Effects of Early Abuse on Adult Affective

EFFECTS OF EARLY ABUSE ON ADULT AFFECTIVE
REACTIONS TO EXPOSURE TO DYADIC CONFLICT

Donald G. Dutton, Lindsey A. Jack, and Annabel N. Webb

Department of Psychology

University of British Columbia

Lee Ryan

University of California, San Diego

Running Head: Early Abuse and Adult Affect

ABSTRACT

To examine the effect of conflict resolution tactics used in the family of origin on current (adult) affective reactions to conflict, 120 college students listened to audiotaped depictions of various family dyadic conflicts. In general, fathers were reported as using the highest rates of physical and verbal abuse in the family of origin while respondents reported themselves as using withdrawal tactics more than did their parents. The total frequency of conflict tactics used in the family of origin correlated significantly with anticipatory anger, anxiety and "subanger" (frustration, irritation) while waiting to hear a family conflict audiotape. Withdrawal, verbal abuse and physical abuse in the family of origin all correlated significantly with anticipatory affect. Conflict tactics used by mothers against fathers or the respondent had the highest correlations with anticipatory affect. Post-exposure impact on affect was weaker. Pre-exposure arousal in groups experiencing high rates of family of origin conflict was nonmonotonic, suggesting that affective flattening and hyperarousal, both aspects of a trauma response, may also be represented in a non-clinical population.

Key Words: Family conflict, emotion, witnessing violence.

EFFECTS OF EARLY ABUSE ON ADULT AFFECTIVE REACTIONS TO WITNESSED DYADIC CONFLICT

In prior studies of witnessed dyadic reactions, Strachan and Dutton (1992) found that both gender and current power position significantly influenced the experience of conflict-generated affect. Specifically, women and persons in low power positions experienced greater anger in response to audiotapes of dyadic conflict about jealousy. Pre-post anger increases were significantly higher for females than for males and were higher for anger than other reported emotions (e.g., anxiety, humiliation, elation, etc.). Highest anger increases for both genders occurred when participants were in a low power position and when an opposite gender person initiated the overheard argument.

Dutton, Webb and Ryan (1994) extended this analysis to witnessed intergenerational dyadic conflicts, replicating the finding that women demonstrated a pre-post anger response double the magnitude of men's. Both men and women anticipated the conflict exposure with moderate anxiety and low anger. Anger increased by 163% for women after exposure, compared to 70% for men. A composite called "sub-anger," comprised of items like irritation and annoyance, also showed larger increases for women. Anxiety, in part because the pre-conflict scores were high, showed less dramatic increases. Nevertheless, post-conflict absolute anxiety scores were extremely high for women (6.97 on a 9 point scale). Dutton et al. (1994) also found that both genders reported greater anger in response to father conflicts than mother conflicts. Depicted conflicts were between either parent and a "child" in their late teens; subjects were college sophomores averaging nineteen years of age. The study led us to speculate that experience with abusive conflict resolution in the family of origin may have long lasting affective sequelae that are manifested upon re-exposure to intimate conflicts in a manner similar to, albeit less extreme than, post-traumatic stress disorder (van der Kolk, 1987).

Examining parent-child interactions where the "child" depicted on the audiotape was demographically similar to the subject (age 19, same sex) raises the question of whether conflict resolution styles in the subject's own family of origin have impact on her/his affective responses to the tapes witnessed at age 19. At present it is not known whether the conflict resolution strategies used in family of origin relationships generate long term affective responses to intimate conflict.

Developmental studies of children's reactions to repeated family conflict indicate that repeated exposure to strong negative emotions like anger becomes, for most children, an aversive experience encompassing strong arousal and behavioral expressions of anxiety (distress) or anger (aggression) (Jaffe, Wolfe & Wilson, 1990). especially when the conflict is not satisfactorally resolved (Davies and Cummings 1994). One mediator of this emotional response is the children's attribution of cause for parental anger (Covell & Abramovitch, 1987). Children who identify more with the angry mother and attribute her anger to themselves are more likely to sustain long term responses to parental anger (Harter, 1982). Studies of older children (reviewed by Jaffe et al., 1990) whose previously assaulted mothers were in shelter houses, and who presumably had been exposed to repeated displays of parental anger and aggression, demonstrated affective disturbances, social-cognitive deficits and problems with peer interaction. Davies & Cummings (1994) reviewed numerous studies of child adjustment to marital conflict and suggested that, based on this literature, it was reasonable to conclude that long term adjustment and regulation of emotional arousal could be adversely affected by long term exposure to parental conflicts. The oldest group studied in the literature they reviewed was 15 to 17 year old children.

