UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of child type and behavioural impact on mothers’ attributions for child behaviour Geller, Josephine Amanda Caroline 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1992_fall_geller_josephine.pdf [ 4.22MB ]
JSON: 831-1.0086477.json
JSON-LD: 831-1.0086477-ld.json
RDF/XML (Pretty): 831-1.0086477-rdf.xml
RDF/JSON: 831-1.0086477-rdf.json
Turtle: 831-1.0086477-turtle.txt
N-Triples: 831-1.0086477-rdf-ntriples.txt
Original Record: 831-1.0086477-source.json
Full Text

Full Text

THE EFFECT OF CHILD TYPE AND BEHAVIOURAL IMPACT ON MOTHERS'ATTRIBUTIONS FOR CHILD BEHAVIOURbyJOSEPHINE GELLERB.A. (hon), University of Ottawa, 1990THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF ARTSinTHE FACULTY OF GRADUATE STUDIES(Department of Psychology)We accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1992© Josephine Geller, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of ^PS9'Ght21-0 I/The University of British ColumbiaVancouver, CanadaDate Sept a 9) /99 2.DE-6 (2/88)11AbstractThis study examined the impact of two relationship-focusedvariables on the formation of parent causal attributions forchild behaviour. Previous work has focused on the influence ofgeneral parent variables, child variables, and situationalfactors on these attributions. The present study highlighted theimportance of the relational context of parent-child interactionsas an influence on parent attributions, and addressed twoimportant aspects of this relational context. First, parentattributions formed in response to behaviours of their ownchildren were compared with those formed in response tobehaviours of unknown children of the same age and gender astheir own children. Second, the behavioural impact(inconvenience or no inconvenience to the mother) of children'snoncompliant behaviour on the mother was examined. Inconveniencewas defined as the extent to which the mother was personallyhassled or bothered by the child behaviour. Results indicatedthat with other children, the behavioural impact of childbehaviours was positively related to stronger affective andbehavioural responses. With their own children, mothers ratedthe cause of their child's noncompliance as less due to globaland stable factors, and anticipated stronger affective andbehavioural responses than with other children. Mothers also sawthe cause of their own child's noncompliance as more due tothemselves and within their own control than with other children.Regression analyses indicated that mothers' affective responseswere predicted by the behavioural impact of the child's behaviouriiiand by mothers' attributions regarding their own as well as theirchildren's role in causing the behaviour. In contrast, mothers'behavioural responses were predicted only by mothers' ratings ofpersonal controllability. Higher perspective taking scores wererelated to seeing the child's behaviour as less due to internal,controllable causes, and to lower anticipated affectiveresponses. Higher empathic concern scores were associated withlower anticipated response ratings, and to higher ratings ofpersonal controllability over the situation. Finally, greaterinvestment in parenting was associated with stronger affectiveand behavioural responses, and to a decreased likelihood ofseeing the child's behaviour as due to internal and controllablecauses. These findings were interpreted within the framework ofDix's (1991) model of affective processes on parenting, and as anapplication of the social cognition phenomenon of positiveattributional biases as extended by parents to their children.ivTable of ContentsAbstract^ iiTable of Contents^ ivList of Tables viAcknowledgement^ viiIntroduction 1Attribution Theory^ 1Parent Attributions and Behaviour^ 3Review of Studies Involving Parent Attributions^ 7Parent Characteristics^ 8Child Characteristics 11Situational Variables^ 14A Move Towards Relationship-Focused Variables^ 18Child Type and Extension of Positive Attributional Bias ^ 21Mother Inconvenience^ 24Secondary Research Questions^ 28Parent Empathy^ 28Parenting Investment^ 30Purpose and Hypotheses 31Study Design^ 33Method^ 36Procedure^ 36Subjects 37Measures^ 37Demographic Information^ 38Instructions for Ratings 39Stimulus Materials^ 39Parent Empathy 45VParenting Investment^ 46Results^ 47Study Design^ 47Alpha-Protecting Strategies^ 47Preliminary Analyses^ 48Aggregation of Attribution Ratings^ 48Relationships Among Coded Responses and Ratings^ 49Manipulation Check^ 51Main Analyses^ 52Child Type and Behavioural Impact^ 52Relationships Among Attributions and Affect^ 57Relationships Among Attributions and Behaviour 59Relationships Among Affect and Behaviour^ 60Empathy and Investment in Parenting 61Discussion^ 63Child Type and Behavioural Impact^ 64Relationships Among Attributions Behaviour and Affect^ 73Empathy and Investment in Parenting^ 76Relationships Among Coded Responses and Ratings^ 79A Relationship-Focused Approach^ 80Limitations and Future Research 82References^ 85Appendix A 100Appendix B^ 101Appendix C 102Appendix D^ 103Appendix E 104Appendix F^ 105viList of TablesTable 1:Factor Loadings of Globality/Stability andInternality/Controllability Factors^ 91Table 2:Pearson Correlations Between Coded Attribution Responses andAttribution Ratings for Child^ 92Table 3:Pearson Correlations Between Coded Behavioural Strategies andAnticipated Behavioural Rating^ 93Table 4:Mean Attribution Ratings and Coded Attribution Responses as aFunction of Child Type and Behavioural Impact^ 94Table 5:Mean Mother attribution Ratings as a Function of Child Type andBehavioural Impact^ 95Table 6:Mean Anticipated Response Ratings and Coded Anticipated Behaviouras a Function of Child Type and Behavioural Impact^ 96Table 7:Regression Analyses For Anticipated Affect and BehaviouralResponses^ 97Table 8:Correlations Among Anticipated Affect Ratings, Coded Attributionsfor Child Behaviour, and Coded Anticipated BehaviouralResponses^ 98Table 9:Correlations Between Mothers' Attributions and AnticipatedAffective and Behavioural Responses with Empathy and ParentingInvestment Subscales^ 99viiAcknowledgementI would like to thank Charlotte Johnston for her invaluableguidance and support in writing this thesis. Her expertise andaccessibility were greatly appreciated, and helped shape thefinished product into something that I feel proud of. Thanks arealso due to Becky Collins and Lynn Alden for their helpfulcomments and encouragement. Finally, thanks to friends andfamily who listened to me when I was at my weariest, and whohelped me remember that good things are worth working for.1This thesis begins by providing general backgroundinformation on attribution theory and related models. Studiesinvolving parents' attributions for child behaviour will then bedescribed, emphasizing the demonstrated relationships betweenparent attributions and parent characteristics, childcharacteristics, and situational variables. The relationshipbetween parent attributions and parent behaviour will also bereviewed. A heuristic model outlining putative relationshipsamong parent characteristics, child characteristics, and parentattributions in the context of the parent-child relationship willthen be presented. This relationship is considered unique andmay be characterized by particular patterns of parentattributions, affect, and behaviour that distinguish the parent-child relationship from relationships the parent has with otherchildren. In elaborating on the uniqueness of the parent-childrelationship, positive attributional biases in attributionformation will be described, and parents' extension of thesebiases to their children will be considered. Factors that mayreduce a parent's tendency to extend positive attributionalbiases, such as parent inconvenience, will be considered.Finally, the relationships among parent empathy, parentinvestment and parent attributions will be addressed.Attribution Theory Attribution theory is an information-processing approachthat considers social behaviour to be dependent on theindividual's ongoing assessment of persons and behaviour (Dix &Grusec, 1985). The underlying assumption of this theory is that2individuals are naive scientists striving to understand, predict,and control the course of events in their lives (Heider, 1958).Generally, this approach emphasizes that behaviour depends onpeople's inferences about what is causing the events around them,the motives and traits that characterize individuals in aninteraction, and the properties inherent in a social situation(Dix & Grusec, 1985).Attribution theorists have focused on various aspects of theformulation of attributions. Most models are based on Heider's(1958) original formulation of attribution theory which stressedthat individuals form beliefs that affect their actions, and thatthese beliefs, valid or not, need to be taken into account inunderstanding behaviour. Some theorists have addressed theprocesses guiding the formulation of attributions, includingJones and Davis' (1965) model of correspondent inference andKelley's (1967) covariation model. Other models have focused onattributional outcomes. For example, Weiner's (1979) modeldescribes the dimensions along which attributions are formulated,including inferences about the stability, locus, andcontrollability of the behaviour. This model also describes theimpact of particular attributions, once formulated, on behaviourand affect. For instance, in a study examining judgmentsconcerning the lending of class notes (Weiner, 1980), offers tolend notes were lowest when the cause of the need was seen asinternal to the actor and as controllable (e.g., lack of effort).It was suggested that ascriptions to internal controllable3factors may maximize negative affect (disgust and anger) andpromote avoidance behaviour.Attributional theory has been extended beyond the bounds ofsocial psychology and a recent emphasis on parent cognitionswithin the developmental literature (Sigel, 1985) has encouragedconsideration of the attributions parents form about theirchildren's behaviour. Parent attributions have been highlightedas important factors influencing both parent and, ultimately,child behaviours (Bugental & Shennum, 1984; Dix & Reinhold, 1987;Mackinnon, 1989). Attributional theory suggests that parentingbehaviour may depend on parents' inferences about causes ofchildren's behaviour. In accordance with these predictions, Dixand Grusec (1985) have summarized research showing that parents'attributions for their children's behaviour are related to childage, the type of child behaviour, and the parent's perception ofthe child's motivation and level of understanding. For instance,parents increasingly attribute child behaviours to internal,dispositional causes as children get older and are perceived tounderstand task demands. Attributional theory therefore mayprovide a useful framework for understanding or predictingparents' responses to child behaviours.Parent Attributions and BehaviourA number of researchers have linked parent attributions forchild behaviour to subsequent parenting behaviour (Alexander,Waldron, Barton, & Mas, 1989; Bugental & Shennum, 1984; Dix &Grusec, 1985; Dix, Ruble, & Zambarano, 1989, Larrance &Twentyman, 1983; Mackinnon, 1989; Murphey & Alexander, 1991).4Using vignettes of child behaviour situations, Dix and hiscolleagues (Dix & Grusec, 1985; Dix, Ruble & Zambarano, 1989)found that mothers predicted that they would use more power-assertive discipline strategies when they inferred that theirchildren understood the rules, had the capacity to actappropriately, and were responsible for their negativebehaviours. Similarly, Mackinnon (1989) found that when mothersperceived their sons' behaviours to be negatively intended, theywere more coercive in interactions with the child. A morepervasive attribution-behaviour link is suggested by the findingthat parents known to have a history of abuse with their childrenform more negative attributions about their children's behaviourthan non-abusive parents (Bauer & Twentyman, 1985; Larrance &Twentyman, 1983). More subtle behavioural differences, includingfacial expression and tone of voice, have been linked to mothers'attributions of their own power and control over the outcome ofmother-child interactions (Bugental & Shennum, 1984). Finally,Murphey and Alexander (1991) found that "internal" parentingattributional styles, defined as parents' beliefs in theirefficacy at affecting transient interactions with their children,were linked to parenting behaviour. For mothers, a more internalparenting style was related to increased provision of help in acooperative parent-child building task. For fathers, a moreinternal parenting style was related to increased directivenessin both cooperative task and play situations.Previous research in our lab (Geller, Johnston, & Gabille,1991) has also addressed the attribution-behaviour relationship5in a study where mothers read descriptions of ambiguous negativeevents involving either themselves or their child. Mothers wereasked to imagine that the situations happened to them or theirchild, answer a series of questions about what they thoughtcaused each of the situations, and describe what they would do ineach situation. Mothers' causal attributions in both types ofsituations (those involving themselves and those involving theirchild) were found to be related to various aspects of theirpredicted behavioural responses. For instance, less favorableattributions about the cause of their child's negative situation(i.e., viewing the cause of the situation as something that wouldbe present in the future, and as something likely to influenceother similar situations) were associated with more negativeanticipated parenting behaviours, such as expressing negativeemotion, actively blaming the child, or telling the child thathe/she should have behaved differently.Studies in the parenting literature addressing attribution-behaviour relationships can be distinguished into two groups.Some of this research has assessed "direct" attribution-behaviourlinks, in which parent attributions and behaviours were assessedin response to the same stimuli (Dix & Grusec, 1985; Dix, Ruble,& Zambarano, 1989). In these studies, parents are presented withdescriptions of situations, asked to form attributions about thechild's behaviour in the situations, and then asked what theywould do if the situation happened in their own lives. Otherstudies have assessed attributions using questionnaires orstimulus materials, and then measured behaviour in a different6situation (Alexander, Waldron, Barton, & Mas, 1989; Bugental &Shennum, 1984; Larrance & Twentyman, 1983; Mackinnon, 1989;Murphey & Alexander, 1991). The convergence of findings (i.e.,establishing an attribution-behaviour link) across the differenttypes of methodological circumstances confirms that parentattributions are related to short-term parenting outcomes and arealso consistent over time and situations.It should be noted that the relationships among parentattributions, parent behaviours, and child behaviours areconsidered to be bi-directional. Although parent attributionsare thought to influence parent behaviours (Dix & Reinhold, 1987;Dix, Ruble, & Zambarano, 1989) which in turn impact on childbehaviours (Patterson, 1982), these influences are alsoacknowledged to occur in the opposite direction. Child behaviourobviously impacts on parent behaviour (Anderson, Lytton, &Romney, 1986; Bugental, Blue, & Lewis, 1990; Lytton, 1990;Maccoby & Martin, 1983; Patterson, 1982) and both parent andchild behaviour contribute to the formation or change of parentattributions (Brunk & Henggeler, 1984). An illustration of thereciprocal influence of child behaviour on parent attributions issuggested in a study comparing the attributions of mothers ofAttention Deficit Hyperactive Disordered (ADHD) children withthose of mothers of nonproblem children (Sobol, Ashbourne, Earn,& Cunningham, 1989). In responding to written descriptions ofchild compliance and non-compliance in which they imagined thatthe child in the descriptions was their own, mothers of ADHDchildren viewed the causes of their child's behaviour as more7unstable than did mothers of control children. It is likely thatmothers' experiences with these children (whose behaviour ischaracterized by known patterns of attentional and impulsecontrol difficulties) has led them to form different attributionsabout the causes of child behaviours than mothers who haveexperience with non-ADHD children. Child behaviours thereforeappear to impact on mothers' attributions.Review of Studies Involving Parent Attributions: In reviewing the parent attribution literature, three linesof research may be identified. First, researchers have focusedon the relationship between parent characteristics andattributions formed about the causes of child behaviours. Forexample, some researchers have compared attributions of abusiveparents with those of non-abusive parents (Bauer & Twentyman,1985; Larrance & Twentyman, 1983). Second, the relationshipbetween parents' attributions and experiences with differenttypes of child behaviour have been elucidated. Studies in thisgroup have examined attributions among parents who have childrenwith learning difficulties (Himmelstein, Graham, & Weiner, 1991),aggressive behaviour, (Dix & Lochman, 1990; Reinhold & Lochman,1991), and attentional difficulties (Sobol et al., 1989).Finally, studies have examined the relationship between parentattributions and situational variables, such as the specific typeof child behaviour (Brunk & Henggeler, 1984; Dix & Grusec, 1985),and parent affect at the time of the encounter (Dix, 1989).Thus, the following review of studies on parent attributions willbe categorized into three groups: research on "parent8characteristics," "child characteristics," and "situationcharacteristics."