Past Research

Within the past several years, research has indicated that people can be led to believe that fictitious events occurred by having them create the false memory in their mind through repeated imagery. For example, Hyman and Pentland (1996) used the fictitious event of spilling a punchbowl at a wedding as a fictitious event. Research has shown that people have difficulty distinguishing between reality and the imagination since the false events include characteristics that would be typical if the event were actually experienced (Johnson et al, 1988). When a person is asked to elaborate on the fictitious or suggested event, such as who was at the wedding or how the person felt when the incident happened, the suggested event becomes more plausible and more familiar to the person, thus confusing which events actually happened and those that are imagined.

Overview

Drivdahl and Zaragoza were interested in whether or not pressing witnesses to elaborate on perceptual characteristics and provide additional details led to false memory for alleged events or objects. They suggested that the follow-up questions used to elicit more information for an investigation encouraged witnesses to embellish mental images, which should result in an increase of false memories. They did, however, acknowledge that an alternative possibility might be that the participants will be more critical in evaluating their memory to distinguish between accurate and false information, which might lead to their hesitation in believing fictitious suggestions.

Two objectives of the study were to test the link between perceptual elaboration and false memory by using a modified experimental paradigm (Zaragoza and Mitchell, 1996) and to evaluate the potential role of personality differences in creating false memories for suggested events by using three different measures associated with imagery, absorption, and suggestibility. So, the researchers hypothesized that:

1. Perceptual elaboration questions would increase false memory for suggested events.
2. Individual differences may relate to the degree in which elaboration would increase false memory for suggested events.

Source test data was collected from one hundred and thirty-two general psychology students from Kent State University. The participants were randomly assigned to either the Perceptual Elaboration Group (PEG) or the No Elaboration Group (NEG), with equal numbers in each group.

A 2 x 2 mixed design included the following:

• Perceptual Elaboration versus No Elaboration
• 1 exposure versus 3 exposures

• The participants watched a videotape of a house burglary. One week later, they read a narrative account of the incident, which mainly included accurate information. However, eight misleading suggestions were also included e.g. the burglar stole a ring when he didn’t even steal any jewelry.
• Some of the suggestions appeared once while others appeared three times.
• For the perceptual elaboration manipulation, participants answered questions (appeared in regular intervals) as they encountered them in the narrative.
• PEG – Follow-up questions were about location, physical appearance, perceived items, suggested items
• When a sentence included a suggested item, the question was always about the suggested item e.g. the location of the ring.
• To evaluate the association between the number of follow-up questions and false memory, participants were asked a different question about the fictitious item every time they encountered it e.g. Did the ring have diamonds? Was the ring on the dresser?
• Two days later – source memory test completed
• False memory measurement – The tendency for the participants to misremember viewing the suggested items in the video
• Did perceptual elaboration questions increase suggestibility? : Compared the incidence of false memory between the PEG and the No Elaboration Control Group
• To control for the effects of number of exposures to suggestion, the follow-up questions NEG participants were asked differed from those asked of the PEG. The NEG questions focused on surface characteristics such as what word in the sentence rhymed with king? This way, no further elaboration would be done on the mental image of the suggested items.

Used three measures:

• Tellegen Absorption Scale (Tellegen, 1982, unpublished manuscript)
• TAS – contains 34 true/false questions; correlates with variables of hypnotizability, imagery, and fantasy proneness (Roche and McConkey, 1990)
• Dissociative Experiences Scale (Bernstein and Putman, 1986)
• DES – self-report measure; relationship between dissociative tendencies and susceptibility to memory illusions (Winograd et al., 1998), imagination inflation effects (Paddock et al., 1998) and constructing false memories for entire autobiographical events (Hyman and Billings, 1998)
• Creative Imagination Scale (Wilson and Barber, 1978)
• CIS – asks participants to rate the similarity between imagined events and corresponding real events.

Participants were in groups of one to fifteen people and were told they would watch a brief video and then engage in other cognitive tasks.

