Journal of Personality and Social Psychology 1990, Vol. 59, No. 2, 298-304
Copyright 1990 by the Amei
n Psychological Association, Inc. 0022-3514/90/$00.75
Victim Attributions and Post-Rape Trauma
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Patricia A. R-azier University of Missouri—Columbia
This study assessed (a) the kinds of attributions victims make, (b) whether behavioral and characterological self-blame are associated with other variables as hypothesized (e.g, perception of future avoidability of being raped), and (c) whether behavioral self-blame is associated with better postrape adjustment (Janoff-Bulman, 1979). Attributions and adjustment were assessed in a sample of adult female rape victims seen at a hospital-based rape crisis program. Many victims blamed themselves but tended to place more blame on external factors. The pattern of relations between behavioral and characterological self-blame and other attributional measures did not support the hypothesized distinctions between them. Both kinds of self-blame were significantly associated with increased post-rape depression (all ps < .05). Attributions strongly predicted adjustment, accounting for up to 67% of the variance in 3-day post-rape depression. The theoretical and clinical implications of these findings are discussed.
Numerous studies have documented common patterns of reactions to rape, as well as factors that are related to individual differences in post-rape trauma. One factor thought to be associated with post-rape distress is the attribution made by the victim about the cause of the rape. Although self-blame is typically assumed to be an unhealthy response (eg, Becker, Skinner, Abel, Howell, & Bruce, 1982; Burgess & Holmstrom, 1974; Notman & Nadelson, 1976; Rose, 1986), it has also been suggested that engaging in self-blame may be a very adaptive strategy in the aftermath of a rape (Janoff-Bulman, 1979). According to this hypothesis, victims who blame themselves may cope more successfully, provided that the blame is directed at specific, controllable behaviors. This behavioral self-blame is thought to be related to better adjustment because victims may feel that, by changing those behaviors, they can avoid being raped in the future. On the other hand, characterological selfblame, which involves attributions to stable and uncontrollable
aspects of the self, is seen as maladaptive, because it does not provide the same sense of control. This theory presents an interesting contrast to other attributional approaches in the literature. For example, in their attributional theory of depression, Abramson, Seligman, and Teasdale (1978) proposed that internal, stable, and global attributions regarding negative events are associated with increased depression. However, Janoff-Bulman and Lang-Gunn (1988) have suggested that internal, unstable, and specific attributions may be most adaptive for rape victims. Thus, whereas Abramson et al. postulated that internal attributions for negative events are associated with increased depression, Janoff-Bulman and her colleagues hypothesized that internal attributions can be adaptive. The clinical implications of the theory are also substantial because the idea that self-blame can be an adaptive response is inconsistent with current approaches to counseling victims of rape (see Katz & Mazur, 1979). Treatment strategies typically focus on providing a nonjudgmental atmosphere for victims that actively discourages self-blame. If, however, behavioral self-blame is associated with greater feelings of control and better adjustment, counselors might need to reconsider this treatment strategy. Despite its importance for both theory and practice, the hypothesis that behavioral self-blame is associated with better adjustment has received relatively little empirical investigation. In a first effort, Janoff-Bulman (1979) surveyed rape crisis counselors regarding self-blame among victims. According to their estimates, 74% of rape victims blame themselves, at least in part, for the assault, with behavioral self-blame (69%) estimated as being more common than characterological self-blame (19%) (see also Libow & Doty, 1979). On the basis of these data, Janoff-Bulman concluded that most victims engage only in behavioral self-blame and that behavioral self-blame represents a functional response. She also speculated that behavioral selfblame may be adaptive only when it is not combined with characterological self-blame. However, the fact that behavioral selfblame is rated by counselors as more common than characterological self-blame does not necessarily imply that it is adaptive.