(i) Parent Characteristics The most compelling evidence for the existence of differentattributional styles associated with parent characteristics stemsfrom the abuse literature. As noted above, compared withnonabusive mothers, mothers with a history of abusive parentinguse a more negative, hostile attributional style in explainingtheir children's behaviour (Bauer & Twentyman, 1985; Larrance &Twentyman, 1983). For example, Larrance and Twentyman (1983)showed 10 abusive mothers, 10 neglectful mothers, and 10 motherswith no previous history of child maltreatment standard sequencesof photographic stimuli of their own and another child in commonsituations. In each series of photographs, it could be inferredthat some negative event had occurred, although the cause of theevent was ambiguous. For instance, photographs depicted crayonedwalls, broken toys, and the results of competitive games. Themothers were asked to state their attributions regarding thecause of the child behaviours described in the photographs.Compared to the nonabusive mothers, abusive mothers were morelikely to attribute their children's good behaviour to unstable,external factors, and to see their children's bad behaviour ascaused by stable and internal sources. In a similar studycomparing abusive, neglectful, and normal mothers, Bauer andTwentyman (1985) found that abusive mothers interpreted childbehaviour as more malevolent than control mothers, withneglectful mothers' attributions falling in between the other two9groups. These studies suggest that negative attributions forchild behaviour may contribute to the precipitation of familyviolence. Further support for the link between parentattributions and violent behaviour comes from a study showingthat in a sample of young mothers considered to be at risk forchild maltreatment, mothers' levels of unrealistic expectationsregarding children were positively correlated with theirattributions of negative child intent, and both of these factorswere related to mothers' level of punitiveness in response tohypothetical childrearing situations (Azar, 1989). In contrastto these findings, Rosenberg and Reppucci (1983) failed to finddifferences between abusive and nonabusive mothers' attributionsfor the cause of child transgressions. The absence ofdifferences in this study compared to the other studies justdescribed may be due to the lack of sensitivity of theattribution measure used by Rosenberg and Reppucci. Whereas theother studies in this section employed Likert scales ranging from7 to 9 points, their study asked parents to answer true-falsequestions. Clarification of this difference in results awaitsfurther study.Maternal depressed mood is a parent variable noted for itsassociation with child behaviour problems (e.g., Cunningham,Benness & Siegal, 1988; Griest, Wells, & Forehand, 1979). Therelationship between depressed mood and a negative attributionalstyle has been repeatedly demonstrated in the adult literature(e.g., Seligman, Abramson, Semmel, & Von Baeyer, 1979) andsupports predictions made by the reformulated helplessness model10of depression (Abramson, Seligman, & Teasdale, 1978). This modelproposes that individuals who make attributions to stable,internal, and global causes of failure are more likely to respondin a helpless fashion to uncontrollable events. Although therelationship between depressed mood and attributional style hasbeen repeatedly demonstrated, the impact of parental depressedmood on attributions formed about their children has seldom beenexplored. One exception is the study we conducted (Geller etal., 1991) examining mothers' attributions for their own andtheir children's behaviour as depicted in written situations. Inthis study, as expected, mothers' depressed mood predictednegative attributions along the dimensions of internality,controllability, globality, and stability in explaining negativeevents occurring in their own lives. More interestingly,depressed mood also predicted use of more internal andcontrollable attributions for negative events involving mothers'children. Depressed mood in mothers therefore was not onlyrelated to explanations about events in their own lives, but alsoinfluenced social cognitions concerning their children.Gender is another parent characteristic that has beenconsidered in relation to parent attributions. Although mostresearch has focused on maternal attributions, a small number ofstudies have also included fathers. In a sample of high-conflictfamilies of adolescents, ratings of dispositional attributionsregarding family members (including children) did not differbetween mothers and fathers (Alexander et al., 1989). Parents'perceived control over caregiving, and parental locus of control11have also been reported as similar for mothers and fathers(Murphey & Alexander, 1991). Furthermore, a recent study in ourlab examining causal attributions formed in response to writtendescriptions of hyperactive and aggressive child behaviours inparents of children with ADHD also found no differences betweenmothers' and fathers' attributions (Johnston & Patenaude, 1991).In contrast, Sobol and his colleagues (1989) found that within asample of parents of ADHD children, mothers rated the cause oftheir child's noncompliance as less due to the child and more dueto the situation than did fathers. Thus, the findings regardingmother-father differences in attributions for child behaviour aresomewhat equivocal although the majority of studies indicate nosignificant gender differences.In sum, previous research has examined the relationshipbetween parent characteristics and parents' causal attributionsfor their children's behaviour. Most notably, parents with ahistory of abuse or neglect use more negative attributionalstyles than parents with no history of child maltreatment.Additionally, preliminary work suggests that parents experiencingdepressed mood are more likely to attribute the causes of theirchildren's negative behaviours to more internal, controllablefactors. Finally, although infrequently examined, resultssuggest few differences between mothers and fathers'attributional responses.(ii) Child Characteristics In addition to examining the relationship between parentcharacteristics and parent attributions for children's behaviour,12parents have also been distinguished according to characteristicsassociated with their children. Himmelstein, Graham, and Weiner(1991) compared attributions regarding the importance ofchildrearing practices on offspring outcomes among mothers ofgifted and special education children. Child outcome wasassessed across three domains: academic performance, socialskills, and personality. Mothers of gifted children were foundto attribute outcome to child-rearing to a greater extent thanmothers of special education children across all three outcomedomains. Additionally, mothers of only children were found toattribute greater importance to child-rearing practices than didmothers of multiple children. In sum, both perceived childsuccess and child singularity in the family were associated withattributions of childrearing as an important cause of childbehaviour.Other work has focused on the attributions mothers ofaggressive children form about their children's behaviour. Aconsistent finding is that mothers of aggressive children makemore negative, hostile attributions for child misbehaviour thando mothers of nonaggressive children (Dix & Lochman, 1990).Mothers of aggressive boys attribute the cause of boys' non-compliant actions to more intentional causes, and see thebehaviours as more reflective of negative personalitydispositions in the child than mothers of nonaggressive boys.This effect has been demonstrated using videotaped depictions ofunfamiliar mothers and children (Dix & Lochman, 1990), andaudiotaped and written vignettes in which mothers were instructed13to imagine that the child in the vignettes was their own (Petit,Dodge, & Brown, 1988; Reinhold & Lochman, 1991). Mothers ofaggressive boys have also been found to become more upset withchild misbehaviour, and to endorse more forceful disciplineresponses than mothers of nonaggressive children (Dix & Lochman,1990; Reinhold & Lochman, 1991). These findings are consistentwith the attribution-behaviour link alluded to earlier. Giventhat, similar to their mothers, aggressive boys also show anegative attributional bias in explaining the cause of others'behaviours, it has been proposed that this style of biased socialinformation processing may be learned in the family (Reinhold &Lochman, 1991).Finally, as stated earlier, mothers of ADHD children formattributions that can be distinguished from those of mothers ofnormal children. In responding to written descriptions of childcompliance and non-compliance in which mothers imagined that thechild in the descriptions was their own, mothers of ADHD childrenviewed the cause of their children's behaviour to be moreunstable and less controllable than did mothers of controlchildren (Sobol et al, 1989). In another study conducted in ourlab, parents of ADHD children who rated their children as moreaggressive were more likely to attribute written descriptions ofaggressive behaviours to controllable causes than were parents ofADHD children who did not rate their own children as aggressive(Johnston & Patenaude, 1992) As noted earlier, these studieshighlight the potential role of child behaviour patterns indetermining parent attributions.14The link between experiences with various types of childbehaviour and parent attributions has also been demonstrated innonclinical samples. Using a community sample of mothers, wefound that mothers' ratings of conduct problems in their ownchildren were significantly correlated with attributions formedin response to written child situations (Geller et al., 1991).Mothers who described their child as having more conduct problemsattributed negative child events to more global and stable causesthan did mothers who rated their child as having fewer conductproblems.In summary, the attributions parents form about children's'behaviours are related to their experiences with their ownchildren. This relationship has been demonstrated across parentsof children with different levels of academic achievement, anddifferent degrees of child aggressiveness, conduct problems, andimpulsivity and hyperactivity.(iii) Situational variables The associations between parent attributions and parentcharacteristics or experiences with their own children havegenerally been explained in terms of broad response sets thatparents hold regarding children's behaviour. That is, particularattributional styles have been associated, for instance, withparent abusiveness or with being the parent of an aggressivechild, and these styles are thought to colour parents'impressions of child behaviour across different situations.Despite the demonstrated importance of parent and childcharacteristics in explaining parent attributions, considerable15variance in attributional ratings has remained unexplained. Thatis, variations in these ratings occur, not only between differentgroups or types of parents, but also within each of the groupsacross situations. It appears that situational factors alsoinfluence attribution ratings. This section will address thesituational context variables that have been found to account foradditional variance in explaining parent attribution scores.Dix and colleagues have shown that variations in childbehaviour are related to variations in parents' social inferences(Dix & Grusec, 1985). Mothers responded differently to scenariosof children engaging in altruistic behaviour (helping, sharing,showing concern), failures to be altruistic (not helping, sharingor showing concern), or explicit norm violations (fighting,stealing, lying). Altruistic behaviours were seen as morestable, general, intentional, controllable, and more the resultof child dispositions than were failures to be altruistic orexplicit defiance. Additionally, failures to be altruistic (forinstance, not sharing) were rated as less intentional andblameworthy than overt defiant behaviours. Dix has suggestedthat the difference between these last two conditions may be areflection of behaviour complexity. Parents may be more upsetwith complex negative behaviours that they infer children don'tunderstand or cannot control than with simpler negativebehaviours that they infer children do understand and control.An even more subtle variation in child behaviour, that oftiming of the behaviour, has also been demonstrated to relate toparent attributions. Dix and Reinhold (1987) made videotapes16that depicted children disobeying simple requests (e.g., "put onyour shoes before you go out") either immediately or following 15seconds of continuing activity. Parents viewed immediatedisobedience as more intended and more reflective of negativedispositions in the child than delayed disobedience. Dix andReinhold (1987) postulated that this phenomenon reflects parents'inferences that a delay makes the task more difficult andtherefore taxes the child's ability to execute requests.Finally, parents' transient affective state is anothercontextual variable that may influence attributions. To test theeffect of parent affect on attributions for child behaviour, Dix(1989) had mothers wait until they found themselves in eitherhappy, angry, or neutral moods and then watch videotapes ofmothers and children in discipline situations. Mothers whoreported feeling angry prior to watching the tapes attributed theobserved child behaviours to more dispositional characteristics,anticipated that compliance would be harder to obtain, and feltthat greater sternness should be used than did mothers who feltin neutral moods prior to watching. Mothers who were angry alsoexpected more negative behaviour from their children in a seriesof situations than did mothers who were unemotional or who werein happy moods. Findings in this study therefore demonstratedmood-consistent biasing of attributions.In another study (Dix & Reinhold, 1987), mothers wereinduced to feel either happy, angry, or unemotional and thenshown videotaped interactions of children disobeying parentalrequests. Unlike the previously cited findings, the angry mood17group did not form significantly more negative attributions thanthe neutral mood group, although there was a trend in thatdirection. The explanation proposed by the experimenters wasthat the mood induction was not fully successful. However,interestingly, the group induced to have a happy mood reportedmore negative cognitions than did unemotional parents. Thesefindings are not in accordance with predictions made from themood-congruent processing perspective (see Blaney, 1986) and aredifficult to integrate with current theories of parent cognitionand affect.Although not specific to the area of attributions, theimpact of situational context variables on general parentalattitudes was investigated in a study examining parents' use ofphysical punishment. Holden (1989) found that mothers fell intoone of three groups with respect to general attitudes towardsphysical punishment; a positive group, comprised of mothers whogenerally endorsed use of physical punishment; a negative group,including mothers who rarely endorsed use of physical punishment;and an ambivalent group, consisting of mothers whose attitudesfell midway between the other two groups. Holden found that asituational variable, whether the discipline setting was privateor public, played a significant moderating role between mothers'attitudes toward punishment and their actual use of physicalpunishment. Although negative attitude mothers were unlikely touse physical punishment in any contexts, positive attitudemothers modulated their punishment intentions slightly in publicsettings, and ambivalent mothers' punishment intentions were18significantly greater in private than in public settings. Thisstudy therefore highlights the importance of considering bothsituational context variables (the type of setting) as well aspre-existing cognitions (attitudes regarding physical punishment)in obtaining a complete picture of parenting behaviour, in thiscase, mothers' actual use of punishment.In sum, a number of factors associated with the situationalcontext of parent-child interactions have been identified asrelated to parents' social cognitions. These range from type ofchild behaviour, to the parent's affective state at the time ofthe interaction. In addition, Holden's punishment studyhighlights the importance of considering both pre-existing andsituational variables in predicting ultimate parenting outcomes.A Move Towards Relationship-Focused Variables In summary then, parent, child, and situational contextvariables have been demonstrated as being related to parentattributions. In addition, a small number of studies haveconsidered the relationships among parent and childcharacteristics and parent attributions in the context of theparent-child relationship (Bugental, 1989; Bugental & Shennum,1984; Kochanska, 1990). This approach is analogous to the trendin marital research, specifically in cognitive approaches tointimate relationships, of examining the way distressed partnersform attributions within the context of the relationship. Aconsistent finding in this literature is that maritalsatisfaction is related, not to particular spouse behaviours, butto the attributions the partners form about one anothers'19behaviour (Berley & Jacobsen, 1984; Bradbury & Fincham, 1989;Fincham & Bradbury, 1987). For instance, locating the cause ofthe relationship problems in one's partner, and viewing the causeas stable and global is associated with marital dissatisfaction.This finding highlights the importance of considering not onlycharacteristics of the individuals involved, but alsocharacteristics of the individuals in the context of a particularrelationship. It also shifts the focus of the research fromindividual pathology to a more systemic view of dysfunctionalrelationships.Illustrating this more systemic approach, Bugental andShennum (1984) developed and tested a transactional model ofadult-child interactions. Parent characteristics, such as self-perceived power as caregivers, and child characteristics, such asresponsiveness and assertiveness, were seen as interacting toinfluence parent attributions and behaviour in parent-childinteractions. For instance, in a series of studies examiningthese characteristics, Bugental and Shennum found that parentswho attributed low power to themselves responded more adverselyto unresponsive children than did parents with high self-perceived power in relationships. These adverse responses weredefined in terms of facial expression, tone of voice, andconflictual messages to the children (e.g., smiling whileattempting to exert control in the interaction). Consistent witha bi-directional, transactional perspective on parent-childrelationships, the adverse responses of mothers with low self-perceived power then functioned to increase child20unresponsiveness. In other words, these mothers used a responsestyle that exacerbated difficult child behaviours, which in turnmaintained mother behaviours and attributional beliefs.Conversely, attributions of high self-power in mothers appearedto act as a buffer against the potentially negative impact ofdifficult child behaviour. These mothers demonstrated nosignificant alterations in their behaviour as a function of childbehaviour. Additionally, when the high self-perceived powermothers interacted with unresponsive children, childunresponsiveness was reduced.Another study illustrating the importance of consideringmother and child variables pertaining specifically to therelationship they share was conducted by Mackinnon (1989). Thisstudy involved observations of mothers and their sons working onan etch-a-sketch task. Prior to engaging in the task, mothersand sons formed attributions about each other's behaviour byresponding to a series of written scenarios involving mother-child interactions in potentially conflictual situations. Whenboth mother and child perceived the other's behaviour in thescenarios to be negatively intended (reflective of a negativeattributional style used in the context of that relationship),their behaviour in the etch-a-sketch task was more coercive thanwhen one or no members of the dyad perceived the others'behaviour as negatively intended. This research thereforesuggests that in predicting the outcome of mother-childinteractions, it is necessary to take into account both parentand child characteristics.21Child Type and the Extension of Positive Attributional Bias This section will describe features of the parent-childrelationship that make it unique from other relationships, andwill report on what is known about the impact of these featureson parent attributions.Attributional biases are well-known phenomena that aredescribed in the social cognition literature (e.g., Fiske &Taylor, 1991). They refer to deviations from normativeprocessing in which the social perceiver systematically distorts(overuses or underuses) some otherwise accurate or appropriateprocedure. For instance, the actor-observer bias refers toindividuals' tendency to ascribe others' behaviours todispositional causes and one's own behaviour to situationaldeterminants (Fiske & Taylor, 1991). In this thesis, "positiveattributional biases" refer to attributional bias hypothesized tohave ego-enhancing effects on the social perceiver. An exampleof a positive attributional bias is the self-serving bias. Thisis the tendency of the actor to attribute desirable actions tointernal psychological causes (dispositions in him/herself) andundesirable actions to the external situation (Miller & Ross,1975). There is currently speculation that parents use positiveattributional biases in explaining events in their children'slives (Dix & Grusec, 1985; Gretarsson & Gelfand, 1988; Larrance &Twentyman, 1983). That is, that parents excuse their childrenfor involvement in negative events and emphasize childdispositional characteristics as causal in positive events.22In describing the context in which parent attributions takeplace, Dix and Grusec (1985) highlight features of the parent-child relationship that explain why a positive attributional biasmight be extended to one's children. They note that parents areenmeshed with children in a closely-bound social and biologicalrelationship, and are at once