IMAGERY AND EMOTION COMPONENTS OF EVENT DESCRIPTIONS ABOUT SELF AND VARIOUS OTHERS

Imagery and emotion have been identified as two of the main component systems of autobiographical events. It is not yet known, however, whether a primary focus on either the self or others may have an impact on these components. To investigate this issue, half of the participants in this study provide a real and made-up event description about themselves, and half provide descriptions about a well-known other. In addition, all participants generated a made-up event description about an unfamiliar other. In accord with predictions generated from a multiple-system model, real events received higher visual detail, imaginability, and positive emotion ratings than made-up events. This pattern was also evident for a novel measure of imagery, in which real events were rated as being much more dynamic than made-up events. However, contrary to theoretical positions which postulate a special enhanced role for self-referent information processing, the selfdescriptive events were not rated as being easier to imagine and did not have more positive emotions or visual detail, than descriptive events about well-known others. This pattern suggests that efficient cognitive schemata may be involved in the processing of information about both the self and wellknown others. In contrast, descriptions of an unfamiliar other received lower imagery and emotionality ratings, suggesting that less well-differentiated cognitive structures are involved in component processing for these individuals. Cuvinte-cheie: imagerie (imagistică), emoţie, sine-altul, real-inventat.


INTRODUCTION
has suggested that autobiographical memories are the product of distinct component systems that include, for example, visual imagery and the emotions associated with the described event.Research investigating this multiplesystems model has determined the relative importance of these distinct components by having persons rate their autobiographical memories on the degree of visual imagery, spatial context, and how strongly the original emotions were reinstated (Rubin, 2005).Across several studies it has been found that the strength of recollection of an event was best predicted by the vividness of visual imagery, followed by emotions (Rubin Schrauf & Greenberg, 2003;Rubin, 2005).
Further research has examined how different types of memories or events may impact on the various components of the multiple-systems model.Crawley and Eacott (2006), for example, contrasted ratings of components for autobiographical memories based on personal recollections (i.e., the birth of a younger sibling) versus memories that were clearly based on second-hand sources (i.e., family stories about one's own birth).Congruent with predictions generated from the multiple-systems model, the autobiographical memories displayed higher ratings for visual detail, spatial location, and emotional feelings than did the memories generated from second-hand sources.Using a different paradigm, Kealy, Kuiper, and Klein (2006) had participants generate both real and made-up events about themselves.Again consistent with the multiple-systems model, real events could be readily distinguished from made-up events, as the former had significantly higher imagery ratings of visual detail, location and time than the latter.
In light of the above, one purpose of the present study was to explore a further important parameter that may also impact on key components of the multiple-systems model.In particular, we examined the extent to which a primary focus on either self or others, when describing events, may also bear on the imagery and emotion components of this model.Although it is clear that we can have distinct memories about self and highly familiar others (Kuiper & Rogers, 1979;Symons & Johnson, 1997), little is yet known about the extent to which these memories can be distinguished in terms of the various components of the multiplesystems model.Accordingly, we examined this issue by having participants in the present study provide both real and made-up event descriptions about themselves and a well-known other currently significant in their lives (e.g., a close friend).One possibility here is that the component ratings for self-referent event descriptions, when compared with descriptions about well-known others, may be richer in imagery and higher in emotionality.Such a pattern would be consistent with the notion that the self provides for very elaborate processing of personally relevant information than is even more enhanced than the processing of information about well-known others (Kircher et al., 2000).In turn, this greater elaboration would result in higher imagery and emotionality ratings for the self-referent events.Furthermore, by including both real and made-up events, we could determine the extent to which such a self-referent enhancement pattern may be evident across both these types of events.
Alternatively, it is also quite possible that there will be no distinction between the component ratings for the self versus well-known other event descriptions.In contrast to the notion that self-referent processing is uniquely elaborate (Kircher et al., 2000), several other investigators have highlighted the substantial similarities between self and well-known other information processing (Gillihan & Farah, 2005;Kuiper & Rogers, 1979;Symons & Johnson, 1997).Some of these researchers have proposed that individuals also develop and employ very efficient cognitive schemata for processing personal information about significant others, thus resulting in quite similar elaboration for both self and well-known others (Kuiper & Rogers, 1979).If this is the case, then one would not expect differences between the component ratings for self versus well-known others.In other words, the selfreferent event descriptions would not display higher imagery and emotionality ratings than those for well-known others.
A further goal of the present study was to examine the potential effects of degree of familiarity of the other-referent target on imagery and emotion component ratings.Thus, in addition to the well-known other condition, participants also provided a made-up event description about an individual they were just meeting for the first time, namely, the experimenter.Previous research indicates that the processing of personal information about such unfamiliar individuals is relatively inefficient and is not schema-based (Kuiper & Rogers, 1979).Thus, without any specific cognitive structure or knowledge base concerning this unfamiliar other, processing would lack the degree of elaboration evident for well-known others or the self.In turn, this suggests that it would be more difficult to image a made-up event about this unfamiliar other, and it would have significantly less emotionality, when compared to the made-up events about well-known others or the self.
Finally, given the central role of visual imagery in the multiple-systems model (Rubin, 2005), we wanted to expand this component of the model to also include consideration of imagery dynamics.As described by Enright (1997) this term refers to the movement characteristics associated with an image.Thus, some events are imagined in a static fashion, like a still picture, whereas others may be more dynamic, like a slide show.Finally, other events may be the most dynamic of all, and imagined more like a movie.In her research, Enright found that real autobiographical memories were much more dynamic than made-up events about the self.Accordingly, this additional component of imagery was assessed in our study by having participants provide an imagery dynamic judgment for each event description.Here, we predicted that real events would be more dynamic than madeup events.Furthermore, to the extent that self-referent event descriptions are indeed more elaborate than event descriptions about a well-known other, they would also display more imagery dynamics.Conversely, if self-referent and wellknown other referent event descriptions are equally elaborated, we would not expect any imagery dynamic distinctions between these two conditions.Finally, we expected that the made-up event descriptions for an unfamiliar other (i.e. the experimenter) would be much more static in terms of imagery dynamics than either the self or well-known other descriptions.Such a pattern would reflect the lack of an efficient cognitive schema for processing personal information about an unfamiliar other.