Portions of this article were presented at the 1988 meeting of the Midwestern Psychological Association in Chicago. Preparation of this research was supported in part by National Research Service Award 1 F31 MH09498-01 from the National Institute of Mental Health and by an Eva O. Miller Doctoral Dissertation Grant from the University of Minnesota to Patricia A. Frazier. This article is based on Patricia A. Frazier's doctoral dissertation submitted to the University of Minnesota and awarded the 1989 American-Psychology Law Society Dissertation Award. I wish to thank Eugene Borgida, my doctoral advisor, for his support, as well as the other members of my doctoral committee: Ellen Berscheid, Jo-Ida C. Hansen, Geoffrey Maruyama, and Patricia McCarthy. Special thanks is also extended to Linda Ledray and the Sexual Assault Resource Service staff for their assistance; to Julie Haugen, Susan Johnson, and Rhonda Thompson for their help with data collection and coding; and to Craig A. Anderson, Eugene Borgida, David Funder, P. Paul Heppner, and three anonymous reviewers for their helpful comments on an earlier draft of this article. Correspondence concerning this article should be addressed to Patricia A. Frazier, who is now at the Department of Psychology, Elliott Hall, University of Minnesota, Minneapolis, Minnesota 55455.
298
299
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
VICTIM ATTRIBUTIONS
In another test of the behavioral self-blame hypothesis, undergraduate women read an account of a rape, imagined that the incident had happened to them, and made attributions accordingly (Janoff-Bulman, 1982). Within this "victim" group, behavioral self-blame was associated with higher self-esteem and characterological self-blame was associated with an external locus of control. In addition, onty behavioral self-blame was associated with a perception that future rapes could be avoided. These results were interpreted in terms of the adaptiveness of behavioral self-blame, although adjustment was not assessed in actual victims. To date, only one study (Meyer & Taylor, 1986) has specifically examined the relation between behavioral and characterological self-blame and post-rape adjustment in a sample of rape victims. Correlations between their attribution factors and symptom measures indicated that both the behavioral and characterological self-blame factors were associated with poorer post-rape adjustment. Meyer and Taylor concluded that their data supported the distinction between the two types of self-blame, although they also noted anecdotal ly that victims did not always seem to view character traits as unchangeable (see also Abbey, 1987). Thus, victims may not make the distinctions between behavior and character that Janoff-Bulman's (1979) theory implies. Katz and Burt (1988) also examined the relation between self-blame and post-rape adjustment in a rape victim sample, although they did not distinguish between behavioral and characterological blame. Correlations between self-blame and postrape functioning indicated that initial (but not current) selfblame was associated with several measures of adjustment, such as fear, depression, and self-esteem. In all cases, self-blame was associated with poorer functioning. Because behavioral and characterological self-blame were not distinguished, these findings do not address Janoff-Bulman's theory, although they are consistent with the results of the Meyer and Taylor (1986) study. In summary, the available data suggest that all kinds of selfblame are related to increased trauma post-rape. However, only two studies have assessed attributions in actual victim samples, and only one attempted to distinguish between behavioral and characterological self-blame. In that study, the two types of blame were not directly assessed, and the hypothesis that behavioral self-blame is adaptive only in the absence of characterological self-blame was not tested. In addition, proposed mediators of the relation between behavioral blame and adjustment have not been measured. Specifically, behavioral blame is hypothesized to be adaptive because it is associated with a belief that future rapes can be avoided. Finally, behavioral and characterological self-blame have not been related to the attributional dimensions of internality, stability, globality (Abramson et al, 1978), and controllability (Anderson & Arnoult, 1985). Making these links would help to clarify the meaning of the two types of self-blame. This study had three primary purposes. The first purpose was to gather more descriptive information about the kinds of attributions victims make about the causes of rape. Specifically, it seemed important to assess the frequency with which victims blame themselves prior to examining the relation between selfblame and adjustment. Anecdotal data suggest that self-blame
is quite common, but to date there is little empirical data regarding the extent to which victims blame themselves or other factors. The second purpose of the study was to assess whether behavioral and characterological self-blame are associated with other attribution variables as predicted by Janoff-Bulman's (1979) model. As mentioned, prior research in this area has not assessed factors such as beliefs about future avoidability that are hypothesized to mediate the relation between behavioral selfblame and adjustment. The final purpose of the study was to test Janoff-Bulman's hypothesis that behavioral self-blame is associated with better adjustment for victims of rape.
