In order for a memory to be classified as a flashbulb memory, Brown and Kulik (1977) required a subject to firstly recall the event and then to provide information which fell into at least one of the following categories: "when they heard the news, where they were, what they were doing and with whom, and how they felt upon hearing the news" (Finkenauer, Luminet, Gisle, El-Almadi, van der Linden, & Philippot, 1998, p.516). In a flashbulb memory, people have, or believe that they have, rather detailed recollections of the specifics regarding the context in which they first learned of important events. According to Weaver (1993), one factor that does seem to differ between flashbulb and regular memories is the overconfidence participants seem to have about the accuracy of their flashbulb memories (Ashcraft, 1998, p. 231).

Brown and Kulik (1977) stated that in order for flashbulb memories to occur, two critical variables are needed. First, the event must be new or unexpected and thus must elicit surprise. Given a sufficient level of surprise, the event is then evaluated in terms of consequentiality or personal importance, which, in Brown and Kulik`s perspective, is equated with emotional arousal. The following study analyzes peoples` memories for two events, the loss of the Savoye family`s house in a tornado in 1992, and the Challenger disaster which occurred in 1986, in light of this model of FBM formation. The first event, the loss of the Savoye family`s house, was obviously much more consequential and personally relevant to members of the immediate family than it was to friends and neighbors. Assuming that the level of surprise was the same for both groups of people, the responses of these two groups are going to be compared in order to determine the effect that consequentiality has on the formation of flashbulb memories. In addition, responses regarding the tornado are going to be compared to those regarding the Challenger disaster, an event which should have similar personal relevance for all participants. This is to control for the confound that members of one group may just be more prone to flashbulb experiences than members of the other group.

Additionally, Brown & Kulik (1977) proposed that rehearsal represents an important mediating variable that leads to further completeness of the FBM. Rehearsal may effect FBM in two ways: (1) it can improve the FBM by reinforcing existing memory traces, and (2)it can modify the FBM content (Finkenauer et. al., 1998, p. 517). According to the model proposed by Brown and Kulik (1977), it is predicted that consequentiality should be associated with both more detailed FBMs and with more overt rehearsal.

RESEARCH PARTICIPANTS
Twelve individuals participated in this study. Research participants were divided into two groups of six members each. Group One consisted of immediate members of the Savoye family (i.e. mother, father, daughter, son, grandmother, grandfather). Group Two consisted of six friends and neighbors of the Savoye family.

MATERIALS
A questionnaire was developed and distributed to research participants. Participants were first asked some specifics regarding the context in which they first learned that the Savoye`s house had been destroyed. These included: (1) when they heard the news (i.e. date and approximate time), (2) where they were when they heard the news, (3) what they were doing at the time and with whom, and (4) how they felt upon hearing the news. They were then asked to rate the following characteristics on a scale of 1 to 5: the vividness of this memory, their confidence about the accuracy of this memory, their level of surprise upon hearing about this event, how consequential they felt this event was to them, and finally, the frequency with which they have talked about this event. These same questions were then asked about the second event, the Challenger disaster.

PROCEDURE
Questionnaires were distributed to members of the Savoye family, as well as to a select group of their friends and neighbors (i.e. those who were aquainted with the family when their house was destroyed in 1992). All distributed questionnaires were returned, and any questions that participants had regarding the nature of the study were answered.

I. WHEN DID YOU HEAR THE NEWS?
Regarding the loss of the Savoye`s house, three out of six members of Group 1 (family) recalled the exact date and time (to the minute) when they learned of this event. The other three members of this group could recall the exact date and an approximation of the time. The responses of all six members of Group 1 appear to be extremel accurate. On the other hand, only one member of Group 2 (friends) could recall the exact date and time when he or she learned of this event. Three out of six members of Group 2 recalled the exact date and an approximation of the time, while two out of six could only recall an approximation of the date (one approximated the time, and the other one did not mention time). Members of Group 2, although overall less specific than members of Group 1 in their responses, still appear to be reasonably accurate. Thus, while members of Group 1 seem to have more detailed memories for the date and the time when they learned of this event than members of Group 2, members of both groups appear to be reasonably accurate in what it is that they do remember.

Regarding the Challenger disaster, four out of six members of Group 1 (family) fave no mention of date or time, while two out of six approximated the date and the time when they learned of this event. In Group 2 (friends), one participant did not remember, one gave no mention of date or time, three out of six gave no mention of date (one provided the exact time, and two approximated the time), and on participant could recall the exact date and an approximation of the time. All of the responses which mention the date and/or the time when the individual learned of this event appear to be reasonably accurate; nobody provided a response which was obviously incorrect.

These results suggest that members of Group 1 (family) do not have better memories for dates and times than members of Group 2 (friends). The significant difference here is that the first event, the tornado which destroyed the Savoye`s house, was much more consequential and personally relevant to members of Group 1 than it was to members of Group 2. Thus, members of this group have more detailed memories of when they learned of this event than do members of the second group.

