Running Head: FAMILY MEMORIES OF SHARED EXPERIENCES
Family Members' Agreement on Memories of Shared Experiences
Rory Remer, Ph.D.
Kimberly S. Gorman, M.S.
Dana Derrickson-Beake, M.S.
William A. Pinto Jr., M.S.
Department of Educational and Counseling Psychology
University of Kentucky
Agreement of family members recollections of shared events was explored. Mothers, fathers and eldest children from 32 families were assessed through both free recall and specific questions. Comparison of 24 usable accounts indicated few points of similarity from the content analysis but somewhat greater agreement on specifics, except for emotional reactions. Results suggest terming families' memories, particularly affective attributes, as "shared" is questionable, except where traumatic events are concerned.
Family Members' Agreement on Memories of Shared Experiences
One of the characteristics often noted as defining a family is the sharing of a common history/experiences (Kaufman, 1993, p. 241). Since we are interested in simulating families for use in research, we decided to explore the validity of this assumption - one of the key criticisms of our contentions that simulated families are similar to actual ones. To do so we conducted a study of shared family memories.
Loftus and Loftus (1980) found that many psychologists thought: "Everything we learn is permanently stored in the mind..."(p.410) Research has clearly demonstrated that memories are usually reconstructions of past events (Loftus, 1993). In fact:
Truth and reality, when seen through the filter of our memories, are not objective facts but subjective, interpretive realities. We interpret the past, correcting ourselves, adding bits and pieces, deleting uncomplimentary or disturbing recollections, sweeping, dusting, tidying things up. Thus our representation of the past takes on a living, shifting reality; it is not fixed and immutable, not a place way back there that is preserved in stone, but a living thing that changes shape, expands, shrinks, and expands again, an amoeba-like creature with powers to make us laugh, and cry, and clench our fists. (Loftus & Ketcham, 1991, p.20)
Although some studies (e.g., Eisenberg, 1985; Fivush, 1991; Fivush & Fromhoff, 1988) have dealt with collective memories and influences on them, most studies have concentrated on individual memories (e.g., Counseling Psychologist, 1995). Little has been done with memories of shared events and the reactions of those involved to them (Edwards & Middleton, 1986; Hirst & Manier, 1996). We wished to find out just how "shared" most family events actually were. Our contention was that even "shared" events had quite different constructed memories for those who shared them.
The 105 members of 32 families volunteered to be interviewed for the study. Families ranged in size from two to five members. Fifty-nine (59) females and 46 males, ages ranging from 64 to 7 participated. All were from lower middle to upper middle SES level with sufficient verbal and written communication abilities to be able to write the required accounts of the situations. Families were not randomly selected, but were those available to those doing the interviews.
Since only families with at least three members - mother, father and child of sufficient age (at least 7 years old) and verbal ability to respond adequately to the research protocol - qualified, two families were immediately eliminated. In addition, the data from 6 families were discarded from the primary analysis because the interviews did not correspond to the protocol and could not be scored. The final sample consisted of 82 individuals from 24 families ranging in size from three to five members.
Interviewers were 18 doctoral and upper level masters students enrolled in a family therapy course. All had had previous course work and practice in conducting therapeutic interviews. Training, including instructions and a demonstration of the interview procedures, was conducted.
The instrument employed consisted entirely of a three page protocol divided into five sections. In Section 1, after the gender age and family role of the participant were indicated, space was provided for listing and ranking (from most memorable to least memorable) five family experiences; sections 2 and 4 were spaces for the participants to write down an account of a family experience; and sections 3 and 5 consisted of five specific questions about the experiences from sections 2 and 4 that were the focus of the study.