Studies Of Adult Affect

Few studies exist which relate childhood experiences to adult affective responding in non-clinical populations. More is known about the intergenerational transmission of abusive behavior (Kalmuss, 1984; Widom, 1989). Since either exaggerated anger or anxiety could impede effective conflict resolution strategies, affect could serve as an important mediating construct that contributes to the intergenerational transmission of abuse (Straus, Gelles & Steinmetz, 1980). Kalmuss (1984) reviewed previous studies on effects of observing mother-father violence and found that both sons and daughters are more likely to be perpetrators and victims of aggression when they have witnessed parental aggression. It is possible that observation produces strong affective responses which increase the likelihood of aggression by being dissonant with effective problem solving. Assaultive males, for example, have poor conflict resolution skills (Dutton & Strachan, 1987) as well as exaggerated anger responses to male-female conflict scenarios (Dutton & Browning, 1988). Since anger increases the likelihood of aggression (Konecni, 1975) and anxiety narrows the range of constructive solutions to perceived conflict (Bower & Bower, 1976), this affective component may shed light on these modelling issues. Browning and Dutton (1986) found that men with histories of wife assault were more likely than control males to experience heightened anger responses (on the Affect Adjective Checklist) to scenes of husband-wife conflict. At present it is not known whether this occurs through specific behavioral modelling or whether childhood events generate chronic long term emotional responses which are themselves risk markers for interpersonal dysfunction and abusive behavior when adult intimate conflict occurs.

The role of affect as a mediating variable between structural aspects of conflict and subsequent behaviors in response to conflict is less well known (see, for example, Frodi, Macaulay & Thome, 1977). Anger is more likely to lead to aggressive tactics of conflict resolution (Konecni, 1975; Rule & Nesdale, 1976); anxiety is believed to lead to withdrawal--the absence of aggressive tactics (Schill & Schneider, 1970). What is not known is the extent to which affective responses to conflict are learned patterns based on prior experiences with specific conflict tactics in certain key relationships. The objective of the current study is to ascertain whether affective responses to depictions of family conflict are influenced by conflict resolution strategies used in the family of origin in college age males and females. A second objective is to ascertain which conflict resolution strategies used in which dyads contribute most to current affective responding.

METHOD

Subjects

One hundred and twenty introductory psychology students, 60 women and 60 men, participated in the study. The mean ages of men and women participants were 19.73 (SD=1.43), and 19.64 (SD=1.98) respectively. Seventy-eight percent of the subject sample rated between 4 and 5 on a 7-point scale of socioeconomic status, and 89% reported their parents to be currently married.

Audiotapes

A set of audio-tapes depicting conflicts between mothers/fathers and their sons/daughters were created specifically for the current study. The conflict scenarios, which were portrayed by professional actors, were written to depict issues that would be both credible and familiar to our subject sample. In one set of conflicts the participants (mother/father and son/daughter) argued about the son/daughter's use of alcohol, and in the other they argued about son/daughter's lifestyle.

Though the scenarios were not scripted, the actors were given written conflict scenarios and character sketches prior to the taping session. The actors were asked to make the conflicts heated and emotionally intense by raising their voices and becoming verbally abusive if necessary. All of the actors were present during the taping of each conflict scenario. They were instructed to pay close attention to the way in which each actor portrayed the conflicts, and were directed to keep the conflicts as similar as possible with respect to verbal content and emotional intensity.

To ensure that the tapes did not differ on intensity or emotionality of the actors, two independent raters rated all participants' emotional intensity using the anger and sub-anger scales of the Affect Adjective Checklist. The correlation between the two raters' scores was .88. There were no rated differences among any of the actors with respect to the amount of anger or hostility that they portrayed; hence, emotional intensity differences between the parents cannot account for any differences in subject ratings.

A second set of tape ratings were completed by eleven independent raters (5 women and 6 men). The raters were randomly assigned to rate mother or father tapes. Raters indicated on nine-point scales how justified they felt the mother/father was in his/her reaction to the son/daughter in the tape, and the degree of seriousness of the argument. They were also asked to choose who, parent or daughter/son, was more convincing with her/his argument.

On all of the tapes except one, the majority of raters (60%+) saw the parents' argument as more convincing. A series of t-tests indicated no significant differences between equivalent (i.e., same parent and same issue) son and daughter tapes (all p>.05) on the degree of seriousness of the conflict and how justified the parent was in her/his reactions.

Males subjects rated the tapes as less realistic (X=6.58) than females (X=7.60); this difference was significant on a t-test (t=3.41, p<.001).

Measures

Conflict in the Family of Origin

The Conflict Tactics Scale (CTS) (Straus, 1979) is a standardized scale designed to measure the frequency and intensity of 19 tactics used in dyads to resolve conflict. The scale includes rational tactics, withdrawal, and a variety of verbally, emotionally and physically abusive strategies. Respondents report both their own use of these tactics and their use by an interactant. For assessment of conflict tactics in the family of origin, dyadic pairs are presented such as father-mother, father-you, and mother-you. Straus et al. (1980) have published population norms for usage of each tactic in a variety of intimate relationships.