II. WHERE WERE YOU WHEN YOU HEARD THE NEWS?
Five out of six members of Group 1 (family) could recall their exact location when they learned of the Savoye`s loss. However, only two of the six members of Group 2 (friends) could recall their exact location. Regarding the Challenger disaster, five out of six members of Group 1 could recall exactly where they were when they learned of this event. Two out of six members of Group 2 could recall their exact locality, and three out of six could recall approximately where they were. While it may look as if members of Group 1 have slightly better memories for location, I think that this is unlikely. Five out of six members of each group have some idea of where they were when they learned of the Challenger disaster.

III.WHAT WERE YOU DOING AT THE TIME AND WITH WHOM?
Regarding hearing about the destruction of the Savoye`s house, all six members of Group 1 (family) could accurately recall exactly what they were doing at the time and with whom. Four out of six members of Group 2 (friends) could recall what they were doing at the time and with whom. Thus, more members of Group 1 have detailed memories of the circumstances surrounding when they learned of this event than do members of Group 2.

On the other hand, three out of six members of each group could remember exactly what they were doing and with whom when they learned of the Challenger disaster. One member of each group recalled approximately what he or she was doing and had a general idea of who he or she was with. In addition, one member of each group had no recollection of what he or she was doing at the time. Thus, these two groups show a similar pattern of remembering what they were doing and with whom when it comes to an event which is assumed to be equally consequential to members of both groups.

IV. HOW DID YOU FEEL UPON HEARING THE NEWS?
Participants` reactions to hearing about the Savoye`s loss differed between the two groups. The most common responses for Group 1 (family) were: shocked (4/6), sad (3/6), relieved (3/6), and frightened or scared (3/6). The most common responses for Group 2 (friends) were: concerned, worried, or sympathetic (4/6), and shocked or surprised (4/6). This is understandable, given the different perspectives of the members of each group. Members of Group 1 were undoubtedly sad about their tremendous loss, relieved that nobody had been injured, and scared about what was to become of then now that they had no home. On the other hand, members of Group 2 were understandably concerned about the Savoye family, and sympathetic towards their loss. Both groups expressed shock at this unexpected event.

Participants` reactions to hearing about the Challenger disaster were more uniform. The most common responses among members of both groups were: shocked (3/6 Group 1, 5/6 Group 2) and sad (4/6 Group 1, 3/6 Group 2).

V.VIVIDNESS,CONFIDENCE,SURPRISE,CONSEQUENT.,FREQ.
Participants were asked to rate the vividness of their memories on a scale of 1 to 5 (1=least vivid, 5=most vivid). Regarding memories for the loss of the Savoye`s house, the mean vividness rating for Group 1 (family) was 5.00 (SD=0), slightly higher than the mean vividness rating for Group 2 (friends) (X=4.33, SD=0.94, range: 3-5). For memories of the Challenger disaster, the mean vividness rating for Group 1 was 4.00 (SD=1.15, range: 2-5), and was also 4.00 for Group 2 (SD=1.41, range: 1-5).

Participants were asked to rate, on a scale of 1 to 5, how confident they were about the accuracy of these memories (1=not confident, 5=very confident). For memories of the Savoye`s house, the mean confidence rating for Group 1 was 5.00 (SD=0), and for Group 2 was 4.33 (SD=1.11, range: 2-5). For memories of learning about the Challenger disaster, members of Group 2 (X=5, SD=0) seem to be more confident on average than members of Group 1 (X=3.83, SD=1.46, range: 1-5) are.

Next, participants were asked to rate, also on a scale of 1 to 5, their level of surprise upon hearing about each of these events (1=not surprised, 5=very surprised). For the loss of the Savoye`s house, the mean level of surprise for Group 1 was 5.00 (SD=0), and for Group 2 was 4.83 (SD=0.37, range: 4-5). For the Challenger disaster, the mean level of surprise was 4.67 for Group 1 (SD=0.47, range: 4-5) and 4.83 for Group 2 (SD=0.37, range: 4-5).

The next variable which participants were asked to rate was how consequential they felt that each of these events was to them personally (1=not very consequential to his/her life, 5=very consequential). Not surprisingly, members of Group 1 (family) rated the loss of the Savoye`s house as more consequential to them (X=5, SD=0) than did members of Group 2 (friends) (X=3.17, SD=1.07, range: 2-5). The Challenger disaster, on the other hand, was rated as similarly consequential by members of both groups. For Group 1, the mean was 2.33 (SD=1.37, range: 1-5), and for Group 2 the mean was 2.17 (SD=1.34, range: 1-5). This event was thus much less consequential to members of both groups than was the loss of the Savoye`s house.