Procedures Data collection. The family interviews were conducted at the family residences by the 18 interviewers at the convenience of the participants. After informed consent was obtained from all family members, the protocol was implemented in two stages. First, all family members were administered section 1. The interviewer then compared the replies of family members to select an event that had been listed in common across the family members' lists. If more than one event was listed in common, the highest ranked event was chosen as the focus of section 2. If no event was common to all lists, then an event listed on only two of the family members' lists was selected. In that case, the third family member was asked to use that event for the focus of section 2 and the event was ranked sixth. Although never the case, interviewers were instructed to select the highest ranked event on the mother's list to use as the focus of section 2, if no commonalities occurred.
The form of the question designed to obtain information about memories is a significant biasing factor in retrieval. Leading questions are particularly problematic (Brown, 1977; Davis & Schiffman, 1985; Fivush & Schwarzmueller, 1995; Loftus, 1975; Loftus & Zanni, 1975; Pirolli & Mitterer, 1984; Weinberg, Wadsworth, & Baron, 1883). Influence of others (Fivush, 1991) - older siblings (Eisenberg, 1985) or mothers (Fivush & Fromhoff, 1987), for example - also introduce unwanted influences.
To avoid such problems as much as possible, in section 2, each family member was asked separately to write a brief account of the common event in the space provided. In some instances, the participant continued the description on the back of the sheet. After the free form account was completed, the responses to five specific questions (section 3) concerning the event were solicited: “Who was the focal point of the event?”, “Where did the event occur?”, “When did did the event occur?”, “What did the you think about the event?”, and “What was your primary emotional reaction?”
For sections 4 and 5 each respondent's highest remaining ranked event was treated in the same manner. Thus each person had his or her highest ranking event as the focus of one account. (Where the highest ranked event was used in section 2, the second highest ranked was the focus of the second account.)
Scoring. Each family's common accounts were scored independently initially by two readers from a group of four. First, the number of words was recorded. Second, the ranked lists of events were compared for the number and rank of the events listed in common. Third, each reader content analyzed the verbatim accounts for points of similarity--each family member’s account was read, then all were reread and any common references, themes or reactions were recorded. Finally, the responses to the specific questions (see previous section) were judged for comparability. The second account (sections 4 and 5) was only used as control on word production. (Where the number of family members was greater than three, only the responses of the oldest child were considered.)
When the independent scoring was completed, the group of four readers met as a whole to compare results. In the case of any disagreement concerning the similarities of the accounts or the comparability of the responses to the specific questions asked, the discrepancies were discussed and consensus was reached through reference back to the original accounts in question and group process, similar to analysis protocols employed in Consensual Qualitative Research (CQR, Hill, Thompson & Williams, 1997) and Synergistic Analysis of Structured Essays (SASE, Tinsley, 1997).
The number of mother/father (M/F), mother/child (M/C), father/child (F/C) and mother/father/child (M/F/C), whole family, matches were produced for events listed in common, points of similarity of accounts, and responses to specific questions. Only three family members' data--mother, father and oldest child--were included in the overall analysis of the 24 families, so that the analyses would not be skewed simply by the inclusion of more members. Summary statistics were produced for all variables by category (see Table 1). Repeated measures ANOVA procedures--comparing the mean number of events in common, points of similarity, and specific question response matches across each type of grouping--were also employed to determine if memories held in common were more likely to occur between or among different family member groupings.. Since no norms exists against which to compare the responses, no such comparisons were (or could be) performed. Similarly, testing to see if the means were significantly different from 0, did not seem to be pertinent.
Looking at the means for each variable (see Table 1), in no instance did the matches rise even to 50% of the number of matches possible (maximum possible for events listed and for specific question responses was 5 for each grouping; for points of similarity the maximum was theoretically infinite, but in reality was limited by the length of the accounts and number of points made in the account). Nor were the means even 50% of the maximum number of matches observed for any one family. In examining the responses to the five individual questions contributing to specific question agreement across all groupings, only 17 of 242 agreements across all 5 questions (7%) came from emotional reactions. Most likely cuing memories by asking about specific details mentioned by other family members would produce more matches, as it did with our cuing, match of emotions would likely remain consistent with our findings.