For purposes of this study the Conflict Tactics Scale was revised and scored as follows:

Arguments: "argued heatedly but short of yelling"

Verbal Abuse: "insulted, yelled, swore, or tried to hurt the others feelings" and "threatened with violence"

Withdrawal: "sulked, walked out and/or refused to talk about it"

Symbolic Display: "threw, smashed, kicked or hit something"

Minor Physical Abuse: "pushed. grabbed, shoved, slapped"

Major Physical Abuse: "kicked or bit, hit with fist or object, beat up the other one"

For each conflict tactic item, scores of 0 (never), 1 (at least once), 2 (occasionally), 3 (frequently), or 4 (weekly) were assigned. Participants were asked to report upon their own and other family members' use of conflict tactics during the entire period of childhood and adolescence. CTS reports were obtained immediately after demographic information. Each subject completed the CTS in a private cubicle.

Emotional Response Scales

The Affect Adjective Checklist (AAC) (Russell & Mehrabian, 1974) was used to assess subjects' emotional responses. This scale consists of sixteen bipolar affect adjective pairs, with affect adjectives anchored differentially on a 9-point scale. Higher scores on the AAC indicate more intense emotional responses.

Russell and Mehrabian's (1974) anger and anxiety composites were used, along with our own composites for sub-anger and arousal. Our sub-anger composite was included to classify anger descriptors that are more inward-directed and less behavioral than those contained in the Russell and Mehrabian anger composite. AAC items were totalled in the following manner: items "angry", "aggressive" and "hostile" were totalled for anger; "tense", "anxious" and "nervous" for anxiety; "annoyed", "frustrated" and "irritated" for the composite sub-anger; and "aroused" and "excited" were totalled for arousal. A single-item humiliation score was also included. Total affect scores are the sum of anger, anxiety, subanger and arousal.

Power and Identification - Perceptual Measures

For every conflict dyad, subjects were asked to rate separately the degree of power held by each participant, the degree to which they identified with each of the participants in the conflict, and their familiarity and experience with this type of conflict. Subjects indicated their responses to these perceptual questions on nine-point rating scales.

Demographic Information

Subjects were given a set of demographic questions that measured sex, age, current relationship status, socioeconomic status (SES), number of siblings, birth order, and parents' marital status.

Design and Procedure

Each experiment consisted of a factorial design, with parent as the within-subject factor, other variables as the between-subject factors. Each subject heard two conflicts, one involving a mother and son or daughter (mother tapes) and the other involving a father and son or daughter (father tapes). The two conflict issues were the son/daughter's use of alcohol, and the son/daughter's lifestyle. The tapes were fully counterbalanced with respect to actor and order. Thus half of the subjects heard mother tapes first, and half heard father tapes first, and none of the subjects heard the same actor twice.

In order to maximize subjects' identification with the sons/daughters in the conflict tapes, we matched subject and actor sex, so that female subjects listened only to conflicts involving daughters (female tapes) and male subjects listened to only conflicts involving sons (male tapes). Furthermore, the sons and daughters in the tapes were demographically similar to our subject sample.

The revised CTS was administered to undergraduate psychology students. A score for Verbal Abuse, Withdrawal and Physical Abuse (symbolic display, minor physical abuse and major physical abuse) was calculated for each interactant in each dyad pair. Dyad pairs included dad/mom, dad/you, and mom/you. Conflict tactics reported between siblings were not measured. Scores were summed across dyad pairs to create total scores for Verbal Abuse, Withdrawal and Physical Abuse.

The experimenter telephoned students who had completed the revised CTS and had indicated an interest in the second portion of the study. She gave them a brief description of the study and asked them if they were interested in participating. She told them that the experiment examined peoples' perceptions of interpersonal behavior and that it would involve listening to an audiotape of interactions between two people and then filling out questionnaires in response to the tape. Both male and female experimenters were used in this study. There were no significant differences in subjects' responding as a function of the sex of the experimenter.

Prior to subjects' listening to the conflict tapes, the experimenter described the procedure and explained the origins of the tapes and warned the subject that the taped interactions might involve conflicts and the use of profane language. The subjects were told the tapes were borrowed from another researcher who had used tape recorders in the homes of families in which one family member was undergoing treatment in order to record clients' interactive behavior in a natural environment. The experimenter told subjects they would be hearing two separate conflict segments which, for the purpose of this study, had been transcribed onto one tape. Subjects were deceived about the origins of the audiotapes in order to prevent any reporting biases or inhibitions that might result if subjects had known the tapes were staged.

The experimental procedure involved having groups of one to four subjects (who were seated in individual cubicles) fill out questionnaires prior to, and in response to, two audiotaped conflicts. Experimenters were blind to the subjects' scores on the revised CTS. Preliminary measures included the demographic information questionnaire and emotional response scale (AAC). To enhance subjects' ability to accurately report affect, a set of focussing instructions were presented directly prior to the AAC. Subjects were asked to shut out distractions, close their eyes, take a deep breath, and ask themselves "What am I feeling right now?". Once they had completed the first AAC, and prior to listening to the conflict tapes, the subjects were given a set of perspective-taking instructions which directed them to try to assume the perspective of the son/daughter in the conflict (see Davis, Hull, Young, & Warren, 1987).