Method Sample and Procedure The participants in this study were part of an ongoing research program at the Sexual Assault Resource Service (SARS) at Hennepin County Medical Center in Minneapolis, Minnesota. The SARS program is staffed by rape crisis nurses who are contacted whenever a rape victim reports to the emergency room. The SARS nurse meets with the victim at the hospital and for up to 1 year post-assault. Assessment measures are completed by the nurses and victims at each of their meetings. The SARS program has been collecting information on their clients since January 1981. Although the specific assessment protocols have changed since that time, the SARS staff routinely collects data on victim characteristics (e.g^ demographics, prior victimization), assault characteristics (e.g., relationship to assailant), post-assault characteristics (e.g., social support), and post-rape adjustment (eg., depression; see Frazier, 1988, for more details). Because the primary goal of this study was to assess the relation between attributions and adjustment, permission was obtained to add a questionnaire assessing attributions about the rape to the SARS protocol. These data were collected from April 1986 to August 1987. Specifically, the SARS nurses were asked to give the attribution questionnaires to their clients at their 3-day, 3-month, and 6-month visits. Clients were asked to complete the questionnaires and mail them back to the SARS office. Altogether, 42% of the 234 questionnaires handed out were returned (n = 98). However, because many victims did not wish to continue with counseling beyond the immediate post-rape period, which is very typical of victims in this setting, most of the questionnaires were returned after the 3-day meeting Ifi - 67). Therefore, the characteristics of this subsample, rather than the entire SARS sample (A1 = 1,420), will be described. It should also be noted at this point that, because this study draws on an archival data set, sample sizes for specific analyses vary somewhat due to missing data. All victims in this subsample were female, at least 16 years of age, and had experienced a completed act of oral, anal, or vaginal penetration. They were primarily young (Af = 27 years of age, SD = 13.59), White (81%), and single (85%). Many of the victims had experienced a prior rape (40%) or incest (27%). Slightly more than half (56%) had been raped by strangers. DemographicaUy, this subsample is comparable to others described in the literature and to the larger SARS sample (Frazier, 1988).
Measures Victim ratings of attributions. Victims responded to two specific questions designed to assess behavioral and characterological selfblame. As a measure of behavioral self-blame, victims were asked, "How much do you blame things you did before the rape (e.g., walking alone at night)?" The characterological self-blame question was, "How
300
PATRICIA A. FRAZIER
Table 1 Victim Ratings of Attributions at 3-Days Post-Rape
M
SD
Behavioral self-blame' Characterological self-blame*
2.64 2.62
.45 .38
Frequency of thinking about "why"b
3.66
.21
Self Rapist* Chance" Society"
2.27 4.62 3.15 3.02
.43 9.81 .29 .30
Internal5 Stable* Global1 Uncontrollable'
2.27 3.00 3.10 3.82
.40 .36 .38 .31
Future avoidability' Past avoidability*
3.66 3.07
.40 .75
Attribution rating
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ioral and Characterological self-blame ratings suggested that the victims in this sample engaged in these two types of self-blame equally. Although most victims stated that they blamed both their behavior and character to some degree, the modal response to both questions was "not at all to blame." The additional attributional ratings indicated that the victims frequently thought about why the rape occurred, which underscored the importance of an attributional analysis in this domain. More specifically, ratings of the four causes suggested that victims blamed themselves the least and blamed the rapist the most. Similarly, scores on the four attributional dimensions indicated that the causes of the rape were generally rated as external and uncontrollable. The mean responses to questions regarding past and future avoidability indicated that most victims believed they could avoid being raped in the future and that the past rape had been avoidable. Factor analysis.