Insert Table 1 Here
Comparison of word counts across accounts and family members yielded a significant difference between accounts (the first account being greater than the second), but no differences among family members. Comparisons of the ranks of events listed and described evidenced neither significant differences for account nor for family members. These analyses indicated that the protocol seemed to function as designed.
Comparison of accounts and questions both grouped pairwise (e.g., mother/father=M/F) and for all three family members (M/F/C) produced no significant differences by grouping for either number of points of similarity (F= 1.57, df= 3,69, p< .204) or number of events listed in common (F= 2.03, df= 3,69, p< .117). A significant difference was obtained for number of specific question agreements (F= 6.67, df= 3,69, p< .001), M/F/C agreement proving significantly higher than pairwise comparisons. This result seems contradictory to expectation. However, further examination provides one possible reason for this occurrence, one that is very illuminating.
Insert Table 2 Here
One exception that emerged should be noted. One family of five members had experienced a traumatic event together. On each list this event was rated first. Every account produced was consistent with the others both factually and affectively. While this agreement could have been the product of the family retelling the story, the impact of the event itself could account for the unusual outcome. In any case, this “outlier” account tended to inflate the averages, by contrast only reinforcing the results obtained from the analysis of the entire sample.
Defining what is a "family" has been difficult. A shared history has often been suggested as one defining feature. Memories of shared events are part of that history. While our results suggest that factual aspects of such accounts produce some agreement, affective reactions are quite often discrepant.
Consistent with our findings, other researchers have characterized "group remembering" as a constructive process (Edwards & Middleton, 1986; Fivush, 1994; Hirst & Manier, 1996). Given that the group remembering process has generalizable characteristics even though "communicable frames vary according to the settings in which they occur (Edwards & Middleton, 1986)", a strong probability exists that this aspect of family definition can be reproduced/simulated, particularly when an appropriate frame is induced.
Terming family memories as "shared" is questionable at best, except when the event in common has been traumatic for all involved. Certainly, "shared family memories" being constructive, do not seem to stand in the way of simulating realistic family interaction. Perhaps other components of a "family history" will prove more discriminating than family memories in preventing such simulation. That contention is yet to be examined and supported.
Brown, L. (1977). Leading questions and the eyewitness report of a live and a described incident. Psychological Reports, 40, 1041-1042.
Davis, J., & Schiffman, H. R. (1985). The influence of wording of interrogatives on the accuracy of eyewitness recollections. Bulletin of the Psychonomic Society, 23, 394-396.
Edwards, D., & Middleton, D. (1986). Joint remembering: Constructing an account of shared experience through conversational discourse. Discourse Processes, 9, 423-459.
Eisenberg, A. R. (1985). Learning to describe past experiences in conversation. Discourse Processes, 8, 177-204.
Fivush, R. (1991). The social construction of personal narratives. Merill-Palmer Quarterly, 37, 59-80.
Fivush, R. (1994). Young children's event recall: Are memories constructed through discourses? Consciousness and Cognition: An International Journal, 3, 356-373.
Fivush, R., & Fromhoff, F. A. (1988). Style and structure in mother-child conversations about the past. Discourse Processes, 11, 337-355.
Fivush, R., & Schwarzmueller, A. (1995). Say it once again: effects of repeated questions on children's event recall. Journal of Traumatic Stress, 8, 555-580.
Hill, C. E., Thompson, B. J., & Williams, E. N. (1997). A Guide to conducting consensual qualitative research. The Counseling Psychologist, 25, 517-573.
Kaufman, T. S. (1993). The combined family: A guide to creating successful step-relationships. New York, NY: Plenum Press.
Hirst, W., & Manier, D. (1996). Remembering as communication: A family recounts its past. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory. (pp. 271-290). New York, NY: Cambridge University Press.
Loftus, E. F. (1975). Leading questions and the eyewitness report. Cognitive Psychology, 7, 560-572.