Subjects listened to conflict tapes on Sony Walkman (model # WM-A53). In order to heighten subjects' arousal levels the volume was set at approximately 3/4 of the maximum. Subjects were asked not to adjust the volume unless they found the conflicts so loud that they could not listen. Following the first conflict scenario, subjects were instructed to repeat the focussing procedure, complete another AAC and the first set of perceptual questions. Subjects then completed the second set of perspective taking instructions, listened to the second conflict tape, completed another emotional response scale, and another set of perceptual questions. Subsequent to questionnaire completion, the experimenter asked participants to verbally rate how realistic the tapes were on a scale of 1 (unrealistic) to 9 (very realistic). The experimenter then asked subjects to briefly specify their reasons for identifying with particular persons in the conflict.

Subjects were given a complete oral and written debriefing, and their questions regarding the purpose of the study were answered by the experimenter.

RESULTS

Participants' ratings of degree of power held by each individual in the dyad, the degree of identification with each individual, and participants' experience and familiarity with the type of dyadic conflict were not significantly correlated with post-conflict Affect Adjective Checklist scores. Additionally, the power and identification measures were not significantly related to specific Conflict Tactics Scale scores, or conflict tactic use within specific dyads. Table 1 demonstrates frequencies for various forms of abuse reported by our college student sample. Both verbal abuse and withdrawal are conflict tactics reported by this group as being used on a weekly basis in the family of origin. The reported abuse scores within each of the six dyad pairs were summed to generate a total score for each type of abuse (Verbal Abuse, Withdrawal, Symbolic Display, Minor Physical, Major Physical). The theoretical minimum and theoretical maximum for each total abuse score is zero and twenty-four, respectively. Total mean scores and standard deviations for various forms of abuse across all family of origin dyad pairs are shown in the top one of Table 1.

_________________________

Insert Table 1 about here

_________________________

MANOVA was utilized to compare the means and standard deviations of the conflict tactics (Verbal Abuse, Withdrawal, Symbolic Display, Minor Physical Abuse and Major Physical Abuse) across specific dyads. Father to mother conflict tactics were compared with mother to father dyad tactics. Additionally, father-you dyad tactics were compared to you-father dyad tactics and, mother-you dyad tactics were compared with you-mother dyad tactics. The results are presented in Table 2. A Bonferroni procedure resulted in an alpha level for significance of .p<.02.

__________________________

Insert Table 2 about here

__________________________

Table 3 demonstrates Pearson correlations between Conflict Tactics Scale measures of violence in the family of origin and current Affect Adjective Checklist composites. A Bonferroni procedure resulted in an alpha level for significance set at p < .02 which required an r of .20 or greater. Withdrawal, Verbal and Physical Abuse scores for the family of origin are all significantly and positively correlated with pre-conflict anxiety, anger, subanger and total affect scores. All preconflict affect scores were significantly and positively intercorrelated. In effect, they form a composite for negative affect comprised of arousal, anger, subanger and anxiety. The average Pearson correlation between total affect and each of these components was .80.

__________________________

Insert Table 3 about here

__________________________

Table 4 demonstrates correlations of General Conflict Tactics Scale scores with post-conflict Affect Adjective Checklist scores after General Witnessed Family Conflict. As with preconflict affect scores, post conflict affect scores formed an intercorrelated composite for negative affect, again comprised of arousal, anger, subanger and anxiety. The average correlation of the total affect scores to each of these was .82. With Bonferroni alpha set at P <.02, none of the correlations were significant.

_________________________

Insert Table 4 about here

__________________________

Table 5 shows Pearson correlations between total affect experienced while anticipating family conflict and specific types of withdrawal or abuse in specific family dyads were computed. For low vs. high abuse comparisons, Bonferroni alpha is set at p <.02.

A Fischer z statistic (z=2.348, p<.05) reveals that the Pearson correlation between total affect and Withdrawal (r=.42) differs significantly from the Pearson correlation between total affect and Major Physical Abuse (r=.14). No other significant differences existed between the other forms of conflict tactics.