Phase 1 – Eyewitness event/DES

• Watched 5 minute clip of house burglary (Ohio Police training tape)
• Young man stole several things from the house; was later caught by police after car chase
• DES questionnaire completed after video

• One week after video – read narrative describing each of 12 scenes from video
• Each participant exposed to 8 misleading suggestions corresponding to specific scenes
• For each participant - 4 suggestions appeared 1 time; the other suggestions appeared 3 times (1 exposure: 3 exposures)
• Repeated suggestions manipulation – reviewed 12 scenes in chronological order 3 times in a row; wording and aspects emphasized varied for the same scenes
• Description of scenes – projected individually on the wall in front of participants through a liquid crystal display
• Participants read each statement while a taped voice read aloud – equated exposure time of suggested items
• 15 seconds to answer follow-up question
• Had to answer every question, even if had to guess
• PEG – questions about location, physical appearance, and physical context
• NEG – superficial questions e.g. rhyming tasks, unscrambling, and grammatical judgment tasks
• The correct answer for NEG questions – always in the suggested item
• TAS questionnaire completed

• After 2 more days – surprise source memory test given
• 32 items were read on tape recorder for 8 seconds each
• 8 – suggestive items, 24 – filler items; 8 from each of 3 sources e.g. video only, both video & statements, or neither sources)
• Indicated source attribution under headings; 1 heading – "video", 1 heading – "statements"
• The questionnaire used a seven point Likert-type scale.
• Before test, participants were accurately informed that there were 4 possible sources for the items (yes – video, yes – statement; yes – video, no – statement; no – video, yes – statement; or neither to both sources).
• After test, CIS was given.
• The guided imagery task – tape recorded male voice

Perceptual elaboration and false memory for suggested events:

The results did strongly support the researchers hypothesis that perceptual elaboration

increases false memory (See Figure 1). PEG participants made significantly more source attribution errors than NEG (when all of the "yes" responses e.g. "maybe yes" were added together and compared to the sum of the NEG group and when looking at the "definitely yes" response from both groups). For both groups, errors increased for misattribution as exposure increased. The relation between repeated exposure and perceptual elaboration was unreliable. Table 2 showed how the perceptual elaboration manipulation influenced memory for the suggested items by indicating that perceptual elaboration increased misattribution to the video and also increased correct attributions to the statements. Repeated exposure increased correct attributions to statements and an ANOVA for the no video/no statement responses indicated the high reliability that repeated exposure also decreased forgetting the suggested item. This suggests that a paradoxical effect exists for repeated exposure and perceptual elaboration. False memory increases while memory is improved for the false item’s source. No differences for filler item source probes between groups indicated that the number of false alarms was comparable.

Individual differences for suggestibility:

TAS was the only personality measure that significantly correlated to suggestibility errors, specifically for ‘definite’ errors only. There also was a significant correlation between TAS and responses of ‘definitely yes’. No other significant findings were found.

Discussion

The results from the present study support the researchers hypothesis that perceptual elaboration about suggested items increases false memories, but there was no support found for repetition interacting with perceptual elaboration. There was not strong evidence that individual personality differences influence false memories for suggested items. TAS was the only personality measure that significantly correlated to suggestibility errors. This was in the case of the "definitely yes" responses and the significant correlations were "rather low."

The researchers claimed that the results provided strong evidence that the nature of the processing, instead of the amount, resulted in the increase of false memory, but they do so without certainty. They also noted that another possibility to explain their results could be that it was the meaningfulness of the elaboration that caused the effects. They do address the high rate of false memory in the NEG, especially in the 3 exposure group. They attribute this to the NEG participants having been exposed to misleading suggestions included in the narrative. They suggested that this more than likely caused them to spontaneously construct a mental image of all described items including the true and suggested. Therefore, the participants in the NEG had many possibilities to reflexively construct the false memory through elaboration. The authors basically provided several possibilities to explain their results, which for the most part seemed plausible.