socializers, regulators, andcaretakers of their children. An additional, contextual factorthat may enhance parents' extension of the positive attributionalbias to their children is parent-child intimacy. Because theparent-child relationship is characterized by powerful emotionalbonds, children's behaviour may be considered to be particularlysignificant to parents. That is, children's behaviour haspersonal relevance for parents that may cause them to shareemotionally in their children's' experiences. This bond mayenhance the parent's ability to place him or herself in theposition of the child, and hence to make the same attributionsfor child behaviour as for their own behaviour. Alternately, tothe extent that parents view themselves in a positive light, ifthey consider their children's behaviour as reflective ofthemselves and their parenting abilities, they can be expected toform attributions that will cast their children, and hencethemselves, in a favorable light.In a study designed to test whether parents use a positivebias in explaining their own children's behaviour, Gretarsson andGelfand (1988) conducted structured interviews in which motherswere asked to describe and form attributions about theirchildren's recent behaviours. Positive parental bias was found23in mothers' perceptions of their children, except when thechildren were seen as difficult to control. That is, mothers ofnonproblem children explained their children's prosocialbehaviour by referring to personality dispositions and attributedtheir children's' misdeeds to situational influences. Theauthors point out that this finding replicates Dix's earlier work(Dix & Grusec, 1985) showing that parents generally attributeboth real and fictional positive child behaviours todispositional, intentional causes, and attribute negativebehaviours to more situational causes. Thus, it appears thatmothers use a favorable attributional style and extend a positiveattributional bias to their children when recalling behaviours oftheir own children, imagining their own child behaving inpredetermined ways, or simply reading about fictional children.However, because these different methods and stimuli were notcompared within a single study, it is not known whether favorableattributions appear more or less in response to parents' ownchild compared to unrelated or fictional children. That is,previous studies fail to address the question of whether mothers'positive attributions for child behaviour reflect a motivationalbias stemming directly from the parent-child relationship or apervasive cultural schema regarding the general capacities ofchildren. To answer this question, examination of the degree ofcorrespondence between mothers' attributions for their own andother children must be performed within the same study.Larrance and Twentyman (1983) examined this question intheir study comparing the attributions of abusive, neglectful,24and control parents which included an "own vs. other" childmanipulation. In accordance with predictions that parents extendpositive biases to their own children, there was a trend forcontrol mothers to form significantly less internal and stableattributions in explaining their own child's behaviour than whenexplaining the behaviour of an unknown child. Interestingly,abusive parents showed the opposite pattern of results, andblamed their own children more than other children for negativebehaviours.The study reported here examined the relationship betweenchild type (own vs. other) and parent attributions for childbehaviour in a nonclinical sample. Given the intimacy andcloseness of the parent-child relationship and the knownpreference of individuals to form attributions casting themselvesin a favorable light, it was hypothesized that attributionsparents form regarding their own children's behaviour would bemore favorable than those formed about unknown children. In thisstudy, written descriptions of child noncompliance were used toelicit mothers' attributions.Mother Inconvenience Aside from the vicarious emotions parents experience inresponse to their own children's behaviour, child behaviour oftenhas a direct impact on the parent. Given the nature of caregiverrelationships, particularly with young children, parents are inclose, and sometimes constant, contact with their children forextended periods of time. As a result, parents are often eitherthe direct beneficiaries or "victims" of their children's25behaviour. For instance, as mentioned earlier, a child'snoncompliant behaviour may upset parents because the behaviour isconsidered to be reflective of poor parenting ability. However,parents may also be upset by children's noncompliant behaviourbecause the misbehaviours inconvenience them in some morepractical manner. For instance, a parent whose child refuses toclean off the table to allow dinner to be served may well beforced to perform the task him or herself. The hassle of havingto perform the task may result in the activation of parentemotion, and may affect parent attributions and choice of parentbehaviour strategies.The parent emotion-attribution link was addressed earlier inDix's work where parents' transient affective states werecorrelated with attributions for child behaviour (Dix, 1989;1991). Dix has also described a model of parenting that stressesthe important influence of emotion on parenting responses tochildren's behaviours (Dix, 1991). Essentially, this modeldescribes how child, parent and contextual factors activateparent emotion, which, once activated, affects parentingbehaviour. Central to this model is the parent's ability toadopt child-centered motives as opposed to parent-centeredmotives. Child-centered motives involve the pursuit of goalsthought to ultimately benefit the child, whereas parent-centeredmotives are directed aimed at profiting the parent. Adoptingchild-centered motives may involve parents setting aside theirown personal desires or wishes, and may be diminished insituations where the parent is emotionally aroused. For26instance, a mother may fail to activate child-centered motivesbecause she feels rushed, angry, or inconvenienced. In such asituation, the parent's behaviour may be fuelled by self-focusedmotives (e.g., use of harsh discipline strategies to terminate anaversive child behaviour). These behaviours are regarded as lessbeneficial for the child because, although they result in short-term gains for the parent (the aversive child behaviour isstopped), they fail to produce long-term gains for either theparent or the child (the child does not learn more appropriatebehaviours). The model therefore ties parent behaviour andaffective state together within the context of specific parent-child interactions.Given the numerous opportunities for parents to beinconvenienced by their children's negative behaviours, thegreater likelihood of emotional activity in situations in whichthe parent is inconvenienced, and the link between parentemotional activity and attributions (Dix & Reinhold, 1987),inconvenience may be an important influence on parents'attributions. Previous research has failed to address theinfluence of the direct impact of child behaviours on parentattributions and responses. For instance, in the example of themother who asks her child to clear his homework off the table, itis not known whether the mother's attributions would differ ifthe child's refusal to clean off the table did not inconvenienceher (e.g., if the child was working at a table other than thedinner table). In this second example, the mother would be lessinconvenienced, may experience less negative affect, and may be27more able to adopt child-centered motives. Thus, herattributions for the child's behaviour may be more favorable.This study examined whether parent attributions for childbehaviour change depending on the direct impact of the behaviouron the parent. We know from previous research using situationsin which the parents are not inconvenienced, that parentsgenerally form favorable attributions about children. Forinstance, when faced with a choice between blaming a negativechild behaviour on the child or on the situation, parentsgenerally blame the situation. In those types of circumstances,after all, parents have little to lose (other than failing toreprimand a "bad" behaviour), and everything to gain (e.g., mychild is innocent, therefore I am a good parent) by "letting thechild off the hook" and faulting a source external to the child.However, if the situation involves the parent instructing thechild to do something, and the child's noncompliance negativelyimpacts on the parent, the parent has more at stake in formingtheir attributions. Some preliminary evidence provided a clue towhat might be expected in those cases. In the previous study weconducted on parent attributions (Geller et al., 1991), parentswere asked to describe recent negative events involving theirchildren, and to respond to attribution questions regarding thelocus, controllability, specificity, and stability of these childbehaviours. In a small proportion of the negative experiencesmothers described for their children the mothers were directlyaffected by the child's negative experience. A comparison of theattributions mothers formed in those "inconvenienced" situations28with those in which they were "uninvolved" revealed significantlymore negative attributions in the "inconvenienced" situations.However, the small sample of "inconvenienced" situations, animprecise definition of "inconvenienced", and the absence ofcontrol imposed on the types of situations recalled render thesefindings difficult to interpret.The study reported here examined the relationship betweenthe behavioural impact (parent inconvenience) of noncompliantchild behaviour and parent attributions. It was hypothesizedthat the attributions parents form in situations in which theyare inconvenienced would be less favorable than those formed insituations in which they are not inconvenienced by childbehaviour. To assess this relationship, an operationaldefinition for inconvenience was provided and predeterminedsituations were used as stimuli so that all parents responded tothe same conditions.Secondary Research Questions The next two sections describe variables that may be relatedto parent attributions for and responses to child behaviours.Because relatively little is known about the relationship betweenthese variables and parent responses in the context of parent-child interactions, predictions are conjectural and the analysesconsidered exploratory.Parent Empathy Definitions of empathy have varied greatly,but recent conceptualizations have included the integration ofboth cognitive and affective components (Eisenberg & Strayer,1987; Feshbach, 1978; Hoffman, 1977). In this integrative-29affective model, the empathic reaction is postulated to be afunction of an individual's cognitive ability to discriminateaffective cues from others and to assume the perspective and roleof another person, and the individual's emotional responsiveness(Feshbach, 1987). This section will briefly review the knownlinks between parent empathy and parent behaviour.Empathy can be considered an interactional, relationship-focused variable that may be related to parent responses in thecontext of parent-child interactions (Egeland & Sroufe, 1981;Feshbach, 1987; Pulkinen, 1982). For instance, parent empathy isconsistently inversely related to parent aggressiveness (Miller &Eisenberg, 1988), and empathy scores in parents are positivelyrelated to child self-control, parent investment, involvement andaffect (Feshbach, 1987). These findings have been consistentacross samples of abusive mothers, nonabusive mothers attendingchild guidance clinics, and normal control mothers (Feshbach,1987). Additionally, in a 12-year longitudinal study, Pulkinen(1982) found that child-centered guidance, which includedparents' consideration of the child's opinions, sustainedinterest and control over the child's activities, and parenttrust and warmth, was a prerequisite for the child's laterdevelopment of self-control. Empathic responses and behavioursin parents are therefore positively related to a number offavorable outcomes in children.The relationship between empathy and parenting behaviour isconsistent with Dix's (1991) work addressed earlier regarding theimpact of parents' emotional activation on subsequent behaviour.30Dix stated that mothers who fail to activate child-centeredmotives because of negative emotional arousal, (e.g., angered bya prior child behaviour) may concentrate on self-focused, ratherthan child-oriented motives. This child- vs.. parent-centereddistinction is similar to what some studies in the parent empathyliterature have referred to as empathic vs. personal distressresponses. Combining the attributional, empathy, and emotionalactivation perspectives, it was expected that parents who reportlower degrees of parental empathy would experience strongernegative emotional activation, form less favorable attributionsregarding their children's behaviour in situations in which theywere inconvenienced, and would employ more negative parentbehaviours. In contrast, it was predicted that high empathyscores would "buffer" the negative impact of beinginconvenienced.Parenting Investment Similar to parent empathy is theconstruct of investment in the parenting role. It is presumedthat the tendency of "normal" parents to perceive the causes oftheir children's behaviour favorably reflects, at leastpartially, a motivation to view themselves in a favorable light(i.e., "when my child does good, I do good"). But what aboutwhen the parent, for some reason, does not see him or herself aslinked to the child, and therefore does not see child behaviouras reflective of him/herself? If the child is not seen as beingan extension or in any way a reflection of oneself, thenmotivation to view that child's behaviour in a favorable lightmay be reduced. Although we do not know why some parents form31more negative attributions regarding their own children'sbehaviour than other parents, it is possible that parents who areless invested in their role as parents do not see their childrenas extensions of themselves, and therefore fail to extendpositive attributional biases to their children. This studyexamined the relationship between intensity of parent investmentand parents' attributions for their children's behaviour to helpclarify this matter.PurposeThe purpose of this study was to determine the impact of tworelationship-focused variables, child type and motherinconvenience on the formation of maternal attributions. Therelationships between parent attributions for child behaviour andparent affect and anticipated parent behaviour were alsoexamined. Finally, the relationships among parent empathy,investment in parenting, and parent attributions for childbehaviour were investigated.32CENTRAL HYPOTHESES:1. Mothers' attributions for the cause of their own children'smisbehaviour will be more favourable (i.e., less internal,global, stable, and controllable, and more external) thanattributions formed about unknown other children.2. Mothers' attributions for the cause of children'smisbehaviours when the misbehaviour does not inconveniencethe mother will be more favourable (i.e., less internal,global, stable, and controllable, and more external) thanattributions formed when the misbehaviour does inconveniencethe mother.SECONDARY HYPOTHESIS:3. Mothers' attributions will be related to anticipated affectiveand behavioural responses for children's behaviour. It waspredicted that less favourable attributions for childbehaviour (i.e., more internal, controllable, global, andstable attributions, and less external attributions) wouldbe related to stronger anticipated affect and to mothers'increased likelihood of doing something about the child'sbehaviour.4. Mothers' attributions will be related to empathy andinvestment in parenting. It was predicted that morefavourable attributions for child behaviour (i.e., lessinternal, controllable, global, and stable attributions, andless external attributions) would be related to higherempathy and investment in parenting scores.33Study DesignTo test the hypotheses of this study, written stimulusmaterials were used to elicit mothers' attributions for childbehaviours. Mothers were asked to imagine various situations inwhich a child fails to comply with their request. In half of thesituations, the child was described as their own, and in theother half, as an unknown child of the same age and gender astheir own child. For half of the mothers, the situation involvednoncompliant child behaviour that inconvenienced the mother, andfor the other half of the mothers, the same noncompliant childbehaviour resulted in a mild form of property damage but did notdirectly inconvenience the mother. After reading each situation,mothers were asked to respond to a series of questions regardingtheir attributions for the cause of the child behaviour(assessing attributions of child internality, externality,stability, globality, and controllability), and their anticipatedaffective and behavioural responses to the child behaviour.The attribution dimensions of stability, locus, andcontrollability are from Weiner's attribution model (1979). Theglobality dimension has been used in assessing the attributionsof depressed individuals (e.g., Seligman, Abramson, Semmel, & VonBaeyer, 1979). Similar questions based on these four dimensionshave been used in a number of studies of parent attributions andhave distinguished between different parent characteristics,(e.g., Larrance & Twentyman, 1983), child characteristics (e.g.,Sobol et al., 1989), and child behaviours (e.g., Dix & Grusec,1985). Previous work in our lab (Geller et al., 1991) has also34used these dimensions, and found that "globality" and "stability"scores predicted mothers' ratings of child conduct problems, and"internality" and "controllability" scores were sensitive tochanges in maternal depressed mood.Whereas most investigators have focused solely on mothers'attributions regarding the child's causal role in noncompliantbehaviour, this study also examined the extent to which motherssaw their own behaviour as causally related to the child'snoncompliance. Bugental and Shennum (1984) found that inpredicting mothers' behaviour in interactions with childconfederates, it was necessary to consider mothers' perceptionsof both her own and the child's ability to influence theinteraction. In addition, although the link between parentaffect and parent attributions regarding the child's role incausing the behaviour has been postulated (e.g., Dix, 1991), therelationship between mothers' attributions regarding their ownrole in the parent-child interaction and their affectiveresponses has not yet been explored. Therefore, in this study,mothers were also asked to rate the extent to which they thoughtthe child's noncompliance occurred because of something aboutthemselves (mother internality) and because of something thatthey had control over (mother controllability).Although the generalizability of responses to writtenstimulus situations to responses to real child behaviours is notassured, previous research in our lab (Geller et al., 1991) hasfound similar patterns of results using mothers' responses toanalogue situations and to recalled events from real life. In35addition, the use of an analogue methodology to control forsituational variations was considered appropriate because of thesituational specificity of cognitions (Miller & Ross, 1975;Nisbett & Ross, 1980). Given that individuals' attributions areinfluenced by the contextual details of situations, analoguesituations that control multiple facets of the situation wereneeded to isolate the effects of the manipulated variables. Anadditional advantage of having subjects imagine writtensituations and their responses to these situations is that thismethodology does not rely on participants' recall of real lifesituations. Therefore, the potential problem of attributionsbeing influenced by participants' knowledge of the ultimateoutcome of the situation is avoided. Finally, although