Journal of Psychopathology and Behavioral Asssment,Vol.26,No.1,March2004(C 2004)
The Moral Dingagement Scale:Extension With
an American Minority Sample
Jennifer Pelton,1Mary Gound,1Rex Forehand,2,3and Gene Brody1
Accepted July29,2003
This study extended an investigation by A.Bandura,C.Barbaranelli,G.V.Caprara,and C.Pastorelli
(1996)by examining the structure,reliability,and correlates of the Moral Dingagement Scale
(MDS),which measures dingagement from moral lf-sanctions.Furthermore,the role of moral
dingagement in the association between parenting and child behavior was examined.Two hun-
dred and forty-five African American single-parent mothers and one of their children(mean age=佛山教育网
11.4years)participated in the study.The MDS had similar factor structure,internal consistency,and
disc
demographic correlates as tho reported by A.Bandura et al.(1996)with an Italian sample of children.
The MDS also correlated with delinquent and aggressive child behavior,and moral dingagement
partially mediated the relationship between parenting and child delinquent behavior.
KEY WORDS:moral dingagement;child disruptive behavior;Parenting.
Cimbora and McIntosh(2003)recently stated that “ay be key to understanding delinquent be-havior”(p.296).Although morality has been a topic of rearch interest for decades,recent work by Bandura (e.g.,1991,1999)has elucidated this construct.He ar-gues that moral dingagement is the process by which a person can justify his or her harmful or aggressive behavior,or,in other words,dingage from the moral lf-sanctions that typically rve to regulate behavior. Bandura and his colleagues examined this concept by de-veloping the Moral Dingagement Scale(MDS;Bandura, Barbaranelli,Caprara,&Pastorelli,1996)and then eval-uating whether moral dingagement is associated with transgressive behavior.Although they found that the con-struct is indeed related to aggressive and delinquent behav-ior of school-age children(Bandura et al.,1996;Bandura, Caprara,Barbaranelli,Pastorelli,&Regalia,2001),the relationships have been exa
mined in only one sample drawn from children living in Italy.The purpos of this study were to extend Bandura et al.’s study to a demo-1Institute for Behavioral Rearch,University of Georgia,Georgia.
2Psychology Department,University of Vermont,Burlington,Vermont. 3To whom correspondence should be addresd at Department of Psy-chology,University of Vermont,230Dewey Hall,2Colchester Avenue, Burlington,Vermont05405-0134;e-mail:rex.forehand@uvm.edu.
graphically different sample of American , African Americans)and to examine the role of moral dingagement in the relationship between parenting and child behaviors(Bandura,1996).
First,however,the construct of moral dingagement derves further explication.Bandura(1999)described how moral standards are developed through socialization and lf-regulation:A person monitors his or her conduct and the conditions under which it occurs,then judges it in relation to his or her moral standards.However,there can be considerable variability in how one’s moral standards operate,as well as in one’s lf-regulatory capabilities.
Specifically,Bandura argued that lf-regulatory mecha-nisms only come into play when activated,and that there are many social and psychological mechanisms by which moral lf-sanctio
ns can be dingaged from transgressive conduct.
Bandura(1999)described eight mechanisms by which moral dingagement occurs.The mechanisms are Moral Justification,Euphemistic Language,Advan-tageous Comparison,Displacement of Responsibility, Diffusion of Responsibility,Distorting Conquences, Attribution of Blame,and Dehumanization.For example, moral justification is a process in which detrimental conduct is made acceptable by portraying it as rving socially worthy or moral purpos.Advantageous 31
0882-2689/04/0300-0031/0C 2004Plenum Publishing Corporation
32Pelton,Gound,Forehand,and Brody
comparison involves contrasting one’s own behavior to behavior that is more reprehensible to exonerate one’s own conduct that is then en as benevolent.Displacement and diffusion of responsibility are two other mechanisms in which personal responsibility is minimized due to social pressures and the dictates of others(in the ca of displacement)or the actions of a group that was engaging in the same behavior(diffusion of responsibility).
