Hyman & Billings, (1998). Individual Differences and the Creation of False childhood Memories. Memory, 6, 1-20.

• To replicate past research where memories have been created via repeated interview sessions.
• To replicate the finding that accessing self-relevant knowledge at interview one predicts the creation of a false memory
• To explore whether various individual differences (i.e., personality characteristics) are predictive of false-memory creation using the implantation paradigm.

We know that some people will create false memories in the laboratory, yet others do not-- why is this? Perhaps some people are more prone to false memory creation because of who they are-- i.e., they might have a personality profile that leaves them more susceptible to false memory creation. If we could predict who is more likely to create a false memory based on responses to personality questionnaires, then we might also be in a better position to assess the veridicality of a memory that arises in a therapy session.

Misinformation (MI) studies have found that some variables do predict whether someone is likely to accept suggested post-event information as part of the original situation:

• Age-- children are more susceptible to MI
• Occupation-- artists and architects are more suggestible
• Gudjohnsson's interrogator suggestibility scale (Hypnotisability, Low assertiveness, High neuroticism)
• Social desirability-- lying to make yourself look better
• Low intelligence
• Poor overall memory ability
• Imagery

So the next question is this-- are these variables related to the full fledged creation of an actual event memory as well?

To answer this question, participants filled out a battery of cognitive/personality measures at one point in time, then in a seemingly unrelated experiment, participants were interviewed about several real and one false childhood event.

• General psych students were asked permission to contact their parents.
• Childhood events questionnaires were sent home to those parents.
• After the questionnaires were returned, those students whose parents included at least two events were contacted and interviews were scheduled.
• 83 were asked to participate, 72 agreed to do so, 66 finished the interviews, but only 48 completed both the interviews and the individual differences questionnaires.

• Interviews were scheduled one day apart, each participant had the same experimenter both times
• Interview 1-- the participant was asked to recall the true events first, the false one last; each event was introduced with a title, the participant's age, and the cues provided by the parents.
• They were asked to describe as much as they could about the event; when they could not recall an event they were encouraged (but not required) to think about it and imagine what the event might have been like.
• They were then encouraged to continue thinking about the events
• Interview 2-- they were asked about each event again; the events were only introduced by title, extra details were only given when the participant could not remember the event, and they were also encouraged to visualize those events that they could not remember.
• At the end of the second interview, they gave a confidence rating of each memory, then were debriefed.

The following cognitive/personality scales were used:

• Creative Imagination Scale (CIS) -- measures vividness of memory and responsiveness to suggestion.
• Tellegen Absorption Scale (TAS) -- measures the extent to which people become involved or immersed in everyday events.
• Dissociative Experiences Scale (DES) -- measures the extent to which one dissociates or separates from reality; the frequency of interruptions of the normal integration of consciousness, memory, and identity.
• Marlow-Crowne Social Desirability Scale (SDS) -- measure the extent to which one is inclined to present the self favorably, whether it is accurate or not.

Results-- as they relate to the three goals of the study

1) Replication of previous implantation studies

• 10 pp created a clear false memory
• 8 pp created a partial false memory
• 18 had no memory, but were trying
• 30 had no memory

** The confidence levels the pp reported on their false memories were comparable to those reported for true recovered/reconstructed memories.

2) Replication of the relationship between including self-relevant knowledge and the creation of a true memory

The inclusion of self relevant knowledge in the first interviewed was scored as a dichotomous variable-- either the participant did or did not include related information.

Self-relevant information was defined as any information the participant provided about their childhood that could have been related to spilling punch at a wedding, whether the information was used to agree with or to deny the event.

Those who did include self-relevant information were more likely to create a false memory-- c ²(3)=18.44, p<.001

• 46 pp discussed SR information, of these, 17 created a false memory of some sort; 1 to of 20 who did not discuss SR information also created a false memory.

3) Individual Differences Variables-- bivariate correlations between the participant's memory status, some interview characteristics, and the five scales were computed.

• Memory status at the second interview was transformed into continuous variable such that: 1= non memory, 2= none but trying, 3= partial false memory, and 4= full false memory.
• # of events interviewed on
• Proportion of recalled events that were true
• Average number of words used to talk about the true and the false events in the second interview
• Whether SR information was included

Interview dynamics and false memory creation:

• talking more was positively related to memory recovery/creation: true memories, r=.52, false memories, r=.51
• the number of words used to describe true and false events were related, r=.73
• when asked about more true events, they tended to use fewer words to describe each event; true, r=-.34, false r=-.26

Two possible explanations for these findings

1. The interviewers built such good rapport that they lead pp to talk a lot, which inadvertently lead the scorers to attribute more words as indicating a false memory

2. Those pp who talked more ended up integrating the false suggestions in with SR information-- the correlations between self-relevant talk and number of words were also significant; true, r=.39, false r=.34.

Individual Difference Variables and False memory creation:

• # of words in the false memories and DES-- r=.34
• # of words in the true memories and DES-- r=.36
• False memory and CIS-- r=.36
• False memory and DES-- r=.48

Self Relevance Issue-- the discussion of self-relevant information at the first interview might have initiated a schematic reconstruction process, that occurred in three steps:

1) self relevant knowledge and the false event knowledge are stored together

2) at retrieval, both types of information come to mind

3) because both come to mind in response to the same cue, source monitoring becomes very difficult and the likelihood of making a source monitoring error (i.e., the acceptance of a false suggestion as a true memory) is increased.

The tendency to experience dissociation, to use mental imagery, and to be more hypnotizable most likely facilitate this process. For example, those people who are more dissociative may be used to integrating external information into their self-concept. As a result, they may use less stringent source monitoring criteria.

An important finding to consider is this as well-- these findings do shed light on the characteristics of the false memory creation process; however they do not shed light on how one might distinguish between a false and a recovered memory.