childconfederates could have been used to control for situationalvariations while still presenting mothers with real childbehaviours, this methodology was considered undesirable due tothe difficulty of making such situations believable, and becauseof the need for deception.36MethodProcedureMothers responded to advertisements in local newspapers andcommunity centers by telephoning the principal investigator.Callers were provided with a rationale and overview of the study,and descriptive information gathered during this telephonecontact determined eligibility for participation. Eligiblemothers had at least one male child in the home between the agesof 6 and 10 years. The target child for the research wasidentified as the oldest male child in the family between 6 and10 years of age. To meet eligibility criteria, this child wasrequired to be biologically-related to the mother. In addition,because the instructions, stimulus materials, and questionnairesrequired a minimum level of literacy in English for completion,mothers who had completed less than 10 years of schooling werenot included in the study.If mothers were interested in participating, they weremailed a packet of questionnaires and were asked to find a quiettime at home to fill out the measures. Mothers were randomlyassigned to one of two behavioural impact conditions. Mothersassigned to the inconvenience condition received packetscontaining child behaviour descriptions which personallyinconvenienced mothers. Conversely, mothers assigned to the noinconvenience condition received packets containing childbehaviour descriptions which resulted in a mild form of propertydamage but did not directly inconvenience the mother. Motherswere informed during the telephone contact that they had the37right to refuse to participate or to withdraw from the study atany time, and that the return of completed questionnaires wouldbe indicative of consent for participation. The questionnairesrequired approximately 45 minutes to complete.Subjects One hundred and thirty-one mothers of children aged 6 to 10years were mailed questionnaire packets, and 100 mothers returnedcompleted packets, representing a 76% response rate. Mothersranged in age from 28 to 46 years (M = 36.52) and 18 were singlemothers. Family socioeconomic status ranged from 1 to 5 on theHollingshead Four Factor Index (Hollingshead, 1975), with anaverage status of 2.82 (upper middle class). The average childage was 7.89 years. Sixty-three of the children were the oldestor the only child in the family. The number of children in eachfamily ranged from 1 to 5, with a median of 2. All mothers werebiologically related to the target child. Forty-two percent ofmothers reported they had at some time sought help for their ownpsychological problems, and 16% reported they had sought help forpsychological problems in the target child. Mothers wererandomly assigned to the inconvenienced (n = 46) or notinconvenienced (n = 54) behavioural impact condition. The twogroups did not differ significantly on any of the precedingvariables.Measures Each mother received a package containing: an explanatorycover letter (Appendix A), a demographic information form(Appendix B), instructions for the attribution ratings (Appendix38C), the stimulus situations each followed by attribution, affect,and behaviour questions (see Appendix D), a modified version ofthe Perspective-Taking subscale and the Empathic Concern subscaleof the Interpersonal Reactivity Index (Davis, 1980) (Appendix E),and the Investment in Parenting subscale of the Parenting RoleScale (Polzien & Abidin, 1991) (Appendix F). Mothers wereinstructed to complete the questionnaires in the order justdescribed. The questionnaire packet also contained threemeasures not addressed in this proposal that were given at theend of the other measures and were not expected to influenceresponses to the preceding questionnaire. These includedmeasures of parent depressed mood, child behaviour problems, andgeneral parent attributional style. Finally, the packagecontained a stamped, addressed envelope in which the completedquestionnaires were returned.Demographic Information Mothers were asked to provide:their age, occupation, highest level of education, the occupationand highest educational level of the father, ethnicity, andfamily status (e.g., two-parent or one-parent family). They werealso asked to provide the age and gender of their child(ren), andto indicate whether they were the biological mother of eachchild. Finally, mothers were asked to indicate whether they hadever sought help for child or maternal psychological problems.This demographic information was used to confirm eligibility ofsubjects. No subjects who returned complete questionnairepackets were eliminated from the study.3 9Instructions for Ratings Prior to reading and responding tothe stimulus situations, mothers read a cover letter providingdetailed instructions for the attribution, affect, and behaviourratings (Appendix C). They were asked to read each situationimagining that, as indicated, the situation involved themselvesand either the target child (in "own child" descriptions), or afictional child, "Johnny" (in "other" child descriptions). Inthe instructions for ratings preceding the stimulus situations,"Johnny" was described as a boy who was the same age as thetarget child. For the "own child" situations, the target child'sname was written in the appropriate spaces in the questionnaireinstructions. Mothers were asked to imagine that each situationwas a separate event, and were reminded that there were no rightor wrong answers.Stimulus Materials Thirty-two stimulus situations (16Inconvenience and 16 No Inconvenience) were pilot-tested usingfemale undergraduate psychology students. Students read varioussituations involving an unknown boy failing to comply with theirrequest. For each situation, students were asked to rate theextent to which they would be inconvenienced by the situation,the extent to which they thought the situation was realistic, andthe age of the boy in each situation. The first two ratings weremade on 9-point Likert scales, ranging from "not at allinconvenienced" to "extremely inconvenienced," and "extremelyunrealistic," respectively. For the boy's age estimation,students were asked to choose among three age groups; preschoolaged (0 to 5), school-aged (6 to 10), or pre-adolescent (11 to4015). From these situations, five inconvenienced situations thatdepicted approximately equivalent degrees of impact of childbehaviour on the mother (M = 6.71, SD = .36) with their five notinconvenienced counterpart situations that depicted approximatelyequal levels of low impact of child behaviour on the mother (M =3.67, SD = .42) were selected. The selected stimulus situationshad mean realistic ratings of 5.70 or more (M = 6.54, SD = 1.80),and were estimated to be reflective of the behaviour of boys inthe 6 to 10 year old range by a mean of 70% of the respondents.Pilot testing thus ensured that the inconvenienced situationswere more bothersome than the not inconvenienced situations, andpermitted selection of situations perceived as realistic anddepicting behaviour typical of 6 to 10 year old boys.Using the 10 situations selected from the pilot testing, atotal of 20 situations was created. For each situation selected,an "own child" and "other child" description was written,involving either the target child, or "Johnny," respectively.Mothers therefore read various situations involving either thetarget child or "Johnny" failing to comply with their request.Half of the mothers read situations in which both children's(target and "Johnny") behaviour inconvenienced the mother, andhalf of the mothers read situations in which the child behavioursdid not inconvenience the mother. In order to eliminate theinfluence of order effects, the situations were presented in arandom order across mothers. However, precautions were taken toensure that two versions (target and "Johnny") of one situationwere not presented sequentially. The following are examples of41inconvenienced and not inconvenienced situation types for the"own" child condition:OWN/NO INCONVENIENCEYou are waiting at the check-out line in the grocery store.You are standing in front of your cart, and child's name isat the back of the cart, holding onto the handle.^ starts swinging on the handle, pushing thecart forwards and backwards with each swing. You ask him tostop playing with the cart and be still. After a fewminutes, you see the cart bang into the check-out counter as^ pushes it forward.OWN/INCONVENIENCEYou are waiting at the check-out line in the grocery store.You are standing in front of your cart, and ^is at the back of the cart, holding onto the handle.^ starts swinging on the handle, pushing thecart forwards and backwards with each swing. You ask him tostop playing with the cart and be still. After a fewminutes, you feel the cart digging into your ankles as^ pushes it forward.After reading each situation, mothers were asked to describein their own words what they thought caused the child'sbehaviour. These responses were content analyzed and categoriesreflecting the various causes mothers reported were rationallyderived. Responses were first categorized into causes consideredto be either internal or external to the child. Responsesreflecting something about the child (e.g., "he may be testing42me") were coded as internal, and responses reflecting somethingabout the situation (e.g., "because of the public place, he mayhave thought my reaction would be minimal") were coded asexternal. Internal causes were subsequently further broken downinto blaming and nonblaming responses. Blaming causes (e.g., "hewas showing off for friends by not responding") were coded asinternal/blaming, and nonblaming causes (e.g., "he had too muchenergy") were coded as internal/nonblaming. Therefore, mothers'written descriptions of what caused the child's behaviour werecoded into one of three categories; internal/blaming,internal/nonblaming, and external. Twenty-five of thequestionnaires were randomly selected and coded separately by twoindependent coders, and kappa reliability coefficients werecalculated for each of the categories. Kappa coefficients forthe three categories were .77, .76 , and .49, forinternal/blaming, external, and internal/nonblaming, categories,respectively, yielding an overall kappa value of .67.Following the open-ended question concerning the cause ofthe child's behaviour, mothers were asked to rate on 9-pointscales, the extent to which they thought the cause of the child'sbehaviour was internal (on a scale ranging from "totally due tochild" to "not at all due to child"), external (on a scaleranging from "totally due to external circumstances' to "not atall due to external circumstances"), global (on a scale rangingfrom "influences just this particular situation" to "influencesmost situations"), stable (on a scale ranging from "lasting" to"a one time thing"), and controllable (on a scale ranging from43"not at all under the child's control" to "completely under thechild's control"). Mothers were then asked to rate the extent towhich they thought the child's noncompliance occurred because ofsomething about themselves (on a scale ranging from "not at alldue to me" to "completely due to me") and because of somethingthat they had control over (on a scale ranging from "not at allunder my control" to "completely under my control"). The ratingscales are provided in Appendix D.In addition to attribution ratings, mothers were asked toindicate their affective and behavioural responses to eachstimulus situation. The affect rating (Appendix C) was designedto provide information about the mother's anticipated emotionalresponses to the situations. Mothers were asked to rate theextent to which they would be upset by each of the situations ona 9-point Likert scale ranging from "not at all upset" to"extremely upset". This measure also served as a check for the"inconvenience" manipulation. This rating scale has been used inprevious work (Dix & Grusec, 1985; Geller et al., 1991; Krech &Johnston, 1991) and found to be sensitive to variations in childbehaviour stimuli.Two items assessed mothers' anticipated behaviouralresponses to the situations. First, mothers were asked to ratehow likely they would be to do something about the child'sbehaviour on a 9-point Likert scale ranging from "not at alllikely" to "extremely likely." Second, they were asked todescribe, in their own words, what they would do if theyexperienced each of the situations (see Appendix D). These44responses were content analyzed and categories reflecting thevarious behaviour strategies mothers reported were rationallyderived. In cases where mothers provided more than onebehavioural strategy, only the first was coded. Responses werecoded into one of five categories; punishment, inactive response,shift of focus, talking, and reprimand. Punishment was coded ifthe mother described either making the child do something (e.g.,"make child apologize") or removing a privilege from the child(e.g., "take his plane away"). Inactive responses includeddescriptions of ignoring the child or failing to respond to thechild behaviour (e.g., "I wouldn't do anything"). Talking wascoded for any response that involved reasoning, explaining, ortalking to the child about the behaviour (e.g., "let him see theconsequences of not listening, and ensure he understood").Reprimand was coded for responses that involved scolding orspeaking sternly to the child about his behaviour (e.g., "Takehim aside and speak to him sternly"). Finally, shift of focuswas coded for responses in which the mother attempted to involvea third person (e.g., "call the janitor" or "talk with a parent")or distracted the child (e.g., "find something more interestingto do"). As above, 25 of the questionnaires were randomlyselected and coded separately by two independent coders, andkappa reliability coefficients were calculated for each of thecategories. Kappa coefficients for the five categories were asfollows: inactive response (.85), talking (.83), shift of focus(.74), punishment (.68), and reprimand (.57), yielding an overallkappa value of .73.45Parent Empathy Parent empathy was assessed using a modifiedversion of the Perspective-Taking (PT) subscale of theInterpersonal Reactivity Index (IRI) and the Empathic Concern(EC) subscale of the IRI (Davis, 1980) (Appendix E). These 7-item subscales measure the tendency to adopt the point of view ofother people and to experience warmth, compassion, and concernfor others, respectively. The PT subscale was adapted to applyspecifically to mothers' relationships with children. Forinstance, the item "When I'm upset with someone, I usually try to'put myself in his or her shoes' for a while" was changed to"When I'm upset with a child, I usually try to 'put myself in hisor her shoes' for a while." For each of the statements, motherswere asked to indicate on a 5-point Likert scale, the degree towhich they agreed with the statement. Higher scores indicatehigher empathy. The adult versions of these scales havedemonstrated moderate internal consistency and test-retestreliability (Davis, 1980). For this study, Cronbach's alphacoefficients were calculated for the two subscales, yieldingvalues of .76, and .44 for the modified Perspective Taking andEmpathic Concern subscales, respectively. Because the internalconsistency of the empathic concern subscale was low, adjustedalphas were calculated deleting each of the subscale items.Deletion of item 3, ("When I see someone being unfairly takenadvantage of, I feel kind of protective toward them") increasedthe internal consistency to .59. As a result, this modified ECsubscale was used in subsequent analyses.46Parenting Investment The extent to which participants wereinvested in their role as mothers was assessed using theIntensity of Investment subscale of the Parenting Role Scale(Polzien & Abidin, 1991) (Appendix F). Mothers were asked toindicate the extent to which they agreed with eight statements on5-point Likert scales. Sample statements include "My childrenare my greatest source of pride" and "It is very important thatmy children feel close to me." Higher scores indicate strongercommitment to and investment in parenting. This scale hasdemonstrated fair internal consistency (standardized item alphais .79) and correlates with measures of parenting stress,attachment, and parenting style (Polzien & Abidin, 1991). Inthis study, the Cronbach's alpha internal consistency coefficientfor the Intensity of Investment subscale was .70.47ResultsStudy DesignA 2 X 2 (Child Type X Behaviour Impact) design was used totest the primary hypotheses. The child type (own/other)manipulation was a within-subjects comparison, and the behaviourimpact (inconvenienced/not inconvenienced) manipulation was abetween-group comparison. Using this design, each motherresponded to two versions of each of five situations. Oneversion of the situations involved the mother's own child, andthe other version involved another child of the same age andgender. Half of the mothers read 10 situations in which theywere inconvenienced (i.e., five own child/mother inconvenienced,and five other child/mother inconvenienced situations) and halfof the mothers read 10 situations in which they were notinconvenienced (i.e., five own child/mother not inconvenienced,and five other child/mother not inconvenienced situations).Alpha-Protecting Strategies In order to minimize the possibility of committing Type 1errors, an alpha-protecting strategy was developed for each typeof planned analysis. For correlational analyses involving morethan one bivariate relationship, a conservative alpha level of.01 was used in interpreting the results. Correlationcoefficients with significance levels in the range of .01 to .05were interpreted as trends. In interpreting multivariateanalysis of variance (MANOVA) results, univariate tests were onlyconsidered when the overall multivariate F value for the relevantfactor was significant. Post hoc analyses were performed using48Tukey's Honestly Significant Difference (HSD) test, a test knownfor maximizing control over familywise error rates when makingpairwise comparisons among group means (Howell, 1987). Ininterpreting regression analyses, the unique contribution of eachblock of variables in predicting variance in the dependentvariables was not examined unless the overall regression wassignificant. Similarly, the unique contribution of individualvariables within each block was not examined unless the block inwhich the variable was entered contributed significantly to thetotal explained variance. Finally, in the two instances in whichit was not possible to use a MANOVA in comparing group means andin which a series of analysis of variance (ANOVA) comparisonswere consequently conducted, alpha levels were adjusted bydividing .05 equally among the number of dependent variables.Preliminary AnalysesAggregation of Attribution Ratings. To determine the extentto which the attribution ratings for child behaviour wererelated, intercorrelations between the five ratings (childinternality, externality, controllability, globality, andstability) were performed. The correlations ranged from .09(internality with externality) to .76 (globality with stability),and more than half of the correlations were statisticallysignificant. Therefore, the ratings did not measure completelyindependent dimensions. As a result, a principle-componentsanalysis was performed to aggregate scores from the fiveattribution ratings for child behaviour. First, ratings on eachof the attribution questions were summed across the 10 situations49(five "own" and five "other" situations) for both inconveniencedand not inconvenienced mothers to yield total internality,externality, controllability, globality, and stability scores foreach mother. Initial principle-components analysis using thesetotal scores revealed two factors with eigenvalues greater than1.0 (2.16 and 1.18). Repeating the analysis forcing a two-factorsolution resulted in factors accounting for 43.2 and 23.7% of thevariance