Although the MDS taps into each of the areas that Bandura has delineated,factor analysis in the two stud-ies by Bandura and his associates(1996,2001)revealed a one-factor structure,accounting for16%of the variance, with all items loading on the principal factor.Reliability analys(internal consistency)for both studies were high (alphas of.82and.86,respectively).Gender differences in moral dingagement emerged in both studies such that boys demonstrated higher levels of moral dingage-ment than girls,but moral dingagement was unrelated to socioeconomic status of the family or age of the child. Finally,for both elementary and middle school children in Bandura et al.’s(1996)Italian sample,child,teacher, and parent ratings of aggression,social competence,and in some cas delinquency correlated with scores on the MDS(Bandura,1996).However,as this scale has only been examined with an Italian sample of children,the question remains whether the MDS would result in simi-larfi,one-factor structure,internal reliability, gender differences)in a U.S.sample with considerably different demographic characteristics.
This study extended parts of Bandura et al.’sfirst study(1996)with a lect sample:African American chil-dren from low income,single-parent houholds residing in rural and urban ttings.This sample was utilized be-cau it differed substantially from the middle income, presumably two-parent,Italian sample in the study by Bandura and his associates.As a result,similarfindings to tho of Bandura et
al.(1996)would suggest that the MDS potentially has utility across children who vary on a number of sociodemographic dimensions.
英语教学文章The purpos of this study were fourfold.The first three purpos were extensions of key aspects of the Bandura et al.(1996)study:(1)Examine the fac-tor structure of the MDS with low income,single-parent,African American children;(2)examine demo-graphic correlates(child gender,child age,family SES) to ascertain the applicability of the MDS to a de-mographically different sample;and(3)examine child behavioral correlates of the MDS.It was hypothe-sized that,similar to Bandura et al.(1996),one fac-tor with acceptable internal consistency would emerge, boys would have higher MDS scores than girls,child age and family SES would not be associated with MDS scores,and child behaviors would correlate with MDS scores.
Thefinal purpo was to examine whether scores on MDS help explain the relationship between parenting and child aggressive and delinquent behaviors.It is im-portant to note that Bandura and his colleagues(1996, 2001)did not examine this question but rather were con-cerned with the internal ,anticipatory guilt) by which moral dingagement influences aggressive and delinquent behavior.The current study did not examine the internal process but instead examined the role of moral dingagement in the link between parenting and child behavior.Previous
rearch has revealed a relation-ship between positive parenting,which provides warmth and structure,and lower levels of aggressive and delin-quent behavior in children(e Kotchick&Forehand, 2002).Moral dingagement may be a mechanism that helps explain the association between parenting and child aggressive and delinquent behavior.For example,positive parenting may enhance moral engagement by providing clear limits and predictable conquences with develop-mentally appropriate discussion regarding the moral im-plications of the transgression or both,thereby reducing aggressive and delinquent behavior.In addition,positive parenting may reduce moral dingagement by increasing the child’s n of personal accountability for his or her actions.Thus,it was hypothesized that moral dingage-ment would help explain the link between parenting and child behavior.
METHOD
Participants
A community sample of245African American fam-ilies headed by single mothers(mean age=33.92years) participated.One child from each family(mean age= 11.40years;124boys and121girls)participated with her or his mother.The sample that participated in this study was initially
recruited3years earlier when the tar-get child was6–11-years old.The primary purpo of the project when it was initiated was to study family function-ing in low-income,African American,single-parent fami-lies.During the fourth asssment,the MDS was added to the battery of instruments administered as this measure ap-peared to hold some promi in helping to understand the association between parenting and child behavior.How-ever,as has been noted,the instrument had not been uti-lized with samples similar to the current one.