Note, n = 67. • Scale = not at all (1) to completely blame (5); * Scale = never (1) to all the time (5); 'Scale = external (\)\o internal^); aScale = unstable (1) lo stable (5); 'Scale = specific (I) lo global (5); ' Scale = controllable (1) to uncontrollable (5); * Scale = no (!) to yes (5).
To assess behavioral and Characterological
self-blame in another way, responses to the 15 attribution statements were factor analyzed using principal-axis factoring with iteration.2 Factors with eigenvalues greater than 1 were rotated using varimax rotation. Initially, six factors met this criterion; however, a three-factor solution was ultimately used that was more interpretable and corresponded well to the factors in the
much do you blame things about your personality (e.g., being too trusting) that you feel you can't change?" Ratings were made on 5-point scales—not at all (1) to completely blame (5). In the interest of replicating the Meyer and Taylor (1986) study, the 15 statements with significant loadings on their attribution factors were also included. Victims rated the importance of these statements in helping them to explain why they were raped (e.g, I should have been more cautious; completely false (1) to completely true (5). To gather additional descriptive data, the victims made several other attributional ratings. First, in an effort to assess potential mediators of the attribution/adjustment relation, victims rated (on 5-point scales) whether they felt a future rape could be avoided and whether they felt the past rape could have been avoided. Second, victims rated (on 5-point scales) how much they blamed the rape on each of four specific causes that were most commonly reported in prior research (i.e, self, rapist, chance, society). Third, victims were asked an openended question about the cause of the rape and were asked to rate that cause on the four attributional dimensions of internality, stability, globality, and controllability on separate 5-point scales. Finally, victims rated how often they thought about why they were raped—never (1) to all the time (5). Adjustment measure. The Beck Depression Inventory (BDI; Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961) was chosen as the primary measure of adjustment.1 The BDI is a 21-item self-report measure of depressive symptoms, with each item containing four symptom statements ordered in increasing severity. For each item, respondents selected the statement that best described their feelings during the past week. Prior research suggests that the BDI has good internal consistency reliabilityfc
Copyright 1990 by the Amei
n Psychological Association, Inc. 0022-3514/90/$00.75
Victim Attributions and Post-Rape Trauma
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Patricia A. R-azier University of Missouri—Columbia
This study assessed (a) the kinds of attributions victims make, (b) whether behavioral and characterological self-blame are associated with other variables as hypothesized (e.g, perception of future avoidability of being raped), and (c) whether behavioral self-blame is associated with better postrape adjustment (Janoff-Bulman, 1979). Attributions and adjustment were assessed in a sample of adult female rape victims seen at a hospital-based rape crisis program. Many victims blamed themselves but tended to place more blame on external factors. The pattern of relations between behavioral and characterological self-blame and other attributional measures did not support the hypothesized distinctions between them. Both kinds of self-blame were significantly associated with increased post-rape depression (all ps < .05). Attributions strongly predicted adjustment, accounting for up to 67% of the variance in 3-day post-rape depression. The theoretical and clinical implications of these findings are discussed.