in ratings, respectively. Following Varimax rotation,globality and stability ratings loaded above .50 on the firstfactor (G/S), and internality and controllability ratings loadedabove .50 on the second factor (I/C). Externality ratings failedto load on either factor (factor loadings were .13 and .11 on theG/S and I/C factors, respectively). Table 1 shows the factorloadings of the G/S and I/C factors. Therefore, for analyses ofattributions for the child's behaviour, two child attributionscores, G/S and I/C, were computed by summing ratings across theglobality and stability dimensions and the internality andcontrollability dimensions, respectively. For each situation,each score could range from 0 to 18, with higher scorescorresponding to more global, stable, internal, and controllableattributions. Mothers' summed totals for G/S and I/Cattributions ranged from 2.4 to 16.7, with a mean of 9.1, andfrom 2.0 to 16.0, with a mean of 9.5, respectively.Relationships Among Coded Responses and Ratings. As notedearlier, mothers responded to two types of questions assessingtheir attributions for children's behaviours and theiranticipated behavioural responses. They provided written, open-50ended descriptions of what they thought caused the child'sbehaviours and what they thought they would do in each situation.They also responded to rating scales examining specificattribution dimensions and made a rating of their likelihood ofreacting to each child behaviour. For purposes of clarity indistinguishing these two types of responses, scores from theopen-ended questions will be referred to as coded responses, andscores from the Likert scale items will be referred to asratings.The relationship between the coded responses and theircorresponding rating scales was examined. Mothers' codedresponses to the open-ended question assessing what caused thechild's behaviour were summed across the 10 situations to obtaintotal scores; internal/blaming (M = 3.3), internal/nonblaming (M= 4.7), external (M = 1.7), and uncodable responses (M = .3).Similarly, total attribution ratings were calculated by summingthe G/S and I/C ratings across the 10 situations (bothinconvenienced and not inconvenienced situations). Therelationships between coded attribution responses and attributionratings for child behaviour were then examined using Pearsoncorrelations. These correlations are reported in Table 1. Asindicated in the table, internal/blaming coded responses werepositively related to G/S attribution ratings, and there was atrend for internal/blaming responses to be related to I/Cattribution ratings. There was also a trend for external codedresponses to be related to lower I/C ratings.51Mothers' coded responses to the open-ended questionassessing anticipated behavioural strategies were also summedacross the 10 situations to yield total scores. The mean of eachof the coded categories were as follows: reprimand (M = .75),shift of focus (M = .84), inactive response (M = .98), punishment(M = 3.33), and talking (M = 2.62), and uncodable responses (M =1.46). Similarly, a mean total anticipated behavioural ratingwas calculated by summing mothers' ratings across the 10situations. These ratings ranged from 1.0 to 8.6, with a mean of5.2 (out of a maximum possible rating of 9). The relationshipsbetween coded behavioural strategies and the behavioural ratingwere then examined using Pearson correlations. Thesecorrelations are reported in Table 2. As indicated, there was atrend for coded reprimand strategies to be positively related toanticipated behavioural ratings, and inactive strategies werenegatively related to behavioural ratings.Manipulation Check. To determine whether the inconveniencedsituations were perceived as more upsetting than the notinconvenienced situations, an independent t test was performedcomparing total anticipated affect ratings provided by mothers inthe inconvenienced condition (summed across the 10 situations)with those from mothers in the not inconvenienced condition.Confirming the manipulation, mothers in the inconveniencedcondition reported they would be significantly more upset aboutthe child behaviour than mothers in the not inconveniencedcondition, t(97) = 2.61, p < .01.52Main Analyses.Child Type and Behavioural Impact. To determine whethermothers' attribution ratings for the cause of children'sbehaviours were related to whether the children were their own(hypothesis 1) and to whether they were personally inconvenienced(hypothesis 2) by the situations, a two-way MANOVA, with onebetween-group factor (inconvenienced, not inconvenienced) and onewithin-group factor (own child, other child), was conducted, withG/S and I/C child attribution scores summed over the fivesituations of each type as dependent variables. Means andstandard deviations for these attribution ratings are presentedin Table 3. With the use of Wilks' criterion, the combineddependent variables were significantly affected by child type,F(2, 95) = 21.56, p < .001. The dependent variables were notaffected by behavioural impact or by a Child Type X BehaviouralImpact interaction. Follow-up univariate analyses revealed thatonly the G/S attribution rating was significantly affected bychild type F(1, 96) = 40.94, p < .001. As predicted, otherchildren's behaviours were seen as more due to global and stablecauses than were own child behaviours.A similar analysis was conducted using summed totals formothers' coded responses to the open-ended questions(internal/blaming, internal/nonblaming, and external) asdependent variables in three separate analyses of variance(ANOVAs), with one between-group factor (inconvenienced/notinconvenienced) and one within-group factor (own/other).Mothers' coded responses for each of the three attribution53dimensions were summed across the five situations of each type(own and other), so that the range of values for each responsewas 0 to 5. Separate ANOVAs, rather than MANOVAS, were used forthis set of analyses because the three variables(internal/blaming, internal/nonblaming, and external) werelinearly dependent. That is, knowledge of scores of any twovariables allowed the third to be predicted. Nonparametricanalyses were not used because this would have resulted in asignificant loss of power, and because of the difficulty of usinga within subjects factor (child type) in nonparametric tests.Means and standard deviations of the coded attribution responsesare presented in Table 3. A conservative alpha of .017 (.05/3)was used because of the number of comparisons. Significant childtype effects were detected for internal/nonblaming, F(1,98) =54.00, p < .001 and external responses F(1,98) = 54.05, p < .001.Mothers made more internal/nonblaming attributions for their ownchildren's behaviour than they did for other children'sbehaviour. However, they were more likely to attribute childbehaviour to external causes when thinking of other childrenversus their own children. Internal/blaming responses were notrelated to the manipulated variables. No behavioural impacteffects or Child Type X Behavioural Impact interactions weredetected.To determine whether mothers' attributions for their ownrole in causing the child behaviours were related to themanipulated variables, a two-way MANOVA was conducted, with totalmother internality and mother controllability (summed across the54five situations of each type) as dependent variables. Thisanalysis examined whether child type and behavioural impact wererelated to the extent to which mothers saw the child's behaviouras due to something about themselves (mother internality) and totheir perception of control over the behaviour (mothercontrollability). Means and standard deviations of the twomother attribution dimensions are presented in Table 4. With theuse of Wilks' criterion, the combined dependent variables weresignificantly affected by child type, F(2, 96) = 113.56, p <.001, and by the interaction between child type and behaviouralimpact F(2, 96) = 7.78, p < .001. Follow-up univariate analysesrevealed a significant child type effect for mother internality,F(1, 97) = 39.97, p < .001 and a significant Child Type XBehaviour Impact interaction for mother controllability, F(1, 97)= 12.28, p < .001. The child type by behavioural impactinteraction for mother controllability was followed up withTukey's Honestly Significant Difference (HSD) post hoccomparisons. This analysis revealed a significant child typeeffect across both inconvenienced and not inconveniencedsituations (p < .05), and no significant differences betweensituation types (inconvenienced/not inconvenienced) in either theown or the other child situations. Mothers rated the cause ofthe child's noncompliance as more due to something aboutthemselves, and as something that was more under their control inthe own child situations than in the other child situations.Finally, to determine whether mothers' affect andanticipated behaviour ratings were related to the manipulated55variables (own/other, inconvenienced/not inconvenienced), a two-way MANOVA was conducted with the total anticipated affect andbehavioural ratings (summed across the five situations of eachtype) as dependent variables. Means and standard deviations ofanticipated affect and behavioural ratings are presented in Table5. With the use of Wilks' criterion, the combined dependentvariables were significantly affected by child type, F(2, 96) =82.74, p < .001, behaviour impact, F(2, 96) = 3.45, p < .05, andby the interaction between child type and behavioural impact,F(2, 96) = 25.64, p < .001. Follow-up univariate analysesrevealed significant Child Type X Behavioural Impact interactionsfor both variables (F(1, 97) = 23.47, p < .001, and F(1, 97) =43.08, p < .001, respectively). Tukey's HSD post hoc comparisonswere used to follow up the significant Child Type X BehaviouralImpact interactions for anticipated affect and behaviouralratings. For the anticipated affect rating, in the notinconvenienced situations, mothers reported stronger affect insituations involving their own child than in those involvinganother child (p < .05). No own/other differences were detectedacross the inconvenienced situations. In situations involvingother children, mothers reported stronger affect in theinconvenienced situations than in the not inconveniencedsituations (p < .05), but no inconvenienced/not inconvenienceddifferences emerged in the own child situations. Therefore,behavioural impact only influenced mothers' affective responsesto other children, but not their own. For anticipatedbehavioural response ratings, mothers anticipated responding more56in the own situations than in the other situations across bothlevels of situation inconvenience (R's < .05), but they were onlymore likely to respond when inconvenienced than when notinconvenienced in the other child situations (p < .05).To further examine the relationship between the manipulatedvariables and mothers' behaviour, mothers' coded anticipatedbehavioural strategies (punishment, inactive response, shift offocus, talking, and reprimand) were used as dependent variablesin five separate 2 X 2 analyses of variance (ANOVAs). Mothers'coded responses for each behavioural strategy were summed acrossthe five situations of each type (own and other), so that therange of values for each behavioural response was 0 to 5. As inthe previous analysis using coded response categories, separateANOVAs were used because the five variables were linearlydependent. Means and standard deviations of the codedattribution responses are presented in Table 5. A conservativealpha level of .01 (.05/5) was used in interpreting the resultsfrom these analyses. For punishment responses, mothers describeduse of punishment responses more with their own children thanthey did with other children, F(1, 98) = 81.82, p < .001.Inactive responses were related to child type, F(1, 98) = 38.54,R < .001, behavioural impact, F(1, 98) = 7.01, p < .01, and tothe interaction between child type and behavioural impact, F(1,98) = 8.42, p < .01. However, Tukey's HSD comparisons revealedsignificant differences only between own and other childsituations across both inconvenienced and not inconveniencedsituations. In both cases, inactive responses were more57frequently used for other children than for own children. Noinconvenienced/not inconvenienced differences were detectedacross either the own or other child situations. For shift offocus responses mothers were more likely to involve another adultor distract the child when the child was not their own, F(1, 98)= 7.90, p < .01. There was a trend for mothers to be more likelyto reason or explain things to the child (talking responses) whenthey were personally inconvenienced by the situation F(1, 98) =5.18, p < .05. Finally, there was also a trend for mothers to bemore likely to reprimand the child in "inconvenienced" situationsthan in "not inconvenienced" situations F(1, 98) = 3.94, p < .05.Relationships Among Attributions and Affect. To determinewhether mother attributions for their own and children'sbehaviour were related to their anticipated affective ratings(hypothesis 3), a multiple regression analysis was performedusing anticipated affect ratings as the dependent variable andchild age, the behavioural impact of the situation, and mothers'attributions for their own and the children's behaviour asindependent variables. In this analysis, attribution ratingswere summed across all 10 situations for each mother. Becausemothers differed in whether they had read situations in whichthey were inconvenienced or situations in which they were notinconvenienced, behavioural impact was dummy-coded (i.e., eitheras a 0 or as a 1 for the "inconvenienced" and "notinconvenienced" mothers, respectively) and used as a blockingvariable.58Variables were entered simultaneously and tested for theirunique contribution in predicting affect ratings. Three blocksof variables were entered into the regression. Behaviouralimpact and child age were entered in one block, the two summedattribution scores for child behaviour dimensions, G/S and I/C,were entered in a second block, and the two attribution ratingsfor mother behaviour, mother internality and controllability,were entered in a third block. The range of mother affectratings averaged across the 10 situations was between .3 and 8.1(M = 4.8). Mother internality and controllability ratingsaveraged across the 10 situations ranged from 0.0 to 6.8 (M =2.4), and from 0.0 to 7.6 (M = 3.6), respectively. Table 6displays the bivariate correlations between the independent andcriterion variables (r), the standardized regression coefficients(b) after entry of all six variables in the equation, the R 2change for each block of variables when entered last into theequation, and the total R 2 for the entire solution.When controlling for the block with child age andbehavioural impact, and the block with mothers' attributions fortheir own role in causing the child behaviour, mothers'attributions for the child's behaviour contributed significantlyto the model F(2, 90) = 6.48, p < .01. Inspection of thestandardized beta weights reveals that within the childattribution block, only I/C attributions contributedsignificantly to the prediction of affective ratings. Whencontrolling for the block with child age and behavioural impact,and that with mothers' attributions for the child's behaviour,59mothers' attributions for their own behaviour contributedsignificantly to the prediction of affective ratings (F(2, 90) =7.19, p < .001). Inspection of the standardized beta weightsreveals that within the block of mother attributions, only motherinternality ratings contributed significantly. Therefore,mothers' attribution ratings for both the children's and theirown behaviour accounted for unique variance in predicting affectratings. Specifically, higher I/C attributions for the child'sbehaviour, and higher personal internality attributions wererelated to stronger affective response ratings.Relationships Among Attributions and Behaviour. Todetermine whether mother attributions for child behaviour wererelated to their anticipated behavioural ratings (hypothesis 4),a similar multiple regression was performed using the anticipatedbehavioural ratings as the dependent variable and child age, thebehavioural impact of the situation, and mothers' attributionsfor their own and the children's behaviour as independentvariables. As in the regression predicting affective ratings,the independent variables were entered in three blocks;behavioural impact and child age, the two summed attribution forchild behaviour dimensions, G/S and I/C, and the two attributionratings for mother behaviour, mother internality andcontrollability. Table 6 displays a summary of the analysis.When controlling for the block with child age andbehavioural impact, and the block with mothers' attributions fortheir own behaviour, mothers' attributions for the child'sbehaviour failed to contribute significantly to the prediction of60behaviour. However, when controlling for the block with childage and behavioural impact, and that of mothers' attributions forthe child's behaviour, mothers' attributions for their own rolein causing the child's behaviour did contribute significantly tothe prediction of behaviour F(2, 90) = 5.50, p <.01. Inspectionof the standardized beta weights reveals that within the motherattribution block, only mother controllability ratingscontributed significantly to the prediction of behaviour.Therefore, in predicting the likelihood that mothers would dosomething about the child's behaviour, only attributions ofpersonal control accounted for a significant amount of variancein behavioural rating scores.Relationships among Affect and Behaviour. To determine therelationship between affect and behavioural ratings, a Pearsoncorrelation was conducted between the two ratings, revealing asignificant, positive relationship, r = .54, p < .001.Similarly, Pearson correlations were conducted betweenanticipated affective ratings and coded behavioural andattribution responses, and between coded attributions andbehavioural strategies. These correlations are reported in Table7. Similar to the regression analyses in which I/C attributionspredicted affect, coded internal child attributions were alsorelated to affect ratings. More internal attributions (bothblaming and nonblaming) were related to stronger affectiveresponses. Unlike the regression analyses in which attributionratings failed to predict the likelihood of mothers responding tothe child's behaviour, some of the specific behavioural61strategies were predicted by the coded attribution responses.For instance, internal/blaming attributions were positivelyrelated to reprimanding, and there was a trend forinternal/blaming attributions to be negatively related toinactive responses and to shifting focus. Internal/nonblamingattributions were negatively related to use of reprimandstrategies, and there was a trend for internal/nonblamingattributions to be related to use of talking strategies.Finally, anticipated affective ratings were positively related toreprimand strategies, and there was a trend for affect ratings tobe related to inactive responses.Empathy and Investment in Parenting. To determine whetherthe two empathy subscales of the IRI (perspective taking andempathic concern) and the Investment in Parenting subscale of theParent Role Scale were related to mothers' attributions for theirown and their children's behaviour, and to their anticipatedaffective and behavioural responses to child behaviours, Pearsoncorrelations were conducted. These correlations are reported inTable 8. As indicated in this table, higher perspective-takingscores were related to lower anticipated affective responses, andto lower internal/controllable attributions. No relationshipswere detected between mothers' attributions for their ownbehaviour and perspective taking. There was a trend for motherswho scored high on empathic concern to be more likely to feelthat they had personal control over the situations, and to beless likely to anticipate reacting to the children's behaviours.Finally, mothers who were