In order to more precisely focus on the effects of high levels of abuse in the family of origin, two extreme groups were established called and high and low abuse groups. T-tests were performed and on each of the six conflict tactic measures (Verbal Abuse, Withdrawal, Symbolic Display, Minor Physical Abuse and Major Physical Abuse), significant differences were found between the high abuse group and the low abuse group, verifying that thet were indeed, different in exposure to abuse in their families. With this established, we proceded to examine the two groups for their pre-conflict total emotion, as well as on pre-conflict anger and subanger scores. The CTS was utilized to measure total verbal and physical abuse experienced in the family of origin. The low abuse group (n=28) reported no physical violence within the family of origin and met our criterion for low family of origin verbal abuse (total verbal abuse less than one half standard deviation below the mean, (M)=8.1, (SD)=6.3). The high abuse group (n=28) reported a total family of origin physical violence score greater than or equal to ten, and a total verbal abuse score greater than one half standard deviation above the mean, (M)=8.1, (SD)=6.3). To determine our criterion for high physical violence in family of origin, a screening questionnaire (CTS) was administered to a normal sample of 287 individuals. The frequency distribution indicated a total physical violence score of ten corresponded to the ninetieth percentile for this particular sample. Thus, a total physical violence CTS score of ten or greater was designated as a criterion for high physical violence in the family of origin. This was done to ascertain whether differences on these scores would appear for extreme group comparisons as well as the across sample correlations already performed. The entire sample might have been too low in scores of extreme family abuse for correlations with current experienced affect to be mainfested. By examining these extreme groups, potential differences might appear. The two groups were significantly different for pre-conflict total emotion, anger and sub-anger. There were no significant differences between the two groups on any of the post-conflict affect scores. The results are presented in Table 5 and Figure 1.

_________________________

Insert Table 5 and Figure 1 about here

_________________________

Figure 1 presents the pre-emotion affect adjective checklist scores of subjects who experienced either high family of origin physical violence (CTS greater than 10) or low family of origin physical violence scores (CTS equal to 0). No differences were found between these two groups on post-emotion AAC scores.Figure 2 shows pre-conflict arousal scores for the high and low abuse groups.

___________________________

Insert Figure 2 about here____________________________

DISCUSSION

The present study found that all forms of conflict tactics were used regularly in the families of origin of a sample of college students. All forms of abuse except major physical abuse were reported as occurring at least fortnightly . Fathers were reported as being significantly more abusive to mothers than were mothers to fathers.

These conflict tactics have long term, generalized affective consequences: college students who anticipated experiencing an audiotape of a family conflict and who had experienced abuse in the family of origin manifested stronger anticipatory affect than those from low or no abuse families. The strongest effects on arousal prior to anticipated conflict was produced by a family history of withdrawal startegies to conflict, but family histories of physical abuse and verbal abuse produced this effect as well. Post -conflict affect was not significantly related to conflict resolution tactics in the family of origin.

The finding that anticipatory arousal generalizes to analog tests situations is consistent with the view that abuse in the family of origin has long term emotional consequences (Jaffe, Wolfe & Wilson, 1990). However, most studies of long term sequelae have been based on clinical samples. Typically, research has not focused on sequelae of abuse in non-clinical samples. Feminist theories of family violence have argued that such violence is not abnormal but is produced by "cultural prescriptions that are cherished in Western society" (Dobash & Dobash, 1979; Burstow, 1992). The present study shows that, at least for the so-called "minor" conflict tactics, this appear to be true. One consequence of this "normal violence" (Straus et al., 1980) is a heightened emotional response to anticipated conflict. The emotions generated are anger, anxiety and "sub-anger" a mix of irritation and frustration. Arguably, this emotional template could create a context for inappropriate conflict resolution tactics used by an adult, since others have shown that heightened anger and anxiety lessen positive problem solving. Margolin, John and Glebermen (1989) found that physically assaultive couples (compared to non-assaultive controls) engaging in short term problem solving exercises manifested heightened arousal and more negative emotional states which began during the anticipation period.

Why abuse experiences had no effects on post-conflict arousal and affect in the current study is unknown. We could speculate that our stimulus scenarios were somehow less aversive than the level experienced by abused subjects in their family of origin. In other words, the anticipatory arousal may have been stimulated by the recollection of family conflicts which were more extreme than what was presented. During the presentation, this arousal level may drop as a form of relief that the presented conflicts are not as bad as expected. Obviously, only physiological monitoring could resolve this issue.

When a high physical abuse group was identified and compared to a low physical abuse group an interesting result emerged, depicted in Figure 2 . van der Kolk (1987) describes two contrasting affective responses to trauma described respectively as autonomic hyperarousal and affective numbing. These are included in the Diagnostic and Statistical Manual of the Mental Disorders-III (American Psychiatric Association, 1980) diagnostic criteria for Post Traumatic Stress Disorder (PTSD) as "numbing of responsiveness" or affective flatness and "hyperalertness/exaggerated startle response" (p. 238). In our current non-clinical sample, subjects reporting high abuse rates in their family of origin demonstrated heightened affect 50% of the time (compared to 22.5% of the low abuse subjects). High abuse subjects showed a low affect pattern 27% of the time compared to 45% of the low abuse. However, when we focus specifically on reports of arousal, the response rates were as follows: high abuse subjects--60% high arousal, 35% low arousal. For low abuse subjects the results were 17.5% high arousal and 20% low arousal. Most of the low abuse subjects manifested moderate levels of anticipatory arousal. Almost all the high abuse subjects manifested a hyperarousal or non-arousal (affective flatness) pattern. These data suggest that, to the extent that self-reported arousal corresponds to actual physiological levels, the clinical arousal pattern of PTSD may be approximated by a non-clinical population. Subjects from high physical abuse families demonstrated higher levels of both affective numbing and extreme arousal in response to conflict. Without the separation of this group from the rest of the subject sample, this effect might have gone undetected.by correlational analysis. Since, high family of origin physical abuse victims split into two groups with essentially opposite arousal reactions, they tend to cancel each other out when physical abuse is simply correlated with arousal across all subjects.