Numerous studies have documented common patterns of reactions to rape, as well as factors that are related to individual differences in post-rape trauma. One factor thought to be associated with post-rape distress is the attribution made by the victim about the cause of the rape. Although self-blame is typically assumed to be an unhealthy response (eg, Becker, Skinner, Abel, Howell, & Bruce, 1982; Burgess & Holmstrom, 1974; Notman & Nadelson, 1976; Rose, 1986), it has also been suggested that engaging in self-blame may be a very adaptive strategy in the aftermath of a rape (Janoff-Bulman, 1979). According to this hypothesis, victims who blame themselves may cope more successfully, provided that the blame is directed at specific, controllable behaviors. This behavioral self-blame is thought to be related to better adjustment because victims may feel that, by changing those behaviors, they can avoid being raped in the future. On the other hand, characterological selfblame, which involves attributions to stable and uncontrollable
aspects of the self, is seen as maladaptive, because it does not provide the same sense of control. This theory presents an interesting contrast to other attributional approaches in the literature. For example, in their attributional theory of depression, Abramson, Seligman, and Teasdale (1978) proposed that internal, stable, and global attributions regarding negative events are associated with increased depression. However, Janoff-Bulman and Lang-Gunn (1988) have suggested that internal, unstable, and specific attributions may be most adaptive for rape victims. Thus, whereas Abramson et al. postulated that internal attributions for negative events are associated with increased depression, Janoff-Bulman and her colleagues hypothesized that internal attributions can be adaptive. The clinical implications of the theory are also substantial because the idea that self-blame can be an adaptive response is inconsistent with current approaches to counseling victims of rape (see Katz & Mazur, 1979). Treatment strategies typically focus on providing a nonjudgmental atmosphere for victims that actively discourages self-blame. If, however, behavioral self-blame is associated with greater feelings of control and better adjustment, counselors might need to reconsider this treatment strategy. Despite its importance for both theory and practice, the hypothesis that behavioral self-blame is associated with better adjustment has received relatively little empirical investigation. In a first effort, Janoff-Bulman (1979) surveyed rape crisis counselors regarding self-blame among victims. According to their estimates, 74% of rape victims blame themselves, at least in part, for the assault, with behavioral self-blame (69%) estimated as being more common than characterological self-blame (19%) (see also Libow & Doty, 1979). On the basis of these data, Janoff-Bulman concluded that most victims engage only in behavioral self-blame and that behavioral self-blame represents a functional response. She also speculated that behavioral selfblame may be adaptive only when it is not combined with characterological self-blame. However, the fact that behavioral selfblame is rated by counselors as more common than characterological self-blame does not necessarily imply that it is adaptive.
Portions of this article were presented at the 1988 meeting of the Midwestern Psychological Association in Chicago. Preparation of this research was supported in part by National Research Service Award 1 F31 MH09498-01 from the National Institute of Mental Health and by an Eva O. Miller Doctoral Dissertation Grant from the University of Minnesota to Patricia A. Frazier. This article is based on Patricia A. Frazier's doctoral dissertation submitted to the University of Minnesota and awarded the 1989 American-Psychology Law Society Dissertation Award. I wish to thank Eugene Borgida, my doctoral advisor, for his support, as well as the other members of my doctoral committee: Ellen Berscheid, Jo-Ida C. Hansen, Geoffrey Maruyama, and Patricia McCarthy. Special thanks is also extended to Linda Ledray and the Sexual Assault Resource Service staff for their assistance; to Julie Haugen, Susan Johnson, and Rhonda Thompson for their help with data collection and coding; and to Craig A. Anderson, Eugene Borgida, David Funder, P. Paul Heppner, and three anonymous reviewers for their helpful comments on an earlier draft of this article. Correspondence concerning this article should be addressed to Patricia A. Frazier, who is now at the Department of Psychology, Elliott Hall, University of Minnesota, Minneapolis, Minnesota 55455.