more invested in their role as parents62were more likely to be upset and to do something about thechildren's behaviours. Mothers who were more invested as parentswere also less likely to attribute the child's behaviour tointernal, controllable causes.63DiscussionThis study examined whether mothers' attributions for childbehaviours and their anticipated affective and behaviouralresponses were affected by child type (own vs. other) and by thebehavioural impact of child behaviour. Results indicated thatwith their own children, mothers made less global and stableattributions about child noncompliance, and greater attributionsof personal internality and controllability than with otherchildren. Mothers anticipated stronger affective and behaviouralresponses with their own children than with other children, butwith other children only, stronger affective and behaviouralresponses were related to the behavioural impact of childbehaviours. Regression analyses indicated that mothers'affective responses were predicted by the behavioural impact ofthe child's behaviour and by mothers' attributions regardingtheir own as well as their children's role in causing thebehaviour. In contrast, mothers' behavioural responses werepredicted only by mothers' ratings of personal control over thechild behaviour. Mothers' ability to perspective take wasrelated to seeing the child's behaviour as less due to internal,controllable causes, and to lower anticipated affective responsesto the behaviour. Mothers' feelings of empathic concern wereassociated with lower likelihood of doing something aboutnoncompliant child behaviours, and to higher ratings of personalcontrollability over the situation. Finally, mothers who weremore invested in their role as parents were less likely to seethe child's behaviour as due to internal and controllable causes,64and were more upset by, and likely to do something about childbehaviour.Child Type and Behavioural ImpactAs predicted, mothers' attributions for child behaviour wereinfluenced by child type. Mothers were more likely to see childnoncompliance as reflective of global and stable causes (G/Sattributions) when the child was not their own than when thechild was their own. Contrary to prediction however,attributions of internality and controllability were not affectedby child type. Given that the design of the study was such thatmothers read each situation twice, with only the child in thedescription differing, the absence of a child type effect oninternality and controllability attributions may be due to thedifficulty of regarding the same behaviour as controllable in onechild and as uncontrollable in another child of the same age andgender. However, because attributions of globality and stabilityreflect mothers' expectations surrounding the likelihood of thebehaviour's recurrence in the future and in other situations,mothers (who have a history with their own children) may feelmore able to distinguish between their own and other childrenwith regards to such expectations. These findings are inagreement with the trend in Larrance and Twentyman's (1983) studyof mothers without a history of abuse to view the cause of theirown child's transgressions as less stable than that of an otherchild. Results from this study are consistent with the proposalthat the relational context of parents and children is animportant determinant of parent attributions of globality and65stability for child behaviours, and supports the proposedpositive bias parents may have concerning their own children'sbehaviours. In this study, mothers' desire to see their ownchildren's behaviour as nonproblematic (and therefore to seethemselves as good parents), may have allowed them to minimizetheir own children's noncompliant behaviour by viewing it as moreof a "one shot deal" than the identical behaviours performed byother children.Child type (i.e., own vs. other child) effects were alsodetected in mothers' coded responses to the attributionquestions. As predicted, internal nonblaming causes weredescribed more frequently with mothers' own child. Contrary toprediction, however, external causes were described morefrequently with another child. Therefore, although mothers wereless likely to blame the situation as the cause of their ownchildren's behaviour, when the cause was identified as within thechild, mothers were less likely to see the behaviours as theirown children's fault. Internal, blaming responses were notaffected by child type. These results can be interpreted byincorporating parents' extension of positive biases to theirchildren as described earlier in this paper with adults' generalperceptions of children's behaviours. Because the intimacy andcloseness of the parent-child relationship may result in parentsviewing children's behaviour as reflective of themselves, it wasproposed that parents may use a positive attributional bias inexplaining events in their own children's lives. In addition,because dispositional and controllable attributions, and66attributions of intentionality increase with child age (e.g., Dix& Grusec, 1985), and adult attributions about children reflectinferences regarding children's general capacities (i.e., use ofless blaming attributions in explaining the behaviour of childrenwho are perceived as less able; Fincham & Emery, 1988), theremay be a pervasive cultural schema to view the general capacitiesof children as low, and a general tendency to use nonblamingattributions in explaining children's behaviours. Combiningthese two ideas, the results from the coded responses can beparsimoniously interpreted. Given that parents have as theirprimary goal a positive conception of themselves (as suggested bythe social cognition literature (e.g., Fiske & Taylor, 1990),they will endorse nonblaming attributions for children and extendpositive attributional biases to their own children as long astheir primary goal (a positive self view) is not compromised.Therefore, consider first mothers' responses to other children.In this study, because other children's behaviour in thesituations may not have been seen as reflective of themselves (asevidenced by mothers' lower overall personal controllability andinternality ratings with other children), and mothers have ageneral tendency to excuse children's behaviours, mothers werefree to blame external causes (e.g., the children's parents) inexplaining the behaviour. With their own children, however,blaming external circumstances would be more difficult, sincemothers may have in part identified themselves as the externalsituation, and ego-preserving needs would take priority over thetendency to excuse children's behaviours. As a result, mothers67used more internal, but nonblaming attributional explanationswith their own children than with other children. Therefore,although with their own children, mothers used explanations thatidentified the cause of the behaviour as within the child, thecause was also described in a nonblaming fashion, such as "hedidn't realize that..." or "he was tired." In this way, theprimary goal of finding an explanation that reflected well onthemselves was accomplished, and both types of children (own andother) were blamed as little as possible. It should be notedthat because this is a post hoc explanation, further testing isrequired to determine the strength of these findings.Contrary to prediction, behavioural impact, or the extent towhich mothers were personally inconvenienced by child behavioursdid not influence mothers' G/S or I/C attribution ratings orcoded attribution responses for child behaviours. A number ofpossible explanations were considered. It is unlikely that theabsence of a behavioural impact effect on attributions for childbehaviour was due to a weakness in the manipulation because thesituations were pilot-tested to ensure that the inconveniencedsituations were significantly more bothersome than the notinconvenienced situations. In addition, a manipulation checkwithin this study determined that the inconvenienced situationswere significantly more upsetting than the not inconveniencedsituations. It is also unlikely that the failure to detect asignificant effect reflected a lack of statistical power. Inthis study, the power to detect a significant difference betweeninconvenienced and not inconvenienced conditions for a medium68effect size (with a sample of 100) was .71. Two possibleexplanations remain. First, it may be that noncompliance issimply a stronger stimulus than situation inconvenience indetermining mothers' attributions for child behaviour. That is,in considering the causes of child behaviour, mothers may be moreinfluenced by children's failure to comply than by the actualconsequences of the noncompliance. An alternative explanation isthat social desirability factors influenced mothers' responses tothe inconvenienced situations. Mothers may have felt that itwould reflect badly on them if they allowed personal afflictionsto influence their attributional responses to children'sbehaviours. Social desirability effects could be reduced byreminding parents that there are no right or wrong answers, andby encouraging honest responses.In contrast to mothers' attribution responses, where onlychild effects were detected, mothers' affective and anticipatedbehavioural response ratings were affected by an interaction ofchild type and behavioural impact of child behaviour. Withregards to affect, mothers indicated that they would be moreupset in response to noncompliance in their own child than inanother child, but only in situations in which they were notpersonally inconvenienced. In situations in which mothers wereinconvenienced, affective responses were the same with their ownand other children. Although contrary to prediction, thispattern of results suggests that the forces acting on mothers'responses to their own children are different from those involvedin their interactions with other children. With their own69children, the intimacy of the relationship and the investmentmothers have regarding their own child's behaviour may causechild noncompliance to be most salient in determining mothers'affective responses, overpowering any possible added influence ofbehavioural impact. On the other hand, the same behavioursperformed by a child about whom mothers did not feel a sense ofpersonal investment, did not elicit strong affective responsesunless mothers were personally inconvenienced by the behaviour.Once again, these results support the proposal that inunderstanding parent-child interactions, child behaviours cannotbe viewed in isolation, and the relational context, or pastinteractional history of the parent and child need to beconsidered.With regards to mothers' anticipated behavioural responseratings, mothers reported that they would be more likely to dosomething about their own child's noncompliance than that ofanother child, and that these own child ratings were highregardless of the extent to which they were inconvenienced.However, mothers anticipated responding more when they wereinconvenienced than when they were not inconvenienced by anotherchild's behaviour. These findings suggest that mothers have alower threshold to act with their own children than with otherchildren, and that similar to affective responses, mothers'behaviours with other children are more under the influence ofsituational cues, such as the extent to which they are personallyinconvenienced by the child behaviour. Relating these results tothe proposal that parents view their children and their70children's behaviour as an extension of themselves, the strongertendency to do something about own child behaviours may reflectmothers' wish to prevent undesirable child behaviours (whichreflect badly on themselves) from recurring. It is interestingthat although mothers made more negative G/S attributions withother children's behaviour, they reported greater anticipatedbehavioural responses with their own children. This pattern offindings, as with mothers' affect ratings, is contrary toprediction, and suggests that variables other than parentattributions for child behaviour may be involved in predictingparents' immediate behavioural responses to child behaviours.Specifically, the relational and situational contexts appear tobe determinants of parent responses to child behaviours. Itshould be noted that because explanations regarding the influenceof the manipulated variables on mothers' affective andbehavioural responses were developed post hoc, further researchis needed to confirm these findings. However, results from thisstudy extend previous research findings identifying situationalcues (e.g., public vs. private setting) as predictors of mothers'use of child discipline strategies (Holden, 1989). Together, thepattern of affective and behavioural results suggests thatoverall, mothers' own children's behaviour has personal relevancefor them, eliciting strong affective and behavioural responses,and that mothers' reactions to other children's noncompliantbehaviour is more under the situational influence of the impactof child behaviours.71The behavioural strategies mothers anticipated using werealso related to the manipulated variables. Contrary to theprediction that more negative attributional responses would beassociated with stronger anticipated behaviour, mothers were morelikely to use punishment strategies and less likely to use aninactive response style (i.e., do nothing or ignore) or a shiftof focus strategy (i.e., distract child or involve a thirdperson) with their own child than with another child. Thesefindings are consistent with mothers' higher anticipatedbehavioural ratings with their own children. Overall, thegreater use of active responses with own children and less activeresponses with other children may reflect the greater importancetheir own child's behaviour holds for mothers, and theirreluctance to discipline a child for whom they do not feelresponsible. Greater use of shift of focus strategies with otherchildren appears to be due to mothers' increased likelihood ofinvolving a third person, specifically, the other child's parent,in dealing with the situations. There was a trend for talkingand reprimand strategies to be used more in situations in whichmothers were inconvenienced by the child behaviour. The impactof situation inconvenience on talking responses may reflect thebelief that children should be made to understand theconsequences of their actions when someone is personallyinconvenienced. Mothers' greater use of reprimand responses insituations in which they were inconvenienced may be related tothe link established in this and other studies (Dix, 1989; Dix &Grusec, 1985; Dix & Reinhold, 1991) between reported affect and72behaviour. This interpretation is consistent with Dix's (1991)model of affective processes in parenting, in which parentemotions are seen as activated by a number of factors, one beingoutcome relevance to the parent (e.g., being personallyinconvenienced by child behaviour). Therefore, according toDix's model, in this study, situation inconvenience may haveinfluenced mothers' behavioural tendencies through the effect itexerted on affect.Finally, mothers' attributions regarding their own role incausing noncompliant child behaviours were also affected by childtype. Not surprisingly, in situations involving their own child,mothers rated noncompliant child behaviours as more due tothemselves and as more under their own control than in situationsinvolving an other child of the same age and gender. Theseresults support the proposal made in the introduction of thispaper that the closely bound relationship parents share withtheir own children causes parents to view their children'sbehaviours as personally relevant to, and therefore reflective ofthemselves. The impact of child type on mothers' attributions ofpersonal internality and controllability may also be related tothe association between mothers' stronger affective andbehavioural responses with their own children. According toDix's (1991) framework of affective processes in parenting, inwhich parent appraisals are identified as one of the multiplefactors eliciting parent emotional activation and subsequentbehavioural responses, mothers' appraisals that they wereinvolved in causing their own child's behaviour may have elicited73an emotional response, which impacted on the likelihood withwhich they anticipated responding. Detection of a child effectin mothers' attribution ratings surrounding their own role incausing noncompliant child behaviour, and the potential rolethese attributions had in eliciting mother affect and behavioursuggests that these self attributions make up an importantcomponent of mothers' cognitive processes during parentinginteractions. The behavioural impact of child noncompliance didnot influence mothers' attributions regarding their own role incausing child behaviour. It may be that in responding toquestions regarding the internality and controllability of theirown actions in causing child behaviour, the actual impact of thebehaviour was not as salient as the noncompliance itself. Thatis, although mothers may have seen themselves as exerting controlover the child failing to respond to their request, they may nothave seen the next step, that is, the impact of the child'snoncompliance, as related to their own behaviour.Relationships Among Attributions, Behaviour and Affect.Consistent with Dix's (1991) model of affective processing inwhich multiple factors are described as influencing parentemotional activation, three predictors of mothers' affectiveresponses to child noncompliance were identified in this study.It should be noted that although some variables are described as"predicting" other variables, conclusions regarding the directionof causality cannot be made. The relationships among parentattributions, parent affect, and parent behaviour are consideredto be bi-directional, and although parent attributions may74influence parent affect and parent behaviours, these influencesare also acknowledged to occur in the opposite direction. First,in predicting affect, the extent to which mothers were personallyinconvenienced by the child's behaviour accounted for uniquevariance in mothers' affect ratings, with higher affectassociated with greater situational inconvenience. Second, aspredicted, the extent to which mothers saw the child's behaviouras controllable and as due to the child (I/C attributions) alsoexplained significant unique variance in affect ratings.Finally, mothers' own attributions of personal internalitysignificantly predicted affective response ratings. Therelationship between affect and I/C attributions, but not G/Sattributions suggests that mothers' feelings about childnoncompliance were most strongly related to their answer toquestions such as "did he do it on purpose?" rather thanquestions concerning the representativeness of the behaviour forthat child in other situations and across time. The finding thatmothers were more upset when they felt that the child behaviourwas due to internal causes is consistent with Jones & Davis's(1965) model of correspondent inference, in which hedonicrelevance is related to stronger affective responses. Situationsin which mothers felt more involved in causing the childbehaviour had greater hedonic relevance for them, and wereconsequently associated with stronger emotional reactions.In contrast to predictors of affect, mothers' anticipatedbehavioural responses were only predicted by mothers'attributions of personal controllability over child75noncompliance. These findings suggest that mothers who fail toperceive their children's behaviour as within their own controlare less likely to respond to child noncompliance. Perhapsmothers who see noncompliance as personally uncontrollable alsofeel that they can do little to prevent the behaviour fromrecurring, or feel unable to take control of the situation oncethe behaviour has occurred. These results suggest that inhelping mothers take an active role in coping with childnoncompliance, it may be of benefit to assist mothers inmodifying their attributions and increase their feelings of