In terms of recruitment,if there were two or more children eligible to participate in a family,with a few
The Moral Dingagement Scale33
exceptions,the oldest child was lected.In the few excep-tions,the participating child was randomly lected.Fam-ilies were drawn from metropolitan and nonmetropolitan counties in the southeastern United States and were re-cruited through community ,schools)and leaders.Initially,each community contact was given the inclusion ,African American).Subquently, community contacts gave rearch staff members the names of families to approach,and staff members con-tacted the families.finish什么意思
Of the participating families,97%received public assistance.Almost all of the families had a per capita
l income of$3,800or less.According to the criteria estab-lished by the Census Bureau(U.S.Census Bureau,1992), thisfigure placed families in thefirst quintile for hou-hold income,which the Bureau defines as poverty status. In the counties from which the sample was drawn,75% of single African American mothers with school age chil-dren live in poverty(U.S.Bureau of Census,1996).Of the mothers in the sample,42%had less than a high school education,38%had received a high school diploma,and 20%had some college or trade school education.Forty-three percent of mothers were employed for at least32-hr per week,19%were employed for fewer than32-hr,and 38%were unemployed.
Measures
The availability of instruments to measure the con-structs of interest was a particular concern becau most measures of family functioning and child adjustment were developed and standardized with Caucasian,middle-class families.Conquently,the concern was that the avail-able measures would not adequately capture the nature of family process among African Americans in this study. This issue was addresd through the formation of focus groups compod of African American community mem-bers in the counties from which the sample was drawn. The focus groups included a total of60people who were reprentative of the population we studied.
The focus groups discusd the relevance of con-structs propod for investigation,as well as the likelihood that the measure would elicit information relevant to the constructs.The groups reviewed each item on the scales and suggested wording changes,as well as the deletion of items that were unclear to them or irrelevant to families in their communities.When items or the respon scale for items were changed,or if the measure had not been uti-lized with similar samples,a confirmatory factor analysis was performed to ensure that each scale was compod of a coherent t of items for this study population.Items were retained if they loaded at.40or above.An alpha co-efficient for each measure was also computed.Measures were completed by the child,the mother,and the child’s teacher.
Demographic Information
Mothers completed a demographic measure that pro-vided information about themlves,their children,and their families.The measure assd basic demographic information such as age,education,income,and commu-nity ,rural vs.urban environment and number of ,“people are shot or knived in my neighborhood”).
Moral Dingagement Scale(MDS)
counton
The MDS is a32-item questionnaire designed to as-ss a child’s proneness to moral dingagement(Bandura, 1995).Items are rated on a three point scale(0=disagree, 1=not sure,and2=agree).The MDS asss eight mechanisms of moral dingagement:Moral justification, Euphemistic language,Advantageous comparison,Dis-placement of responsibility,Diffusion of responsibility, Distorting conquences,Attribution of blame,and Dehu-manization.However,as noted earlier,one factor emerged accounting for16%of the variance,with an alpha coeffi-cient of.82.In addition,correlations with child behavior in the expected ,higher MDS scores corre-late with aggression and delinquency)emerged(Bandura et al.,1996).
For the purpos of the prent project,the four items tapping Euphemistic language were eliminated due to their complexity and children’s lack of comprehension of the meaning of the items.For example,“To hit ob-noxious classmates is just giving them a‘lesson”’and “Slapping and shoving someone is just a way of joking”were removed becau the children with whom the mea-sure was pilot tested had difficulty comprehending the meaning of the euphemistic language.Additionally,mi-nor modifications were made to some items to make it more culturally nsitive and to increa comprehensi-bility with the prent sample.The modifications were made subquent to testing with veral pilot participants. Examples of modifications are“It is okay to treat someone badly if they
act like a‘worm’(original)”to“If someone acts like a jerk,it is okay to treat them badly(revid)”;“If kids are not disciplined,they should not be blamed for misbehaving(original)”to“Kids who are not taught to obey the rules should not be blamed for misbehaving
34Pelton,Gound,Forehand,and Brody
(revid).”In thefirst example,the word“worm,”which was not culturally meaningful to the current population, was changed to the word“jerk.”In the cond example, the word“disciplined”was not understood by the pilot participants and was changed to“not taught to obey the rules.”
Positive Parenting
Two aspects of parenting,monitoring and mother–child warmth/support,were assd by mother report.