298
299
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
VICTIM ATTRIBUTIONS
In another test of the behavioral self-blame hypothesis, undergraduate women read an account of a rape, imagined that the incident had happened to them, and made attributions accordingly (Janoff-Bulman, 1982). Within this "victim" group, behavioral self-blame was associated with higher self-esteem and characterological self-blame was associated with an external locus of control. In addition, onty behavioral self-blame was associated with a perception that future rapes could be avoided. These results were interpreted in terms of the adaptiveness of behavioral self-blame, although adjustment was not assessed in actual victims. To date, only one study (Meyer & Taylor, 1986) has specifically examined the relation between behavioral and characterological self-blame and post-rape adjustment in a sample of rape victims. Correlations between their attribution factors and symptom measures indicated that both the behavioral and characterological self-blame factors were associated with poorer post-rape adjustment. Meyer and Taylor concluded that their data supported the distinction between the two types of self-blame, although they also noted anecdotal ly that victims did not always seem to view character traits as unchangeable (see also Abbey, 1987). Thus, victims may not make the distinctions between behavior and character that Janoff-Bulman's (1979) theory implies. Katz and Burt (1988) also examined the relation between self-blame and post-rape adjustment in a rape victim sample, although they did not distinguish between behavioral and characterological blame. Correlations between self-blame and postrape functioning indicated that initial (but not current) selfblame was associated with several measures of adjustment, such as fear, depression, and self-esteem. In all cases, self-blame was associated with poorer functioning. Because behavioral and characterological self-blame were not distinguished, these findings do not address Janoff-Bulman's theory, although they are consistent with the results of the Meyer and Taylor (1986) study. In summary, the available data suggest that all kinds of selfblame are related to increased trauma post-rape. However, only two studies have assessed attributions in actual victim samples, and only one attempted to distinguish between behavioral and characterological self-blame. In that study, the two types of blame were not directly assessed, and the hypothesis that behavioral self-blame is adaptive only in the absence of characterological self-blame was not tested. In addition, proposed mediators of the relation between behavioral blame and adjustment have not been measured. Specifically, behavioral blame is hypothesized to be adaptive because it is associated with a belief that future rapes can be avoided. Finally, behavioral and characterological self-blame have not been related to the attributional dimensions of internality, stability, globality (Abramson et al, 1978), and controllability (Anderson & Arnoult, 1985). Making these links would help to clarify the meaning of the two types of self-blame. This study had three primary purposes. The first purpose was to gather more descriptive information about the kinds of attributions victims make about the causes of rape. Specifically, it seemed important to assess the frequency with which victims blame themselves prior to examining the relation between selfblame and adjustment. Anecdotal data suggest that self-blame
is quite common, but to date there is little empirical data regarding the extent to which victims blame themselves or other factors. The second purpose of the study was to assess whether behavioral and characterological self-blame are associated with other attribution variables as predicted by Janoff-Bulman's (1979) model. As mentioned, prior research in this area has not assessed factors such as beliefs about future avoidability that are hypothesized to mediate the relation between behavioral selfblame and adjustment. The final purpose of the study was to test Janoff-Bulman's hypothesis that behavioral self-blame is associated with better adjustment for victims of rape.
Method Sample and Procedure The participants in this study were part of an ongoing research program at the Sexual Assault Resource Service (SARS) at Hennepin County Medical Center in Minneapolis, Minnesota. The SARS program is staffed by rape crisis nurses who are contacted whenever a rape victim reports to the emergency room. The SARS nurse meets with the victim at the hospital and for up to 1 year post-assault. Assessment measures are completed by the nurses and victims at each of their meetings. The SARS program has been collecting information on their clients since January 1981. Although the specific assessment protocols have changed since that time, the SARS staff routinely collects data on victim characteristics (e.g^ demographics, prior victimization), assault characteristics (e.g., relationship to assailant), post-assault characteristics (e.g., social support), and post-rape adjustment (eg., depression; see Frazier, 1988, for more details). Because the primary goal of this study was to assess the relation between attributions and adjustment, permission was obtained to add a questionnaire assessing attributions about the rape to the SARS protocol. These data were collected from April 1986 to August 1987. Specifically, the SARS nurses were asked to give the attribution questionnaires to their clients at their 3-day, 3-month, and 6-month visits. Clients were asked to complete the questionnaires and mail them back to the SARS office. Altogether, 42% of the 234 questionnaires handed out were returned (n = 98). However, because many victims did not wish to continue with counseling beyond the immediate post-rape period, which is very typical of victims in this setting, most of the questionnaires were returned after the 3-day meeting Ifi - 67). Therefore, the characteristics of this subsample, rather than the entire SARS sample (A1 = 1,420), will be described. It should also be noted at this point that, because this study draws on an archival data set, sample sizes for specific analyses vary somewhat due to missing data. All victims in this subsample were female, at least 16 years of age, and had experienced a completed act of oral, anal, or vaginal penetration. They were primarily young (Af = 27 years of age, SD = 13.59), White (81%), and single (85%). Many of the victims had experienced a prior rape (40%) or incest (27%). Slightly more than half (56%) had been raped by strangers. DemographicaUy, this subsample is comparable to others described in the literature and to the larger SARS sample (Frazier, 1988).