self-efficacy in dealing with child behaviours. Indeed, recentresearch has shown that parenting sense of competence increasesin parents of ADHD children as a result of participation inBehavioural Parent Training (BPT) groups (Pisterman, Firestone,McGrath, Goodman, Webster, Mallory, & Goffin, 1992). Thedetected relationship between personal control and behaviouralresponses in this study is consistent with previous researchlinking internal parenting styles to parenting behaviour(Bugental & Shennum, 1984; Murphey & Alexander, 1991), andsuggests that mothers' self-efficacy beliefs surroundingchildrearing are most important in determining whether motherswill intervene in parent-child interactions.A number of relationships were also detected among affectratings, mothers' open-ended written descriptions of the cause ofchildren's behaviour, and the behavioural strategies they woulduse in response to the behaviour. The strongest attribution-behaviour link was between blaming attributions and greater use76of reprimand strategies. Therefore, although attribution-behaviour links were not detected using mothers' ratings of thelikelihood with which they would respond to child behaviours, arelationship was detected between specific behavioural strategiesand coded attribution responses. This discrepancy in findingssuggests that type of behavioural strategy may in some cases be amore sensitive measure of behaviour than the likelihood ofresponding at all. Internal, blaming attributions were relatedto stronger affective ratings, and to greater use of reprimandstrategies, supporting the appraisal, emotional activation,behavioural tendency sequence described in Dix's (1991) model.The pattern of attribution-affect correlations for both codedattributions and attribution ratings showed that locating thecause of the behaviour as within the child was related tostronger anticipated affective responses.Empathy and Investment in ParentingThis study examined relationships among perspective taking,empathic concern, and parenting investment as they relate tomothers' attributions and anticipated affective and behaviouralresponses to child behaviours. There was a trend for mothers'perspective taking scores to be related to less internal andcontrollable attributions about child behaviour, indicating thatthe more mothers "put themselves in children's shoes," the lesslikely they were to see the cause of noncompliant behaviours aswithin the child, and as under the child's control. This linksuggests that thinking about the problem from the child'sperspective may elicit what Dix (1991) refers to as "child-77centered" concerns, and less blaming attributions. Given thepositive relationship between internal blaming coded attributionresponses and use of reprimand strategies, it may be thatperspective taking ultimately results in less punitive responsesthrough the effect it exerts on I/C attributions. That is,perspective taking may result in a more thorough consideration ofthe forces involved in causing child behaviour, and this may inturn generate less blaming attributions, lower affectiveresponses (also detected in this study), and to the adoption ofmore "child-centered" behavioural tendencies. Together, thesefindings suggest that the relationships noted in previousresearch between parent empathy and favourable parentcharacteristics such as lower parent aggressiveness (Miller &Eisenberg, 1988) and greater parent self-control (Feshbach, 1987)may reflect an association between perspective taking and use ofless blaming attributions in explaining children's behaviour. Itis possible that the beneficial effect parent empathy exerts onchild outcome variables, such as the development of self-control(Feshbach, 1987; Pulkinen, 1982), may also reflect the impactthat empathy has on attributions, affect, and parents' use ofchild-centered child-rearing.Because the reliability of the empathic concern scale waslow, results using this variable need to be interpreted withcaution. However, the negative trend between empathic concernscores and anticipated behavioural responses suggests that higherfeelings of sympathy and concern are associated with a decreaseddrive to act on children's noncompliant behaviours. The positive78trend between empathic concern and mother controllabilitysuggests that mothers who had greater feelings of self-efficacyin their interactions with children sympathize more withchildren's experiences.There was a trend for parenting investment to be related tolower internality and controllability (I/C) attributions forchild behaviour, and to higher affective and behaviouralanticipated response ratings. The negative association betweenparenting investment and child I/C attributions is in accord withthe proposal made earlier that the more invested parents are intheir role as parents, the greater personal relevance theirchildren's behaviour holds for themselves, and the greatertendency they may have of using positive attributional biases inexplaining their children's behaviour. The relationship betweenparenting investment and anticipated affective and behaviouralresponses may reflect the association between parentinginvestment and the hedonic relevance of child behaviours. Giventhat child behaviours are more important for highly investedparents, they may be more likely to respond affectively, and todo something to prevent noncompliant behaviours from recurring.It should be noted that because many of the relationshipsamong perspective taking, empathic concern, and parentinginvestment with mothers' responses to child behaviours were onlysignificant at the .05 level, these findings should beinterpreted with caution. The goal of this study was to examinethese variables in an exploratory fashion, and further research79is clearly needed to determine the extent to which these findingsare generalizable to other situations.Relationships Among Coded Responses and Ratings The use of rating scales and open-ended questions inassessing mothers' attributions for child behaviour andanticipated behavioural responses allowed the relationshipbetween the two types of measures to be examined. With regardsto mothers' attributions for child behaviours, the significantpositive relationships between internal blaming codedattributional responses and both G/S and I/C attribution ratingssuggest that written responses were generally consistent withratings, although the magnitude of these relationships (r = .28and .23, respectively) was not overwhelming. Similarly, thenegative relationship between external coded responses and I/Cattribution ratings is in accord with expectations, though again,not of overwhelming magnitude (r = -.20). Although mothersspontaneously described external causes in their writtenresponses, and these coded responses were sensitive to childtype, the externality rating failed to load significantly ontoeither of the child attribution factor scores. Mothers' ratingsregarding the likelihood with which they would do something aboutchild behaviours were also related to actual behaviouralstrategies. As expected, anticipated behavioural responseratings were negatively related to inactive responses andpositively related to reprimand strategies. However, althoughreprimand and talking strategies were affected by behaviouralimpact across child type, anticipated behavioural responses were80only affected by behavioural impact in the other child condition.Furthermore, although the regression results failed to detect arelationship between mothers' attributions for child behaviourand anticipated behavioural response, coded internal blamingattribution responses were related to reprimand strategies.Together, although the convergence of findings on mothers'attribution and behavioural responses suggests that mothers'responses to ratings and to open ended questions were consistent,the two measure types clearly provide different, but equallyvaluable information. With regards to attributions, thediscrepancy between mothers' responses to rating scales and toopen ended questions suggests that the measures may in some caseshave been tapping qualitatively different dimensions. Forinstance, in the case of attributions of externality, the ratingscale may not have been as sensitive as mothers' coded writtenresponses. On the other hand, the rating scales providedinformation about a greater number of attributional dimensionsand were more readily quantifiable than coded responses. Inaddition, past research has reported higher reliability estimatesand better construct validity using rating scales than codedopen-ended attributions (Russell, McAuley & Tarico, 1987).Together, these findings suggest that inclusion of both measuretypes is advisable in future attribution research.A Relationship-focused Approach to Parent-Child Interaction The detection of child effects supports the proposal thatparents may extend positive attributional biases to their ownchildren, with whom they share a close, intimate bond. That is,81parents may not only tend to view their own negative actions asdue to situational factors (while others' negative actions areseen as due to dispositional causes), but they may also viewtheir children's negative behaviours as due to situationalfactors. Mothers' lower endorsement of globality and stabilityattributions for their own children's noncompliance when comparedto those formed about other children support this proposal.Higher globality and stability attributions suggest that thesebehaviours were seen as more stable across situations and time,implying more dispositional, as opposed to situational factors asdeterminants of other child behaviours, and a tendency to viewown child noncompliance as more of a "one shot deal." Withregards to mothers' coded responses, use of more externalattributions with other children and more internal nonblamingattributions with own children was explained by consideringmothers as balancing their ego-preserving needs and their driveto use nonblaming strategies with children in general. However,the unexpected relationship between external causes and otherchildren suggests that in predicting the extent to which mothers'extend positive attribution biases to their children, mothers'own ego-preserving needs may need to be considered. Thisrelationship could be further tested by using similar stimulusmaterials in which mothers are not directly involved in thesituations (e.g., Larrance & Twentyman, 1983). The negativerelationship between parenting investment and I/C attributionsstrengthened the proposal that parents' view of their children as82linked to themselves accounts for their extension of positiveattributional biases to their children.The effect of child type on mothers' responses to childrensupports use of a relationship-focused framework in understandingparent-child interactions. Specifically, mothers' experienceswith their own children appear to have a significant impact onthe way they see their own and their child's role in causingchild behaviours, as well as their affective and behaviouralresponses to the behaviours. There is, therefore, somethingunique about the parent-child relationship which distinguishesmothers' responses in interactions with their own and otherchildren. The child type by behavioural impact interactions alsosuggest that the forces acting on mothers' responses to their ownchildren are not the same as those involved in their interactionswith other children. Specifically, the relationship betweenbehavioural impact and mothers' affective and behaviouralresponses to other, but not their own children's noncompliance,suggests that child noncompliance in mothers' own childinteractions may overpower the added influence of behaviouralimpact.Limitations and Future ResearchAlthough analogue research is advantageous in its ability tocontrol manipulated variables, the generalizability of researchfindings to real-life parenting situations is not assured. Forinstance, in this study, it was not clear whether the failure todetect a behavioural impact effect on mothers' attributions forchild behaviour reflected social desirability factors, or whether83behavioural impact simply is not related to mothers'attributions. Future research could simulate real-lifesituations more closely by having parents recall past incidencesin which they were inconvenienced by child behaviours andprompting them to make attributions about the behaviours (e.g.,Gretarsson & Gelfand, 1988), or by showing parents videotapes ofthemselves interacting with their children and assessing theirattributions in response to selected child behaviours (e.g.,Bugental, 1991).The findings from this study suggest that mothers' self-efficacy beliefs were a central determinant of anticipatedbehavioural responses. Given the known links between parentbehaviours and the development of child behaviour problems (e.g.,Patterson, 1982), results from this study suggest that in helpingclinicians assist mothers in taking an active role in coping withchild noncompliance, it may be beneficial to identify ways inwhich these self-efficacy beliefs can be modified. For instance,parents' self confidence surrounding their ability to controltheir children may be increased by learning ways to spendenjoyable time with their children in which they are "armed" withpredetermined strategies to deal with problematic childbehaviours (e.g., Pisterman et al., 1992). Finally, therelationships between parent responses and perspective taking,empathic concern, and investment in parenting also providedfruitful avenues for future research. In particular, therelationship between perspective taking and I/C attributionssuggests that attributions may mediate the positive impact that84empathy has on parent and child outcome variables. The negativerelationship between parenting investment and I/C attributionsalso suggests that parents' use of a positive bias inattributions for their own children may be related to the extentto which they are invested in their role as parents. However,the exploratory manner in which these variables were investigatedin this study precludes any conclusions surrounding the strengthor the directionality of these relationships, and future researchis clearly warranted.85ReferencesAbramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978).Learned Helplessness in Humans: Critique and Reformulation.Journal of Abnormal Psychology, 87, 49-74.Alexander, S. W., Waldron, H. B., Barton, C. & Mas, C. H. (1989).The minimizing of blaming attributions and behaviours indelinquent families. Journal of Consulting and Clinical Psychology, 57(1) 19-24.Anderson, K. E., Lytton, H., & Romney, D. M. (1986). Mothers'interactions with normal and conduct-disordered boys: Whoaffects whom? Developmental Psychology, 22(5), 604-609.Azar, S. T. (1989). Unrealistic expectations and attributions ofnegative intent among teenage mothers at-risk for childmaltreatment: The validity of a cognitive view of parenting. Paper presented at the annual meeting for theadvancement of behaviour therapy, Washington, D.C.Batson, C. D. . (1987). Prosocial motivation: Is it ever trulyaltruistic? Advances in experimental social psychology(Vol. 20, pp. 65-122) New York: Academic Press.Bauer, W. D., & Twentyman, C. T. (1985). Abusing, neglectful,and comparison mothers' responses to child-related and non-child-related stressors. Journal of Consulting and ClinicalPsychology, 53(3), 335-343.Berley, R. A., & Jacobsen, N. S. (1984). Causal attributions inintimate relationships: Toward a model of cognitive-behavioural marital therapy. In P. Kendall (Ed.), Advancesin cognitive-behavioural research and therapy (Vol. 3, pp.1-90). New York; Academic Press.Billings, A. G., & Moos, R. H. (1983). Comparisons of childrenof depressed and nondepressed parents: A social-environmental perspective. Journal of Abnormal ChildPsychology, 11(4), 463-486.Blaney, P. H. (1986). Affect and memory: A review. PsychologicalBulletin, 99, 229-246.Bradbury, T. N., & Fincham, F. D., (1989). Cognition and maritaldysfunction: The role of efficacy expectations. Paperpresented at the 23rd Annual Convention of the Associationfor the Advancement of Behaviour Therapy, Washington, D.C.Brunk, M. A., & Henggeler, S. W. (1984). Child influences onadult controls: An experimental study. Developmental Psychology, 20, 1074-1081.86Bugental, D. B. (1989). Caregiver cognitions as moderators of affect in abusive families. Paper presented at the Meetingof the Society for Research in Child Development, KansasCity.Bugental, D. (1991). Adult processing deficits in stressful caregiving interactions. Presented at the Society forResearch in Child Development, Seattle.Bugental, D. B., Blue, J., & Lewis, J. (1990). Caregiver beliefsand dysphoric affect directed to difficult children.Developmental Psychology, 26(4), 631-638.Bugental, D. B., and Shennum, W. A. (1984). "Difficult" childrenas elicitors and targets of adult communication patterns:An attributional-transactional analysis. Monographs of theSociety for Research and Development, 49 (Serial No. 205,1 )•Cunningham, S. E.., Benness, B. B., & Siegal, L. S. (1988).Family Functioning, time allocation, and parental depressionin the families of normal and ADD-H children. Journal ofClinical and Consulting Psychology, 17(2), 169-177.Cohen, J. (1988). Statistical power analysis for the behavioural sciences. (Second Edition). New York: Academic Press.Cross Calvert, S., & Johnston, C. (1990). Acceptability oftreatments for child behaviour problems: Issues andimplications for future research. Journal of Clinical ChildPsychology, 19, 61-74.Davis, M. H. (1980). A multidimensional approach to individualdifferences in empathy. JSAS Catalogue of selecteddocuments in Psychology, 10.Davis, M. H. (1983a). Measuring Individual differences in empathy: Evidence for a multidimensional approach. Journalof Personality and Social Psychology, 44(1).Davis, M. H. (1983b). The effects of dispositional empathy onemotional reactions and helping: A multidimensionalapproach. Journal of Personality, 51(2), 167-184.Dix, T. (1989). Affective processes in parent cognition. Paperpresented at the biennial meeting of the Society forResearch in Child Development, Kansas City.Dix, T. (1991). The affective organization of parenting:Adaptive and maladaptive processes. Psychological Bulletin,110(1), 3-25.Dix T., & Grusec, J. E. (1985). Parent attribution processes inthe socialization of children. In I.E. Sigel (Ed.),87Parental beliefs systems (pp. 201-234). Hillsdale, N.J.:Erlbaum.Dix, T., & Lochman, J. (1990). Social cognition and negativereactions to children: A comparison of mothers ofaggressive and nonaggressive boys. Journal of Social andClinical Psychology, 9, 418-438.Dix, T., & Reinhold, D. (1987). Mood and context in parents' reactions to childrens' disobedience. Paper presented atthe meeting of the Society for Research and Development,Baltimore.Dix, T., Ruble, D. N., & Zambarano, R. J. (1989). Mothers'implicit theories of discipline: Child effects, parenteffects, and the attribution process. Child Development,60, 1373-1391.Egeland, E., & Sroufe, A. (1981). Developmental sequelae ofmaltreatment in infancy, New Directions for ChildDevelopment: Developmental Perspectives on Child Maltreatment. 