The Monitoring and Control Questionnaire(MCQ) was ud to asss the extent to which a mother moni-tored child behavior.This17-item scale was developed for this study with help from some of our colleagues(e Kotchick et al.,1997).The MCQ is bad on monitor-ing measures ud by Patterson and Stouthamer-Loeber (1984).It asss parents’perceptions of their knowl-edge about various as
pects of their children’s lives.Items are rated on a4-point Likert scale ranging from1(never) to4(always).Sample items include,“How often do you know about where[target child]is and what s/he is do-ing when away from home?”,“How often do you know about[target child’s]u of alcohol?”,and“How often do you know what his/her grades are?”Scores can range from17to68,with higher scores indicating higher levels of maternal monitoring.For the prent sample,a factor analysis indicated that all17items loaded,and the alpha coefficient was.86.
The short form of the Interactive Behavior Question-naire(IBQ;Prinz,Foster,Kent,&O’Leary,1979)was ud to asss warmth and support in the mother–child re-lationship.This form consists of the20items that have the highest phi coefficients and the highest item-to-total corre-lations among the75items in the original IBQ.The short form correlates.96with the longer version.The items, which are endord as True or Fal,include“You enjoy spending time with your child,”and“You think you and your child get along well with each other.”Prinz et al. (1979)and Robin and Weiss(1980)reported adequate internal consistency and discriminant validity.A confir-matory factor analysis indicated that14of the20items loaded on a single construct at.40or above;accordingly, only the14items were included in the analys.The alpha coefficient for the14items was.85.Scores can range from0to14,with higher scores indicating more warmth and support in the mother–child relationship.
bowel
The two measures of parenting were correlated(r= .28,p<.01)at a low to moderate level.As the two mea-sures reprent two different dimensions of parenting,the magnitude of the correlation is not surprising and is not in-consistent with otherfindings(Conger et al.,1993).As the two constructs together constitute authoritative or positive ,Kotchick&Forehand,2002),the two mea-sures were each standardized and then averaged to form the positive parenting construct.
Child Psychosocial Adjustment
Three areas of child psychosocial adjustment were examined:social competence,delinquent behavior,and aggressive behavior.
Social Competence.The Perceived Competence Scale for Children(PCSC;Harter,1982),the Parent’s Rating Scale of Children’s Actual Competencies(PRS; Harter,1982),and the Teacher Rating Form(TRF;Harter, 1982)were administered to children,mothers,and teach-ers,respectively,to asss for child competence.
when i look into your eyesThe PCSC measures a child’s perceptions of his or her competence in veral domains,including social com-petence.The Social Competence subscale includes items such as,“You are popular with others of your age”and “Youfind it hard to make friends.”Harter reported ade-quate reliability and vali
dity data for the PCSC;however, the measure was not standardized with samples similar to the prent one.Therefore,a confirmatory factor anal-ysis was conducted on the Social Competence subscale, resulting infive of the ven items being retained.The al-pha coefficient for thefive items for the prent sample was.67.
lordiThe PRS,completed by mothers,asss parents’perceptions of their child’s competence in the same ar-eas as assd by the PCSC.Factor analysis of the PRS revealed a virtually identical factor pattern to the PCSC (Harter,1982).A factor analysis with the sample in the current study resulted infive of the ven items loading at .40or higher and being retained.Thefive items had an alpha coefficient of.63.
Teachers’perception of child competence was bad on the Teacher Rating Form(TRF).Like the PCSC and PRS,the TRF asss a child’s competence in veral areas,including social competence.Internal consistency of the subscales has been found to range from.93to.96 (Harter,1982).A factor analysis resulted in four of the ven items loading at.40or higher and being retained. The four items had an alpha coefficient of.67.
Aggressive and Delinquent Behavior.Two domains of child transgressive ,externalizing p
rob-lems),aggressive behavior and delinquency,were exam-ined using child report,parent report,and teacher report of the Child Behavior Checklist(CBCL;Achenbach,1991a, 1991b;Achenbach&Edelbrock,1986).
