Greenhoot, A.F., Ornstein, P.A., Gordon, B.N., and Baker-Ward, L., (1999). Acting out the details of a pediatric check-up: The impact of interview condition and behavioral style on children's memory reports. Child development, 70, 363-380.
Goals of the study:
1) An attempt to diminish age differences in recall of a past event by manipulating the recall context-- children were either verbally interviewed or asked to re-enact the doctors visit with props
2) An attempt to account for previously unexplained variance in children's recall by looking at two individual differences factors-- temperament (behavioral style) and verbal ability.
Why?
1) To learn more about children's memory processes
2) To help design better interview situations to facilitate child eyewitnesses' performance
Method
Kids who were scheduled for a regular pediatric check up:
-- 32 three year olds
-- 30 five year olds
Kids were interviewed in the university lab about their trip, and were randomly assigned to either the verbal or enactment condition
-- enactment condition-- the room was set to look like the doctor's office and had dolls and props for the child to use (see table 1), the props were both actual and distracter items; kids asked to "show and tell" what happened
-- verbal condition-- the room was similar but did not have the examination table or the props and instruments; kids asked to tell what happened
Overview of Procedure:
-- A researcher meet the family at the doctor's office for preliminary consent/assent
-- after the visit, the individual difference assessments were made
-- the first interview happened within on eweek of the visit, the second happened about six weeks later
Interviews:
-- Questions in both conditions went from being very general ("what happened during your check-up?") to very specific ("Did the doctor shine a light in your eye?")
-- were asked questions about what the nurse, and what the doctor did, in counterbalanced order
-- enactment condition: kids did a practice run with a camera, then were asked to show and tell; they were encouraged to use all the props
-- both conditions, some distracter questions were asked as well to check for accuracy-- absent feature questions were ones that could have occurred but didn't ("did you get a shot?") and extra event questions that did not and were very unlikely to have occurred ("did the nurse take your temperature?") why is getting a shot more likely than having your temperature taken?
Individual differences:
-- Test of Early Language Development (TELD): assesses child's ability to express and understand language forms and language content
-- Parent form of the Temperament Assessment Battery for Children: parent's filled out a behavioral style questionnaire that covers the child's activity, adaptability, approach-withdrawal, emotional intensity, ease of management, and persistence.
Coding of Recall transcripts:
Total Recall-- sum of the proportion of features recalled from open-ended and specific probe questions; correct denials were included in this tally, as were intrusions (false information in the free-recall) and false alarms (incorrect assertion of "yes" to an absent feature question).
Elaborative detail:
2= complete elaboration of a procedure
1= some but not all detail included
0= no recall or demonstration at all
Play and off-topic talk/demonstration was not included
Noteworthy Results
-- Context --
Present Feature recall: five year olds recalled more detail than three year olds; enactment lead to more recall than verbal interview
Non-administered features: enactment condition lead to more intrusions, in particular for the three year olds at the 1 week interview and for both groups at the second interview
Unrelated features: details not reported by medical staff or parent, not on the list of suggested features, but medically related, frequencies were tabulated; More intrusions occurred in the enactment than in the verbal interview, and more occurred in the 6 week interview
Because of these findings, they looked at the proportion of unrelated to present features for each child as an index of accuracy; enactment seemed to lead three year olds to be less accurate to open ended probes, and enactment (but not verbal) lead both groups of kids to be less accurate over time.
Questions about open-ended demonstrations-- (I don't really get this variable) -- Anyone?
Specific Probe questions-- at both interviews, the kids did not perform above chance; the five year olds were better at the non-administered features; from a signal-detection type analysis did not show any differences in response biases
-- Individual Differences --
Language-- not a mediator
Temperament-- looked at manageability and persistence
Manageability and intrusions in the enactment condition were positively related for the three year olds at one week; no relationship at 6 weeks
Persistence was negatively related to the three-year-olds answers; less persistence was related to higher total and cued recall and to more false alarms in the verbal condition; lack of persistence seems to have lead to a "yes" bias in responding
Discussion
It looks like the context and individual differences do matter in shaping the memory reports of young children, however not in the way that was expected
The enactment context lead to increased accurate AND inaccurate reporting, manageability lead to more inaccurate reporting and persistence lead to a "yes" bias in the younger children!
Perhaps the easy children were more inclined to play and the less persistent children were more inclined to say "Yes" than to really think about it
Or, it could be that the children were not really relying on individual memories, rather they were thinking about their general doctor-visit knowledge
Forensic and Clinical Psychologists beware-- enactment has been used and is thought to be helpful!