PARTICIPANTS
Ninety students (61 females and 29 males) enrolled in introductory psychology courses at the University of Western Ontario participated.The mean age was 19.43, with a range from 18 to 30.
Event Descriptions and Instructions.Each participant provided a total of three written pleasant event descriptions, consisting of one real and one made-up event about either self or a well-known other, and one made-up event about an unfamiliar other (i.e., the experimenter).The real events were to have happened within the past 3 months.The made-up events were to be realistic and plausible, and were also to be written as if they had happened within the last 3 months.The instructions for each event description were as similar as possible, and each emphasized a focus on the specific target for that particular event (i.e., self, well-known other, and unfamiliar other).Participants were instructed to write about the event without stopping, and to write about the feelings and sensations they experienced with the event.
Imagery Ratings.Participants rated the specific degree of visual detail evident for each event by using a 7-point scale, with 1 indicating "none at all", 4 indicating "a moderate amount", and 7 indicating "very much".Participants also rated how easy it was to imagine each event, with this imagability rating made on a 7-point scale, with 1 indicating "not at all", 4 indicating "moderately", and 7 indicating "very easily".Finally, participants also indicated how dynamic their images were by selecting one of the following descriptors for each event: (1) "Still, like a photograph"; (2) "A series of still pictures, like a slide show"; (3) "Moving, like a movie; or (4) "No image is present".
Emotion Ratings.Participants used a 7-point scale to rate the degree to which each event contained positive feelings, with 1 indicating "not at all", 4 indicating "a moderate amount", and 7 indicating "very much".An identically labeled second 7-point rating scale was used to provide a separate rating of the negative feelings associated with each event.
Selection of a Well-Known Other Target.Specific well-know other targets were selected using a brief form that highlights three individuals currently important in each participant's life (Kuiper & Rogers, 1979).Participants indicated the extent to which each of these individuals was known to them, using a 5-point familiarity scale ranging from "Moderately well" (1) to "Extremely well" (5).
Participants also indicated how many hours per week they spent with each of these individuals, and the type of relationship involved (e.g., close friend, boyfriend/ girlfriend, siblings, or parents).The individual with the highest rating on the 5-point scale was then selected as that participant's well-known other target.The mean familiarity rating for the 46 well-known others selected as targets in the present study was 4.87 (sd of .34).If there was a tie in familiarity ratings, then the individual the participant spent the most time with per week was selected as the target.On average, participants spent 28.33 hours per week with their selected targets.Twenty-two of the selected well-known others were close friends, 8 were boyfriends or girlfriends, 7 were mothers, 6 were siblings, and 3 were fathers.

PROCEDURE
Participants were tested in groups of up to 4, with each session lasting approximately 40 minutes.Alternating with each session, participants were placed in either the self or well-known other condition (n's of 44 and 46, respectively).Therefore, participants either provided one real and one made-up description about themselves, or one real and one made-up event description about a well-known other.Each participant (regardless of self or well-known other condition), also provided a made-up event description about an unfamiliar other (the experimenter).The presentation order for these events was varied across sessions.
After reading and signing an informed consent form, participants in the wellknown other condition completed the brief form to select their own well-known other target.Participants were then read the instructions for completing the real and made-up event descriptions for the self and other targets.After each event was written, participants completed the imagery and emotion ratings for that target.Following completion of the final event description and ratings, each participant was given a debriefing form and credit for participation in the study.