Measures Victim ratings of attributions. Victims responded to two specific questions designed to assess behavioral and characterological selfblame. As a measure of behavioral self-blame, victims were asked, "How much do you blame things you did before the rape (e.g., walking alone at night)?" The characterological self-blame question was, "How
300
PATRICIA A. FRAZIER
Table 1 Victim Ratings of Attributions at 3-Days Post-Rape
M
SD
Behavioral self-blame' Characterological self-blame*
2.64 2.62
.45 .38
Frequency of thinking about "why"b
3.66
.21
Self Rapist* Chance" Society"
2.27 4.62 3.15 3.02
.43 9.81 .29 .30
Internal5 Stable* Global1 Uncontrollable'
2.27 3.00 3.10 3.82
.40 .36 .38 .31
Future avoidability' Past avoidability*
3.66 3.07
.40 .75
Attribution rating
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ioral and Characterological self-blame ratings suggested that the victims in this sample engaged in these two types of self-blame equally. Although most victims stated that they blamed both their behavior and character to some degree, the modal response to both questions was "not at all to blame." The additional attributional ratings indicated that the victims frequently thought about why the rape occurred, which underscored the importance of an attributional analysis in this domain. More specifically, ratings of the four causes suggested that victims blamed themselves the least and blamed the rapist the most. Similarly, scores on the four attributional dimensions indicated that the causes of the rape were generally rated as external and uncontrollable. The mean responses to questions regarding past and future avoidability indicated that most victims believed they could avoid being raped in the future and that the past rape had been avoidable. Factor analysis.
Note, n = 67. • Scale = not at all (1) to completely blame (5); * Scale = never (1) to all the time (5); 'Scale = external (\)\o internal^); aScale = unstable (1) lo stable (5); 'Scale = specific (I) lo global (5); ' Scale = controllable (1) to uncontrollable (5); * Scale = no (!) to yes (5).
To assess behavioral and Characterological
self-blame in another way, responses to the 15 attribution statements were factor analyzed using principal-axis factoring with iteration.2 Factors with eigenvalues greater than 1 were rotated using varimax rotation. Initially, six factors met this criterion; however, a three-factor solution was ultimately used that was more interpretable and corresponded well to the factors in the
much do you blame things about your personality (e.g., being too trusting) that you feel you can't change?" Ratings were made on 5-point scales—not at all (1) to completely blame (5). In the interest of replicating the Meyer and Taylor (1986) study, the 15 statements with significant loadings on their attribution factors were also included. Victims rated the importance of these statements in helping them to explain why they were raped (e.g, I should have been more cautious; completely false (1) to completely true (5). To gather additional descriptive data, the victims made several other attributional ratings. First, in an effort to assess potential mediators of the attribution/adjustment relation, victims rated (on 5-point scales) whether they felt a future rape could be avoided and whether they felt the past rape could have been avoided. Second, victims rated (on 5-point scales) how much they blamed the rape on each of four specific causes that were most commonly reported in prior research (i.e, self, rapist, chance, society). Third, victims were asked an openended question about the cause of the rape and were asked to rate that cause on the four attributional dimensions of internality, stability, globality, and controllability on separate 5-point scales. Finally, victims rated how often they thought about why they were raped—never (1) to all the time (5). Adjustment measure. The Beck Depression Inventory (BDI; Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961) was chosen as the primary measure of adjustment.1 The BDI is a 21-item self-report measure of depressive symptoms, with each item containing four symptom statements ordered in increasing severity. For each item, respondents selected the statement that best described their feelings during the past week. Prior research suggests that the BDI has good internal consistency reliabilityfc