77-92.Eisenberg, N. & Miller, P. A. (1987). The relation of empathy toaggressive and externalizing/antisocial behaviour.Psychological Bulletin, 103(3), 324-344.Eisenberg, N. & Strayer, J. (Eds.). (1987). Empathy and its development. New York: Cambridge University Press.Feshbach, N. D. (1978). Studies of empathic behaviour inchildren. In B. A. Mahler (Ed.), Progress in experimentalpersonality research, 8. New York: Academic Press.Feshbach, N. D. (1987). The construct of empathy and thephenomenon of physical maltreatment of children. InCicchetti, D. & Carlson, V. (Eds.), Child Maltreatment (pp.349-373). Cambridge University Press: Cambridge.Fincham, F. D. & Emery, (1988). Limited mental capacities andperceived control in attribution of responsibility. BritishJournal of Social Psychology, 27, 193-207.Fincham, F. D. & Bradbury, T. N. (1987). Cognitive processes andconflict in close relationships: An attribution-efficacymodel. Journal of Personality and Social Psychology, 53(6),1106-1118.Fiske, S. T., & Taylor, S. E. (1991). Social Cognition. (2nded.). New York: McGraw-Hill.Geller, J., Johnston, C., & Gabille, T. (1991). Mothers' negative mood and attributions for their own and their children's experiences. Paper presented at the meeting of88the association for the advancement of behaviour therapy,New York.Goldstein, A. P. & Michaels, G. Y. Empathy: Development, Training, and Consequences. New Jersey: Erlbaum.Gretarsson, S. J., and Gelfand, D. M. (1988). Mothers'attributions regarding their children's social behaviour andpersonality characteristics. Developmental psychology,24(2), 264-269.Griest, D. L., Wells, K. C., & Forehand, R. (1982). Anexamination of maternal perceptions of maladjustment inclinic-referred children. Journal of Abnormal Psychology,88, 277-281.Heider, F. (1958). The Psychology of Interpersonal Relations.New York: Wiley.Himmelstein, S., Graham, S., & Weiner, B. (1991). Anattributional analysis of maternal beliefs about theimportance of child-rearing practices. Child Development,62, 301-310.Hoffman, M. L. (1977). Empathy, its development and prosocialimplications. In C. B. Keasey (Ed.), Nebraska symposium onmotivation, 25. Lincoln: University of Nebraska Press.Holden, G. W. (1989). Parental selection of responses tomisbehaviour: The case of physical punishment. Symposiumconducted at the biennial meeting of the Society forResearch in Child Development, Kansas City.Howell, D. C. (1987). Statistical methods in psychology.Boston: PWS Publishers.Johnston, C. & Patenaude, R. L. (1991). Attributions for hvIperactive and aggressive child behaviours in parents of children with attention deficit hyperactivity disorder.Manuscript submitted for publication.Jones, E. E. & Davis, K. E. (1965). From acts to dispositions:The attribution process in person perception. In L.Berkowitz (Ed.), Advances in Experimental Social Psychology(Vol. 2, pp. 219-266). New York: Academic Press.Kelley, H. H. (1967). Attribution theory on social Psychology.In D. Levine (Ed.), Nebraska Symposium on Motivation.Lincoln: University of Nebraska Press, 15, 192-238.Kochanska, G. (1990). Maternal beliefs as long-term predictorsof mother-child interaction and report. Child Development,61, 1934-1943.89Krech, K. & Johnston, C. (1991). The relationship of depressedmood and life stress to maternal perceptions of child behaviour. Manuscript submitted for publication.Larrance, D. T. & Twentyman, C. T. (1983). Maternal attributionsand child abuse. Journal of Abnormal Psychology, 92(4),449-457.Lytton, H. (1990). Child and parent effects in boys' conductdisorder: A reinterpretation. Developmental Psychology,26(5), 683-697.Maccoby, E. E. & Martin, J. A. (1983). Socialization in thecontext of the family: Parent-child interaction. In E. M.Hetherington (Ed.), P. H. Mussen (Series Ed.), Handbook of child psychology: Vol. 4. Socialization, personality, andsocial development (pp. 1-101). New York: Wiley.Mackinnon, C. E. (1989). The relation between mother/sonattributions and the coerciveness of their interaction.Paper presented at the biennial meeting of the Society forResearch in Child Development, Kansas City.Miller, D. T. & Ross, M. (1975). Self-serving biases in theattribution of causality: Fact or fiction? PsychologicalBulletin, 82(2), 213-225.Murphey, D. A., & Alexander, S. W. (1991). Parents' beliefs andparental behaviour: A multi method study. Paper presentedat the biennial meeting of the Society for Research in ChildDevelopment, Seattle, Washington.Nisbett, R. E. & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ:Prentice-Hall.Offord, D., Rae-Grant, T. L., Links, G. S., Cadman, R. W., Byles,S. R., Crawford, D. M., Monroe-Blum, C. R., Byrne, E. D.,Thomas, W. S., Woodward, T. R., Boyle, I. L., & Szatmari, P.(1987). Ontario Child Health Study: II. 6-month prevalenceof disorder and rates of service utilization. Archives of General Psychiatry, 44, 832-836.Patterson, G. R. (1982). Coercive Family Process. Eugene,Oregon: Castalia.Petit, G. S., Dodge, K.A., & Brown, M. M. (1988). Early familyexperience, social problem solving patterns and children'ssocial competence. Child Development, 59, 107-120.Pisterman, S., Firestone, P., Mcgrath, P., Goodman, J. T.,Webster, I., Mallory, R., & Goffin, B. (1992). CanadianJournal of Behavioural Science, 24, 41-58.90Polzien, P. A., & Abidin, R.R. (1991) The parenting role scale: A measure of parenting commitment. Poster presented at thebiennial meeting of the Society for Research in ChildDevelopment, Seattle, Washington.Pulkinen, L. (1982). Self-control and continuity from childhoodto late adolescence. In Baltes, P. B., & Brim, 0. G. (Eds.)Life Span Development and Behaviour (Vol. 4, pp. 63-105),New York: Academic Press.Reinhold, D. & Lochman, J. E. (1991). Attributional and affectlabeling biases in mothers of aggressive sons. Paperpresented at the biennial meeting of the Society forResearch in Child Development, Seattle, Washington.Rogers, C. R. & Truax, C. B. (1967). The therapeutic conditionsantecedent to change: A theoretical view. In Rogers, C. R.(Ed.) The therapeutic relationship and its impact: A studyof psychotherapy with schizophrenics. Madison: Universityof Wisconsin Press.Rosenberg, M. S. & Reppucci, N. D. (1983). Abusive mothers:Perceptions of their own and their children's behaviour.Journal of Consulting and Clinical Psychology_, 51(5), 674-682.Seligman, M. E., Abramson, L. Y., Semmel, A., & Von Baeyer, C.(1979). Depressive attributional style. Journal of Abnormal Psychology, 88(3), 242-247.Sigel, I. E. (Ed.). (1985). Parental Belief Systems. Hillsdale,NJ: Erlbaum.Sobol, M. P., Ashbourne, D. T., Earn, B. M., & Cunningham, C. E.(1989). Parents' attributions for achieving compliance fromattention-deficit disordered children. Journal of Abnormal Child Psychology, 17(3).Weiner, B. (1979). A theory of motivation for some classroomexperiences. Journal of Educational Psychology, 71(1), 3-25.Weiner, B. (1980). A cognitive (attribution)-emotion-actionmodel of motivated behaviour: An analysis of judgments ofhelp-giving. Journal of Personality and Social Psychology,39(2), 186-200.91Table 1Factor Loadings of Globality/Stability andInternality/Controllability Factors Attribution RatingsFactorsGlobality/Stability Internality/ControllabilityInternality -.05 .56Externality .13 .11Controllability -.11 .60Stability .53 -.11Globality .53 -.0892Table 2Pearson Correlations Between Coded Attribution Responses andAttribution Ratings for ChildChild Attribution RatingsaCoded AttEibutionResponses G/Sc I/CdInternal/Blaming .28** .23*Internal/Nonblaming -.14 -.03External -.11 -.20*aHigher ratings indicate greater endorsement of the attributions.bHigher scores indicate endorsement of the attribution in moresituations.cG/S = Globality/StabilitydI/C = Internality/Controllabilty.* p < .05. ** p < .01.93Table 3Pearson Correlations Between Coded Behavioural Strategies andAnticipated Behavioural RatingCoded BehaviouralStrategiesAnticipated Behavioural RatingReprimand .23*Shift of Focus -.13Inactive Response -.37***Punishment .13Talk with Child .11* p < .05. *** p < .001.94Table 4Mean Attribution Ratings and Coded Attribution Responses as aFunction of Child Type and Behavioural ImpactChildOwn OtherM SD M SDChild Attribution RatingsaInconvenience^(n=46)Globality/Stability 8.82^3.10 10.33 2.56Internality 12.50 2.00 12.35 2.18No Inconvenience (n=52)Globality/Stability 7.97 3.67 9.40 2.95Internality/Controllability 12.20 2.64 12.07 2.69Coded Attribution ResponsesbInconvenience^(n=46)Internal/Blaming 1.83 1.27 1.54 1.41Internal/Nonblaming 2.74 1.29 1.85 1.43External .39 .98 1.54 1.64No Inconvenience (n=54)Internal/Blaming 1.63 1.28 1.52 1.33InternalNonblaming 3.06 1.41 1.76 1.47External .22 .74 1.26 1.50aHigher ratings indicate greater endorsement of the attributions.Maximum score = 18.00. Minimum score = 0.00.bHigher scores indicate endorsement of the attribution in moresituations. Maximum score = 5.00. Minimumscore = 0.00.95Table 5Mean Mother Attribution Ratings as a Functon of Child Type andBehavioural ImpactChildOwn^ OtherM SD M SDInconvenience (n = 46)Internality 3.02 1.47 1.92 1.52Controllability 4.70 1.69 2.66 1.61No Inconvenience (n = 52)Internality 2.68 1.57 1.91 2.02Controllability 5.26 1.62 1.95 1.49Note. Higher ratings indicate greater endorsement of theattribution.Maximum score = 9.00. Minimum score = 0.00.96Table 6Mean Anticipated Response Ratings and Coded Anticipated Behaviouras a Functon of Child Type and Behavioural ImpactChildOwn^ OtherSD M SDAnticipated Affect and Behaviour Ratings aInconvenience (n = 46)Affect^ 5.23^1.43^5.21Behaviour 6.15^1.74^4.87No InconvenienceAffect^ 5.06^1.63^3.77Behaviour 6.94^1.72^3.001.711.921.862.01Coded Behavioural ResponsesbInconvenience (n = 46)Punishment 2.20 1.24 1.04 1.05Inactive Response .13 .45 .50 .94Shift of Focus .20 .50 .57 .69Talk 1.57 1.31 1.54 1.26Reprimand .46 .75 .57 .98No Inconvenience (n = 54)Punishment 2.50 1.42 .93 .95Inactive Response .13 .39 1.15 1.20Shift of Focus .39 .71 .52 .69Talk 1.22 1.11 .96 .99Reprimand .35 .85 .17 .42aHigher ratings indicate greater anticipated affective andbehavioural responses. Maximum score = 9.00. Minimum score =0.00.bHigher scores indicate endorsement of the behavioural strategyin more situations. Maximum score = 5.00. Minimum score = 0.00.97Table 7Regression Analyses for Anticipated Affect and Behavioural ResponsesVariableAffect Behaviourr b^R2incremental r b^R2incrementalChild Age -.04 -.08 -.08 -.08SituationInconvenience -.28 -.25** .07* -.19 -.17 .03G/S .27 .10 .20 .12I/C .27 .28** .10** .16 .15 .05Mother Internality .31 .28** .22 .11MotherControllability .22 .12 .11*** .29 .25* .10**Total R2 = .29***^Total R2 = .18**G/S = Child Globality/Stability; I/C = ChildInternality/Controllabilty.* p < .05. ** p < .01. *** p < .001.98Table 8Correlations Among Anticipated Affect Ratings, Coded Attributions forChild Behaviour, and Coded Anticipated Behavioural Reponses AnticipatedAffectRatingCoded AttributionsInternal/Blaming Internal/Nonblaming ExternalBehaviouralStrategyTalking -.06 -.05 .17* -.10Punishment .11 .09 -.01 -.02Inactive -.18* -.18* .10 .05ResponseShift of -.07 -.18* .12 .15FocusReprimand .28** .33*** -.13 -.21*AttributionsInternal/Blaming .29**Internal/Nonblaming .25**External -.02* p < .05. ** p < .01. *** p < .001.99Table 9Correlations Between Mothers' Attributions and Anticipated Affective andBehavioural Responses with Empathy and Parenting Investment Subscales Empathy Parenting RolePerqpectiveTaking EmpathicConcern ParentingInvestmentG/S -.09 -.05 .12I/C -.23* .05 -.17*Mother Internality -.07 .16 .14Mother Controllability .01 .23* .03Affective Rating -.30** .03 .22*Behavioural Rating .05 -.17* .18*Note. Higher values indicate higher scores on the measures.* p < .05. ** p < .01.Appendix ADepartment of PsychologyUniversity of British ColumbiaJosie Geller 689-9974Dr. Charlotte Johnston 822-6771This questionnaire is part of a research project being conductedby graduate student Josie Geller and Dr. Charlotte Johnston inthe Psychology Department at the University of British Columbia.We are interested in how parents explain things their ownchildren do and things that other children of the same age andgender do. Findings from this project will better ourunderstanding of how parents think about children's behaviour andhow this relates to particular parenting behaviour.We recommend that you choose a quiet time to complete the forms(e.g. after your child(ren)'s bedtime). There are four sectionsto the questionnaire. The first asks you to provide some generalinformation about yourself and your child. The second sectionasks you to respond to eight descriptions of situations involvingyourself and a child. For half of the descriptions, we wouldlike you to imagine that the child in the situations is^ , and in the other half a child you don't knowbut who is the same age and gender as ^ . Afterreading each description you are asked to indicate why thebehaviour occurred and how you would react. The third section ofthe questionnaire consists of a series of questionnaires aboutyourself, and finally, the last section asks you to describe^ . More detailed directions will be provided atthe beginning of each section.We would appreciate it if you could complete the questionnairesand return the packet as soon as possible. The questionnairesshould take about 45 minutes to complete.You are free to withdraw from the study at any time. If thecompleted questionnaires are returned to UBC, then it will beassumed that you have consented to participate. Your responsesto the questions will be treated as confidential. The externalenvelope we receive from you will be discarded and your packagewill be identified by subject number only. Your participation inthe project is strictly voluntary.Thank you for expressing an interest in this study. If you haveany comments or questions regarding the project, please callJosie Geller at 689-9974.100Appendix B^ 1011. Mother's Age: ^2. Mother's Occupation:^3. Highest level of education received for mother:4. Mother's ethnic background: ^5. Father's Occupation:^6. Highest level of education received for father:^7. Please check the situation that best describes your family:a) Two parent familyb) One parent familyc) Step-familyd) Other. Please describe^8. Please provide the following information about each of yourchildren:CHILD^GENDER^ARE YOU THE CHILD'SAGE (M/F):^BIOLOGICAL MOTHER?9a. Have you ever sought psychological help for personal problems? YN9b. If yes, please describe briefly:10a. Have you ever sought psychological help for child behaviourproblems?Y N10b. If yes, please identify which child and describe briefly:Appendix C^ 102We would like you to read a series of scenarios describing problemchild behaviours and answer questions about each of them. For eachsituation involving your own child, please try to imagine that thesituation happened to ^ . For situations that do notinvolve your own child, please imagine that the situation happened with"Johnny," a child you have never met but who is the same age and genderas ^ . Try to imagine each situation as an entirelyseparate event, and not to let your ratings for one situation influenceratings on subsequent situations.For each situation, you will first be asked to describe what you thinkcaused the child's behaviour. Then, you will be asked to answerspecific questions about the cause of the behaviour. In seekingexplanations for the behaviour of others, people often make judgementsalong the following dimensions:1. The extent to which the behaviour is caused by somethingabout the child.2. The extent to which the child has control over the behaviour3. The extent to which the behaviour is caused by factors whichare present every time the situation arises.4. The extent to which the behaviour is caused by factors thatwill be present in the future.In addition, you will be asked to rate how intense your response wouldbe to each situation, how likely you would be to respond to eachsituation, and what you would do, if anything, if the situationactually happened.You may find it difficult to make these ratings. Remember, there areno right or wrong answers, just go with your first impression.••••••••••••••••••••••••••••••"••.............••••••••••..;••••••••••••••Please try to vividly imagine yourself in the situations that follow.Although some of the situations may not be typical of your experienceswith your child or other children, please stretch your imagination andtry to put yourself in each situation. For each situation, describewhy you think the child failed to do as he/she was told, and answer thequestions about what you think would have caused the child to behave inthat way. Finally, rate your reaction to the situation, and describewhat you would do in each circumstance..Aprendix DYOU are riding on the city bus, and ^ is sitting neat to you. ^ is holding a CARof Pop and begins to PlaY with the metal ring that ovens the tan. You ask his not to 421504 it. Atioute later you hear the •pftff . sound -of the can being opened and you notice that the pop hag(1)^khat do you think is the main reason the child did not do as you asked?(2) To what extent was the child's behaviour caused by something about the child?0^1^2^3^4^5^6^7^8^9totally due Not at allto the child due to the child(3) To what extent was the cause of the child's behaviour something about theexternal circumstances or the situation?0^1^2^3^4^5^6^7^8^9Totally due to Not at all due toexternal circumstances external circumstances(4) To what extent was the child's behaviour something that was within hiscontrol?0^1^2^3^4^5^6^7^8^9Not at all under Completely underchild's control child's control(5) To what extent was the cause of the child's behaviour something that islasting vs. a one-time thing?0^1^2^3^4^5^6^7^8^9Lasting A one-time thing(6)^To what extent was the cause of the child's behaviour something that willinfluence his behaviour in other circumstances?0^1^2^3^4^5^6^7^8^9Influences just this Influences mostparticular situation situations(7) To what extent would you be upset by the child's behaviour?0^1^2^3^4^5^6^7^8^9Not at all upset Extremely upset(8) To what extent was the cause of the child's behaviour something aboutyourself?0^1^2^3^4^5^6^7^8^9Totally due to me Not at all due to me(9) To what extent was the child's behaviour something that was within your control ?0^1^2^3^4^5^6^7^8^9Not at all Completelyunder my control under my control(10) If this situaaboutttion heactualchil ly s happened, to what extent would you be likely to dosomethingd' behmilour?0^1^2^3^4^5^6^7^8^9Not at all Extremelylikely likely(11)^If this situation actually happened, what, if anything, would you do?Appendix E^ 104The next set of questions ask about your thoughts surrounding children.Please read each statement carefully and decide whether you agree ordisagree with it. Indicate how much you agree or disagree with eachstatement according to the following scale:1 1 1 1 11 2 3 4 5Strongly Agree Neutral Disagree StronglyAgree Slightly Slightly Disagree^ 1. I sometimes find it difficult to see things from a child'spoint of view.2. I try to look at children's side of a disagreement beforemaking a decision.3. I sometimes try to understand children better by imagininghow things look from their point of view.^ 4. If I'm sure I'm right about something involving my children,I don't waste much time listening to children's arguments.5. I believe that there are two sides to every question and tryto look at them both.^ 6. When I'm upset with a child, I usually "put myself in his orher shoes" for a while.^ 7. Before criticizing a child, I try to imagine how I would feelif I were in his or her place.^ 8. I often have tender, concerned feelings for people lessfortunate than me.^ 9. In emergency situations, I feel worried and ill-at-ease.10. When I see someone being taken advantage of, I feel kind ofprotective toward them.^ 11. Other people's misfortunes do not usually bother me a lot.^ 12. When I see someone being treated unfairly, I sometimes don'tfeel much pity for them.^ 13. I am often very touched by things I see happen.^ 14. I would describe myself as a pretty soft-hearted person.Appendix F^ 105The next set of questionnaires ask about your feelings and thoughtssurrounding parenting. Please read each statement carefully and decidewhether you agree or disagree with it. Indicate how much you agree ordisagree with each statement according to the following scale:1^2^3^4^5Strongly^Agree^Neutral^Disagree^StronglyAgree Slightly Slightly Disagree^ 1. My children are my greatest source of pride.2. It is very important to me that my children feel close to me.^ 3. When I do something which makes my children happy, I amhappy.^ 4. Being a parent makes me feel important.^ 5. No matter where I am or what I am doing, I think of mychildren at least every couple of hours.^ 6. My children will accomplish more than I did.^ 7. Being a parent is something I want to be good at.^ 8. Being a parent is my greatest responsibility.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items