巴黎恐怖袭击
The Moral Dingagement Scale35
To examine transgressive behavior from the child’s perspective,the Youth Self-Report form(YSR)was ud. Children completed both the Aggressive Behavior and the Delinquent Behavior Subscales(Achenbach,1991a).This measure,lected becau it asss the types of exter-nalizing problems typically displayed by children in the age range included in this study,has acceptable reliability and validity data(Achenbach,1991a);however,it has not been standardized with children as young as some of tho included in this investigation.Thus,a factor analysis was conducted on each subscale.All19-items of the Aggres-sive subscale loaded at.40or greater and were retained. The alpha coefficient for this subscale with the current sample was.87.Seven of11items of the Delinquency subscale loaded at.40or greater and were retained.The alpha coefficient for this sample was.72.
Mothers completed the Child Behavior Checklist (CBCL;Achenbach,1991b),a113-item instrument tha
t consists of nine subscales.Achenbach(1991b)reported mean test–retest reliability of.87and evidence for con-tent and criterion-related validity with samples similar to the current one.Consistent with the YSR,the subscales of interest for this study were the Delinquent and Aggressive Behavior Subscales,which had alpha coefficients in the current sample of.91and.79,respectively.
Teachers completed the Teacher Report Form of the CBCL(CBCL-TRF).The CBCL-TRF is completed by teachers who have known a pupil in a school tting for at least2months.It is a113-item instrument modeled after the parent version of the CBCL.Achenbach and Edelbrock (1986)reported mean test–retest reliability of.86and ev-idence for content and criterion validity.The subscales of Delinquency and Aggressive Behavior,with alpha coeffi-cients of.75and.97,respectively,were examined in this study.
Procedure
Families were recruited through community leaders and ,schools).Each community contact gave the rearch staff member the names of families who ex-presd interest in participation,and the staff member con-tacted the families.Of the families contacted,67%agreed to participate.
Two data collection ssions,each of which lasted between1and2hr,were scheduled at each asssment. During thefirst ssion,the mother completed informed connt forms and granted permission to contact the child’s teacher.The mother and child subquently com-pleted an interview focusing on demographic information. In the cond ssion,the study ,moral dingagement,child adjustment)were assd.At both data collection ssions,lf-report questionnaires were administered in an interview format to the mother and child.Each interview was conducted privately between the mother or child and an interviewer,with no other fam-ily members prent or able to overhear the conversation. Families were paid$50for their participation in each data collection ssion.Data for parenting were collected at an asssment which occurred15months before the primary asssment in this study.
六级准考证丢了怎么办
For teacher report,a rearch assistant contacted the teachers individually,informed them of the purpo of the study,and then hand-delivered the appropriate instruments to the teachers.The teachers mailed the completed instru-ments to the rearch assistant.Teachers were paid$20. Data Analytic Strategies
The data analytic strategy consisted of four steps. First,similar to Bandura et al.(1996),a principal-components analysis with varimax orthogonal rotation was conducted on the MDS and an alpha relia
bility co-efficient for emerging factors was calculated.Second and third,again similar to Bandura et al.,demographic and behavioral correlates were examined.
Finally,to test whether moral dingagement helps explain the relationship between parenting and both child delinquency and child aggression,the procedures outlined by Baron and Kenny(1986)for testing medi-ation were utilized.To demonstrate mediation,veral conditions must be met.First,positive parenting must be significantly related to moral dingagement.Second, positive parenting must also be significantly related to delinquency/aggression.Third,moral dingagement must be significantly related to delinquency/aggression. If the three conditions are met,evidence for mediation is demonstrated if the association between parenting and delinquency/aggression is no longer significant after controlling for moral dingagement.Thus,two ts of regression analys were conducted with aggression rv-ing as the dependent variable in one t and delinquent behavior as the dependent variable in the cond t. Parenting rved as the primary independent variable, with moral dingagement,the mediator,also entered as an independent variable.
The mediational analys were conducted with par-enting data collected15months earlier than moral dis-engagement and aggression/delinquent behavior data.As parenting has been shown to be a relatively stable con-struct over a1-to2-year period(r values above.50and no change in mean level)in
samples similar to the current