Finally, participants were asked to rate the frequency which with they have talked about each of these events (1=not at all, 5=quite frequently). Members of Group 1 reported talking about the loss of the Savoye`s house on average more often than did members of Group 2 (Group 1: X=4.83, SD=0.37, range: 4-5; Group 2: X=2.83, SD=0.69, range: 2-4). Neither group reported talking about the Challenger disaster with much frequency. The mean for Group 1 was 1.83 (SD=0.69, range: 1-3), as was the mean for Group 2 (X=1.83, SD=0.37, range: 1-2).

VI. REHEARSAL EFFECTS
According to the model proposed by Brown and Kulik (1977), consequentiality should be associated with the frequency of overt rehearsal. This is exactly what was found in this study. A +0.87 correlation was found between the consequentiality of the loss of the Savoye`s house for participants and their self-reported frequency of talking about this event. This result was statistically significant (p<.001, two-tailed). The correlation between the consequentiality of the Challenger disaster for participants and their overt rehearsal of this event was slightly lower (r=+0.50, p<.10, two-tailed).

Frequency of overt rehearsal was also related to the self-reported vividness of the memories. In this study, a +0.52 correlation was found between the vividness of participants` memories for hearing about the loss of the Savoye`s house and their frequency of talking about this event (p<.10, two-tailed). A +0.83 correlation was found between the vividness of participants` memories for hearing about the Challenger disaster and their frequency of talking about this event. This result was statistically significant (p<.001, two-tailed).

A correlation was also computed between the vividness and consequentiality to determine if events which are judged by participants to be more consequential produce more vivid flashbulb memories. A +0.48 correlation was found between self-reported vividness and the consequentiality of the loss of the Savoye`s house. A +0.43 correlation was found between self-reported vividness and the consequentiality of the Challenger disaster. However, these results were not statistically significant.

Consequentiality seems therefore to be related to more detailed FBMs. Three members of Group 1 remembered the exact minute when they became aware that their house had been destroyed. Two of these individuals (Mr. Savoye and his son) watched the tornado destroy their house while they were sitting in their car in the driveway. Both recall that the clock in the car read exactly 11:03am when a large tree landed on top of the house, causing it to collapse. In addition, five out of six members of Group 1 recall exactly where they were located when they learned of this event. As said previously, Mr. Savoye and his son were in the driveway. The son even recalls the color of the car that they were sitting in (green). Mrs. Savoye was driving home when this event occurred. She was stopped by her husband about two houses away, and was informed of what happened, and what she was about to see. The daughter was at school, when she had a strange sensation that something was wrong; she tried calling home from a payphone near the school gym, and eventually reached her father, who told her what had happened. She remembers numerous details from that day, such as who she ate lunch with and even what she ate for lunch. Even grandma remembers what she was doing when she got the call: she was fixing herself cottage cheese and fruit for lunch. In discussing this event with the family, I have learned taht all four members of the immediate family remember exactly what they were wearing on this day. However, this could be due to the fact that the family was forced to wear these clothes for a few days following this incident.

Not only are the accounts given for the Savoye`s loss given by Group 1 (family) more detailed than the accounts given by Group 2 (friends), but the majority of the memories provided by both groups for this event are more detailed than are the memories for the Challenger disaster. This event was rated as more consequential for all participants than was the Challenger disaster. Moreover, memories for the context in which participants heard about the Challenger disaster are equally detailed for both groups. This is assumed to be because this event had the same personal relevance for all participants in this study. However, another iterpretation of the more detailed memories for the Savoye`s house than for the Challenger disaster is that the tornado which destroyed the Savoye`s house struck in 1992, while the Challenger exploded in 1986, six years earlier. I think, however, that differences in consequentiality provide for a much more adequate explanation of these findings. It has been found that peoples` forgetting curves for FBMs are far less affected by time than are their forgetting curves for other types of memories (Bohannon & Symons, 1992). Thus, the passage of time should theoretically have no effect on the detail of a flashbulb memory. Therefore, the present study supports the notion that, assuming an adequate level of surprise or shock, differences in amount recalled and in the detail of a FBM are the result of how consequential the particular event was to that person.

Bohannon, J.N., & Symons, L.V. (1992). Flashbulb Memories: Confidence, consistency, and quantity. In E. Wingograd & U. Neisser (Eds.), Affect and accuracy in recall: Studies of "flashbulb" memories (pp. 65-91). Cambridge: Cambridge University Press.

Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5, 73-99.

Finkenauer, C., Luminet, O., Gisle, L., El-Ahmadi, A., van der Linden, M., & Philippot, P. (1998). Flashbulb memories and the underlying mechanisms of their formation: Toward an emotional-integrative model. Memory and Cognition, 26, 516-531.

Weaver, C.A., III. (1993). Do you need a "flash" to form a flashbulb memory? Journal of Experimental Psychology: General, 122, 39-46.