Finally, it may be that the effects of abuse in the family of origin on analog conflict scenarios are mediated by personality or chronic styles which are themselves related to early parenting modes. Dutton, Starzomski, Saunders and Bartholomew (1994) found that attachment style was strongly related to experiences of affect in adult relationships and especially to anger, jealousy and conflict tactics. A recent study (Starzomski & Dutton, 1993) found that attachment style was strongly related to affect reports after exposure to conflict scenarios. Attachment style may be the mediator between early experiences with conflict in the family of origin and adult post-exposure affective reactions to intimate conflict. Future work in this area should include personality assessments as mediators of post exposure affect.

The relatively strong effect of withdrawal is somewhat surprising. Withdrawal had the strongest correlation with pre-conflict (anticipatory) affect and was the only conflict tactic significantly associated with post-conflict affect. This is not merely an effect of frequency of use. Verbal abuse was reported as occurring more frequently than withdrawal. The symbolic meaning of withdrawal in intimate dyadic conflict may warrant further study. Of particular interest is the interpretation of withdrawal to the other's disdain for the respondent as opposed to seeing it as motivated by anxiety on the other's behalf. These attributions themselves may be related to attachment (Starzomski & Dutton, 1993).

REFERENCES

American Psychiatric Association (1980). Diagnostic and Statistical Manual of the Mental Disorders, Third Edition. Washington, DC: APA Press..

Bower, S. A., & Bower, G.H. (1976). Asserting yourself: A practical guide for positive change. Reading, MA: Addison-Wesley.

Browning, J. J., & Dutton, D. G. (1986). Assessment of wife assault with the conflict tactics scale: Using couple data to quantify the differential reporting effect. Journal of Marriage and the Family, 48, 375-379.

Burstow, B. (1992). Radical feminist therapy: Working in the context of violence. Newbury Park: Sage.

Covell, K., and Abramovitch, R. (1987). Understanding emotion in the family: Children's and parents' attributions of happiness, sadness and, anger. Child Development, 58(4), 985-991.

Davies, P.T. & Cummings, E.M. (1994) Marital conflict and child adjustment:An emotional security hypothesis. .Psychological Bulletin, 116, 3, 387 -

Davis, M.H., Hull, J.G., Young, R.D., & Warren, G.G. (1987). Emotional Reactions to Dramatic Film Stimuli: The Influence of Cognitive and Emotional Empathy. Journal of Personality and Social Psychology, 52 (1), 126-133

Dobash, R.E. & Dobash, R.P. (1979). Violence against wives: The case against the patriarchy. New York: Free Press.

Dutton, D.G., & Browning, J.J. (1988). Concern for power, fear of intimacy and wife abuse. In G.T. Hotaling, D. Finkelhor, J.T. Kirkpatrick, & M. Straus (Eds.), New Directions in Family Violence Research. Beverly Hills, CA: Sage.

Dutton, D.G., Saunders, K., Starzomski, A., & Bartholomew, K. (1994). Intimacy-anger and insecure attachment as precursors of abuse in intimate relationships. Journal of Applied Social Psychology, 24(15), 1367 - 1386.

Dutton, D. G., & Strachan, C. E. (1987). Motivational needs for power and dominance as differentiating variables of assaultive and non-assaultive male populations. Violence and Victims, 2(3), 145-156.

Dutton, D.G., Webb, A.N., & Ryan, L. (1994). Gender differences in anger/anxiety reactions to witnessing dyadic family conflict. Canadian Journal of Behavioural Science, 26(3), 353 - 364.

Frodi, A., Macaulay, J. & Thome, P. (1977). Are women always less aggressive than men? Psychological Bulletin, 84, 634-660.

Harter, S. (1982). The perceived competence scale for children. Child Development, 53(1), 87-97.

Jaffe, P., Wolfe, D. & Wilson, S.K. (1990). Children of battered women. Newbury Park, California: Sage.

Kalmuss, D.S. (1984). The inter-generational transmission of marital aggression. Journal of Marriage and the Family, 36(1), 11-19.

Konecni, U.J. (1975). Annoyance, type and duration of postannoyance activity, and aggression: The "cathartic effect." Journal of Experimental Psychology, 104(1), 76-102.

Margolin, G., John, R.S. & Glebermen, L. (1989) Affective responses to conflictual discussions in violent and non-violent couples. Journal of Clinical and Consulting Psychology, 56 (1), 24 -33.

Rule, B.G., & Nesdale, A.R. (1976). Emotional arousal and aggressive behavior. Psychological Bulletin, 83, 851-863.