RESULTS
Real and Made-up Events: Self versus Well-Known Others.The means and standard deviations for ratings of imagery and emotion components are shown in Table 1.Each measure was analyzed using a 2 (Event: Real, Made-up) × 2 (Target: Self, Well-known Other) analysis of variance (ANOVA).Imagery.The ANOVA for the visual detail rating revealed the expected significant main effect of Event, F = 6.45, p < .05;with real events having significant higher amounts of visual detail than made-up events (respective main effect means of 5.75 vs. 5.20).In a similar fashion, the ANOVA on the imagability ratings also revealed the expected significant main effect of Event, F = 29.88,p < .001.Here, real events were deemed to be easier to imagine than made-up events (with respective means of 6.33 vs. 5.44).For both of these imagery ratings, neither the main effects of Target, nor the two-way interactions were significant, all F's < 2. Thus, self-referent events were not easier to imagine than well-known other events (with respective means of 5.80 vs. 5.98).Nor did self-referent events contain more visual detail than events focusing on a well-known other (respective means of 5.63 vs. 5.33).Furthermore, it was also not the case than any imagability distinctions between self and well-known others emerged as interactions involving Event by Target.Table 2a shows the frequency distribution for the imagery dynamic categories for all of the real versus made-up events.As expected, a significant chisquare analysis, χ2 (3) = 8.93, p < .01indicated that real events were considered more dynamic than made-up events.In particular, real event images were most often viewed as "moving" or being a "series" of images.In contrast, made-up event images were more often seen as being "still" or as a "series" of images; and less often as "moving".This pattern supports a clear distinction between real and madeup events in terms of imagery dynamics.
Two further chi-square analyses were then performed to investigate whether the self or well-known other nature of the target had any impact on the imagery dynamics for either the real or the made up events.No significant relationships were found for either the real or made-up distributions.These findings indicate that the imagery dynamics associated with event descriptions about the self could not be reliably distinguished from those about a well-known other, for either type of event.
Emotions.In accord with prior research, a significant main effect of Event was found for positive emotions, F = 23.46,p < .001,with the real events having more positive feelings associated with them than the made-up events (respective main effect means of 6.50 vs. 5.86).Self-referent events, however, did not have more positive feelings than well-known other events (respective means of 6.18 vs. 6.17),F < 1; nor was the two-way interaction significant, F < 1.Finally, the ANOVA for the negative feelings rating revealed that none of the three sources of variance were significant, all F's < 1.This non-significance may reflect the fact that only pleasant events were being described, and thus there may have been an overall paucity of negative feelings associated with these events.3 presents the imagery and emotion ratings for the made-up events only, for all three targets.Two separate sets of paired-sample t-tests were performed on this data: one set for the group of 44 participants that provided made-up events about both themselves and an unfamiliar other (as shown in the left portion of Table 3); and a further set for the remaining 46 participants that provided made-up event descriptions about both a well-know other and an unfamiliar other (right portion of Table 3).

Made-up Events Only: Self or Well-known Others versus Unfamiliar Other
Self vs. Unfamiliar Other.The self-referent made-up events displayed significantly more visual detail than the unfamiliar other events, t (43) = 4.44, p < .001,and were significantly easier to imagine, t (43) = 4.17, p < .001.Finally, greater positive emotions were also associated with the self-referent made-up events, when compared with the unfamiliar other made-up events, t (43) =3.86, p < .001.In summary, the made-up events about the self were clearly distinguishable from those about an unfamiliar other, with the only non-significant comparison involving negative feelings.204 Well-known Other vs. Unfamiliar Other.The t-test results for these comparisons are shown in the right panel of Table 3. Well-known other events were rated as being easier to imagine, t (45) = 4.60, p < .001,but did not contain significantly more visual detail than the made-up events about an unfamiliar other.The wellknown other made-up events also had significantly more positive emotions than the unfamiliar other events, t (45) = 4.46, p < .001,but did not display any difference in terms of negative feelings.Overall, these findings suggest that although wellknown other made-up events can still be distinguished from made-up events about an unfamiliar other, this distinction is not quite as clear as that found for selfreferent versus unfamiliar other made-up events.
Imagery Dynamics for Made-up Events.A preliminary chi-squared analysis revealed that the distribution of imagery dynamic categories for the made-up events was not contingent upon the self versus well-known other nature of the target.Accordingly, Table 2b presents the frequency distribution of imagery dynamic categories for the combined self and well-known other targets versus the unfamiliar other made-up events.A chi-square analysis on this distribution revealed a significant relationship, χ 2 (3) = 12.55, p < .01,such that made-up events about the self and well-known other were much more dynamic relative to made-up events about an unfamiliar other.For example, participants reported considerably more "moving" images for the self and well-known other than for an unfamiliar other.Conversely, there were almost twice as many "still" images for the unfamiliar other target (i.e., the experimenter) than for the combined self and well-known other targets.Finally, it should be noted that participants had considerably more difficulty in generating images for the unfamiliar other than for the self or wellknown other event descriptions (i.e., 12 vs. 4 "no images").

