & F. I. M. Craik (Eds.), Varieties of memory and consciousness (pp. 3-41). Hillsdale, NJ : Erlbaum.

ROEDIGER, H. L. , III , WELDON , M. S., STADLER, M. L., & RIEGLER, G. L. (1992). Direct comparison of two implicit memory tests: word fragment and word stem completion. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 1251-1269.

SCARBOROUGH, D. L. , GERARD, L. , & CORTESE, C. (1979). Accessing lexical memory: The transfer of word repetition effects across task and modality. Memory and Cognition, 7, 3-12.

SCHNEIDER, W. (1988). Micro Experimental Laboratory: An integrated system for IBM PC compatibles. Behavior Research Methods, Instruments, and Computers, 20, 206-217.

SEAMAN, M. A , LEVIN , J. R. , & SERLIN, R. C. (1991). New developments in pairwise multiple comparisons: Some powerful and practicable procedures. Psychological Bulletin, 110, 577-586.

WETZEL, C. D. (1975). Effect of orienting tasks and cue timing on the free recall of remember- and forget-cued words. Journal of Experimental Psychology: Human Learning and Memory, 104, 556-566.

WETZEL, C. D., & HUNT, R. E. (1977). Cue delay and the role of rehearsal in directed forgetting. Journal of Experimental Psychology: Human Learning and Memory, 3, 233-245.

WILHELM, S., MCNALLY, R. J., BAER, L., & FLORIN, I. (1996). Directed forgetting in obsessive-compulsive disorder. Behavioral Research and Theory, 34, 633-641.

WOODWARD, A E., & BJORK, R. A (1971). Forgetting and remembering in free recall : Intentional and unintentional. Journal of Experimental Psychology, 89, 109-116.

ZACKS, R. T. , & HASHER, L. (1994). Directed ignoring: Inhibitory regulation of working memory. In D. Dagenbach & T. H. Carr (Eds.) , Inhibitory processing in attention, memory, and language (pp. 241 -264). San Diego, CA: Academic Press.

ZACKS, R. T. , RADVANSKY, G., & HASHER, L. (1996). Studies of directed forgetting in older adults. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22, 143-156.

222 FLECK ET Al.