Russell, J., and Mehrabian, A. (1974). Distinguishing Anger and Anxiety in Terms of Emotional Response Factors. Journal of Consulting and Clinical Psychology, 42, 79-83.

Schill, T., & Schneider, L. (1970). Guilt and self report hostility. Psychological Reports, 27(3), 713-714.

Starzomski, A. J. & Dutton, D.G. (1993). Emotional and attributional reactions to conflict: The role of attachment. Unpublished manuscript.

Strachan, C.E., & Dutton, D.G. (1992). The role of power and gender in anger responses to sexual jealousy. Journal of Applied Social Psychology, 22, 1721-1740.

Straus, M. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics Scale. Journal of Marriage and the Family, 41, 75-88.

Straus, M. A., Gelles, R. J., & Steinmetz, S. (1980). Behind closed doors: Violence in the American family. Doubleday, NY: Anchor Press.

van der Kolk, B. (1987). Psychological Trauma, Washington, DC: American Psychiatric Press.

Widom, C.S. (1989). Does violence beget violence? A critical examination of the literature. Psychological Bulletin, 106(1), 3-28.

Table 1

Reports of Conflict Tactics in Family of Origin in College Undergraduate Sample

Verbal Withdrawal Symbolic Minor Physical Major Physical

Abuse Display Abuse Abuse

Total 8.1 (6.3) 7.4 (4.6) 2.7 (3.2) 2.0 (2.4) .96 (1.8)

DM 1.6 (1.7) 1.4 (1.2) .52 (.84) .34 (.76) .16 (.58)

MD 1.2 (1.5) 1.2 (1.1) .34 (.65) .16 (.51) .06 (.27)

DY 1.6 (1.8) .95 (1.2) .52 (.90) .62 (.96) .35 (.80)

YD .86 (1.1) 1.4 (1.1) .45 (.77) .11 (.31) .01 (.09)

MY 1.6 (1.6) .87 (.98) .34 (.64) .61 (.83) .35 (.73)

YM 1.3 (1.2) 1.6 (1.1) .54 (.83) .19 (.45) .03 (.22)

0 = never, 1 = at least once, 2 = occasionally, 3 = frequently, 4 = weekly

DM= father to mother, MD= mother to father, DY= father to child, YD= child to father, MY= mother to child, YM= child to mother

Total= sum of conflict tactics used across each of the six dyad pairs (0=theoretical minimum, 24=theoretical maximum)

Table 2

T-tests

Family of Origin Conflict Tactic		t	p
Verbal Abuse
Dad 3.2 (3.1)	Mom 2.9 (2.8)	1.42	n.s.
Dad 3.2 (3.1)	You 2.1 (1.9)	4.34	.000*
Mom 2.9 (2.8)	You 2.1 (1.9)	2.96	.004*
Withdrawal
Dad 2.4 (2.1)	Mom 3.0 (1.8)	1.61	n.s.
Dad 2.4 (2.1)	You 3.0 (2.0)	-4.99	.000*
Mom 2.1 (1.8)	You 3.1 (2.0)	-2.92	.004*
Symbolic Display
Dad 1.0 (1.6)	Mom .70 (1.1)	2.43	.01*
Dad 1.0 (1.6)	You 1.0 (1.5)	.20	n.s.
Mom .70 (1.1)	You 1.0 (1.5)	-2.76	.007*
Minor Physical
Dad 1.0 (1.4)	Mom .80 (1.0)	1.56	n.s.
Dad 1.0 (1.4)	You .30 (.60)	5.03	.000*
Mom .80 (1.0)	You .30 (.60)	5.29	.000*
Major Physical
Dad .50 (1.2)	Mom .40 (.90)	1.00	n.s.
Dad .50 (1.2)	You .30 (.20)	4.39	.000*
Mom .40 (.90)	You .04 (.20)	5.07	.000*

* denotes significance at Bonferroni level (p<.0167)

Table 3

MANOVA

Conflict Tactic by Dyad Pair		F	p
Verbal Abuse
DM 1.6 (1.7)	MD 1.2 (1.5)	8.17	<.005*
DY 1.65 (1.8)	YD .86 (1.1)	28.83	<.001*
MY 1.6 (1.6)	YM 1.3 (1.2)	6.95	<.01*
Withdrawal
DM 1.4 (1.2)	MD 1.2 (1.1)	4.58	.05
DY .95 (1.2)	YD 1.4 (1.1)	15.14	.001*
MY .87 (.98)	YM 1.6 (1.1)	44.18	.001*
Symbolic Display
DM .52 (.84)	MD .34 (.65)	5.13	<.05
DY .53 (.91)	YD .46 (.78)	.61	n.s.
MY .34 (.64)	YM .54 (.83)	10.77	<.001*
Minor Physical
DM .34 (.76)	MD .16 (.51)	10.86	<.001*
DY .62 (.96)	YD .11 (.31)	32.95	<.001*
MY .61 (.83)	YM .19 (.45)	38.31	<.001*
Major Physical
DM .16 (.58)	MD .06 (.27)	5.33	<.05
DY .35 (.80)	YD .01 (.09)	22.52	<.001*
MY .35 (.73)	YM .03 (.22)	23.65	<.001*