DISCUSSION
Employing the multiple-systems model of autobiographical memory (Rubin 2005), the present study examined imagery and emotion component ratings for real and made-up events about self and various others.In doing so, we found that real pleasant events were easier to imagine and contained more visual detail and positive emotions than made-up events.This pattern is consistent with the findings reported by Kealy et al. (2006) for both pleasant and unpleasant events about the self.These findings are also congruent with research demonstrating that real autobiographical memories can be clearly distinguished from memories derived from secondary sources, such as other family members (Crawley & Eacott, 2006).As such, these findings provide additional support for a basic tenet of the multiplesystems model, namely that different categories of memories or recollections are reliably distinct from one another in terms of associated imagery components and emotional feelings (Rubin, 2005).
The present findings also extend the multiple-systems model to incorporate consideration of imagery dynamics.Several different qualities of imagery are already recognized in this model, including vividness, visual detail, spatial elements, and visual perspective (Rubin, 2005;Rubin et al., 2003).The present findings suggest that a further aspect of imagery, namely, its dynamic versus static quality, may also play an important role in distinguishing between various types of event descriptions and memories.In particular, our findings indicated that images of real events were viewed as being much more dynamic in nature, and thus have much more perceived movement (i.e., being like a movie) than images of made-up events (i.e., being like a still picture).
Interestingly, we did not find any distinctions in the various ratings for self versus well-known others.Event descriptions for both of these targets were found to be equivalent in terms of imagability, visual detail, imagery dynamics, and positive emotions.This equivalence fails to support the notion that self-referent processing is distinct or unique (e.g., Kircher et al., 2000;Symons & Johnson, 1997).If this had been the case, we would have expected the most elaboration for the self-referent condition, resulting in higher imagery and emotion ratings, when compared with well-known others.Instead, the equivalence of findings supports the proposal that efficient cognitive schemata are involved in the processing of information about both the self and well-known others (Kuiper & Rogers, 1979).These rich cognitive schemata provide detailed background knowledge and wellorganized structures for processing further information for both types of targets.
Differences between self and others only emerged when considering the comparisons involving made-up events about an unfamiliar other (the experimenter).
Here, the self-referent events had significantly higher ratings for all of the measures, except negative emotions.Thus, participants viewed the made-up events about the self as having more visual detail, being easier to imagine, being more dynamic and being more positive, than their made-up events about an unfamiliar other.This pattern further supports the proposal that an efficient cognitive schema is involved in the elaboration of made-up events about the self; and that, in contrast, the processing of information about unfamiliar others is relatively inefficient and is not associated with a person-specific schema (Kuiper & Rogers, 1979;Symons & Johnson, 1997).Thus, the dearth of specific background information about the unfamiliar target, coupled with the lack of a distinct cognitive structure for organizing and processing information about this individual, would limit the degree to which a made-up event about an unfamiliar other could be richly imagined or emotions could be relived.
Additional findings from our study speak even more directly to the issue of familiarity.In particular, several significant differences emerged between the two other-referent targets, with the well-known other event descriptions being rated as easier to image and having more positive emotions than the made-up event descriptions about an unfamiliar other.Once again, this highlights the contrast between the use of efficient cognitive schemata for well-known others and much less efficient or elaborate processing for unfamiliar others.Interestingly, however, there were no distinctions in terms of visual details.As such, there may be a certain degree of similarity in visual details for other-referent targets ranging across a wide range of familiarity.
Although the present findings are informative, there are several limitations which should be acknowledged and addressed in future research.To begin, the present study focused only on pleasant events.Thus, it may still be the case that negative self-referent events may differ from negative events about a well-known other in terms of the various components and beliefs associated with the multiplesystems model.In particular, the increased stress associated with personally relevant events may foster enhanced imagery, along with greater negative emotions, when compared with unpleasant events about a well-known other.Furthermore, it is not yet known if more distant past events (i.e., more than several years old: see Sporer & Sharman, 2006) would lead to self versus well-known other differences in the various components of the multiple-systems model.Finally, it would also be of interest to move beyond a university sample to consider an older, community based sample that may also display a different pattern of self-other similarities and differences.
a) For Real versus Made-up Events (Combining Self & Well-known Others) b) For Made-up Events Only (Self & Well-known Others versus Unfamiliar Other)

Table 1
Event Component Ratings: Means (and Standard Deviations) by Event Type and Target

Table 2
Distribution of Imagery Dynamic Categories

Table 3
Made-up Event Component Ratings: Means (Standard Deviations) and t-test p-values