* denotes significance at Bonferroni level (p<.0167)>>

Table 4

Pearson Correlations between General Conflict Tactics Scale scores and Affect Adjective Checklist Composites (PreConflict) for General Anticipated Family Conflict

PreAnx PreAng PreSub PreArous PreEmot

Total Withdrawal .22** .31*** .42*** .08 .42***

Total

Verbal Abuse .19 .30*** .35*** .12 .36***

Total

Physical Abuse .14 .17 .31*** .09 .30***

** p<.01 *** p<.001 N = 120

Bonferroni adjustment p<.003

PreAnx = anxiety reported before exposure to dyadic conflict

PreAng = anger reported before exposure to dyadic conflict

PreSub = subanger reported before exposure to dyadic conflict

PreArous= arousal reported before exposure to dyadic conflict

PreEmot = total affect reported before exposure to dyadic conflict

Total Physical Abuse includes symbolic display, minor and major physical abuse.

Table 5

Pearson Correlations between Conflict Tactics Scale scores and Affect Adjective Checklist Composites (PostConflict) for General Witnessed Family Conflict

PostAnx PostAng PostSub PostArous PostEmot

Total Withdrawal .14 .09 .09 .13 .13

Total Verbal Abuse .04 .05 .04 .09 .05

Total Physical Abuse .02 .09 .03 .08 .09

** p<.01 *** p<.001 N = 120

Bonferroni adjustment p<.003

PostAnx = anxiety reported after exposure to dyadic conflict

PostAng = anger reported after exposure to dyadic conflict

PostSub = subanger reported after exposure to dyadic conflict

PostArous= arousal reported after exposure to dyadic conflict

PostEmot = total affect reported after exposure to dyadic conflict

Total Physical Abuse includes symbolic display, minor and major physical abuse.

Table 6

Pearson Correlations between Specific Conflict Tactics Scale Scores and General Anticipated Family Conflict Affect (PreEmot)

Verbal Abuse DM MD DY YD MY YM

.37*** .19* .34*** .21** .26** .32*** .21*

Withdrawal

.42*** .29*** .28*** .33*** .27** .30*** .28***

Symbolic Display

.32*** .06 .32*** .11 .28*** .24** .34***

Minor Physical

.27*** .04 .09 .15* .12 .37*** .20**

Major Physical

.14* .09 .10 .04 .19* .18* -.03

* p<.05 ** p<.01 *** p<.001

M = mom, D = dad , Y = you

Table 7

Pearson Correlations between Specific Conflict Tactics Scale scores and Affect Adjective Checklist Composites (PostConflict) for General Witnessed Family Conflict

Verbal Abuse	DM	MD	DY	YD	MY	YM
.05	.03	.03	.06	.15	-.04	.08
Withdrawal
.13	.06	.15	-.01	.13	.05	.15
Symbolic Display
.14	.00	.02	.10	.09	.20*	.18
Minor Physical
.09	.04	.03	.06	.03	.11	.05
Major Physical
.02	.07	.02	.03	.05	-.04	.01

* p<.01

M = mom, D = dad , Y = you

Table 8

Comparisons of Conflict Tactics and AAC measures

between groups reporting high physical and high verbal abuse and groups reporting no physical and low verbal abuse in family of origin

n=25 n=28

Low Abuse High Abuse t p

Conflict Tactics

Verbal Abuse 1.0 (1.4) 16.2 (3.5) 21.2 p<.001

Withdrawal 3.3 (2.8) 11.1 (2.8) 10.3 p<.001

Symbolic Display 0.0 (0.0) 6.6 (2.4) 14.8 p<.001

Minor Physical 0.0 (0.0) 4.9 (2.2) 11.7 p<.001

Major Physical 0.0 (0.0) 2.6 (2.7) 5.1 p<.001

PreAAC

EmotPre 2.5 (0.8) 3.6 (1.5) 3.6 p<.001

PreAnx 4.0 (1.5) 4.8 (2.0) 1.7 p>.05

PreAng 2.1 (1.2) 3.0 (1.8) 2.3 p<.05

PreSub 1.9 (0.9) 3.7 (2.0) 4.3 p<.001

PreArous 3.5 (1.9) 4.2 (1.8) 1.2 p>.05

Figure 1

Scores of PreEmotion Affect Adjective Checklist Measure by Violence in Family of Origin

Confam3.doc revision July 13, 1993 LJ

Confam4.doc revision August 5,1993 LJ

Confam5.doc revision Dec. 1 1993 DGD

Confam5.doc final January 4 1994 BM

Confam51.doc revision September 6 1994 LJ

Confam6.doc revsion November 25, 1994 AK & DG

Confam7 revision November 28, 1994 AK

Novmeber 30, 1994 DG & AK

November 30 R & R CJBS