anak di pukul

DOI: 10.1542/peds.2009-2678 2010;125;e1057-e1065; originally published online Apr 12, 2010; Pediatrics

Catherine A. Taylor, Jennifer A. Manganello, Shawna J. Lee and Janet C. Rice Aggressive Behavior

Mothers' Spanking of 3-Year-Old Children and Subsequent Risk of Children's

http://www.pediatrics.org/cgi/content/full/125/5/e1057located on the World Wide Web at:

The online version of this article, along with updated information and services, is

rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275. Grove Village, Illinois, 60007. Copyright 2010 by the American Academy of Pediatrics. All and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elkpublication, it has been published continuously since 1948. PEDIATRICS is owned, published, PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

. Provided by Indonesia:AAP Sponsored on December 10, 2010 www.pediatrics.orgDownloaded from

Mothers Spanking of 3-Year-Old Children andSubsequent Risk of Childrens Aggressive Behavior

WHATS KNOWN ON THIS SUBJECT: Dozens of studies haveshown a signicant statistical link between the use of CP withchildren and child aggression, including many studies thatcontrolled for the childs initial level of aggression.

WHAT THIS STUDY ADDS: With controlling for the childs initiallevel of aggression, demographic features, and 8 potentialparenting risk confounders, which to our knowledge have notpreviously been controlled simultaneously, more-frequent use ofCP increased the risk for higher levels of child aggression.

abstractOBJECTIVE: The goal was to examine the association between the useof corporal punishment (CP) against 3-year-old children and subse-quent aggressive behavior among those children.

METHODS: Respondents (N 2461) participated in the Fragile Fami-lies and Child Well-being Study (19982005), a population-based, birthcohort study of children born in 20 large US cities. Maternal reports ofCP, childrens aggressive behaviors at 3 and 5 years of age, and a hostof key demographic features and potential confounding factors, includ-ing maternal child physical maltreatment, psychological maltreat-ment, and neglect, intimate partner aggression victimization, stress,depression, substance use, and consideration of abortion, wereassessed.

RESULTS: Frequent use of CP (ie, mothers use of spanking more thantwice in the previous month) when the child was 3 years of age wasassociated with increased risk for higher levels of child aggressionwhen the child was 5 years of age (adjusted odds ratio: 1.49 [95%condence interval: 1.21.8]; P .0001), even with controlling for thechilds level of aggression at age 3 and the aforementioned potentialconfounding factors and key demographic features.

CONCLUSIONS: Despite American Academy of Pediatrics recommen-dations to the contrary, most parents in the United States approve ofand have used CP as a form of child discipline. The current ndingssuggest that even minor forms of CP, such as spanking, increase riskfor increased child aggressive behavior. Importantly, these ndingscannot be attributed to possible confounding effects of a host of othermaternal parenting risk factors. Pediatrics 2010;125:e1057e1065

AUTHORS: Catherine A. Taylor, PhD, MSW, MPH,a JenniferA. Manganello, PhD, MPH,b Shawna J. Lee, PhD, MSW,MPP,c and Janet C. Rice, PhDd

Departments of aCommunity Health Sciences and dBiostatistics,School of Public Health and Tropical Medicine, Tulane University,New Orleans, Louisiana; bDepartment of Health Policy,Management, and Behavior, School of Public Health, Universityat Albany, State University of New York, Rensselaer, New York;and cMerrill-Palmer Skillman Institute for Child and FamilyDevelopment, School of Social Work, Wayne State University,Detroit, Michigan

KEY WORDScorporal punishment, child discipline, spank, childhoodaggression, child aggressive behavior

ABBREVIATIONSCPcorporal punishmentFFCWSFragile Families and Child Well-being Study

www.pediatrics.org/cgi/doi/10.1542/peds.2009-2678

doi:10.1542/peds.2009-2678

Accepted for publication Jan 7, 2010

Address correspondence to Catherine A. Taylor, PhD, MSW, MPH,Tulane University School of Public Health and Tropical Medicine,Department of Community Health Sciences, 1440 Canal St, Suite2301 TW19, New Orleans, LA 70112. E-mail: [email protected]

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).

Copyright 2010 by the American Academy of Pediatrics

FINANCIAL DISCLOSURE: The authors have indicated they haveno nancial relationships relevant to this article to disclose.

Funded by the National Institutes of Health (NIH)

ARTICLES

PEDIATRICS Volume 125, Number 5, May 2010 e1057. Provided by Indonesia:AAP Sponsored on December 10, 2010 www.pediatrics.orgDownloaded from

When parents discipline their children,they generally do so to teach their chil-dren a lesson, to instill values, and/orto improve their childrens currentand future behavior. Corporal punish-ment (CP) is one disciplinary strategythat remains highly prevalent in theUnited States despite controversy sur-rounding its use.1 Estimates of US par-ents who have used CP vary from 35%to 90%, depending on key modierssuch as the age and gender of the childand the type of punishment specied(eg, spanking or slapping).25 In a 2005US poll, 72% of adults reported that itwas OK to spank a child, with ap-proval ratings being highest in theSouth and lowest in the Northeast.6

The normativeness of CP in the UnitedStates stands in contrast to the recom-mendations of the American Academyof Pediatrics, which are consistentwith those of other professional orga-nizations,79 that parents be encour-aged and assisted in the developmentof methods other than spanking formanaging undesired behavior.(p723)10

Such concerns are rooted in the in-creasing body of empirical evidencesuggesting that the risks of using CPagainst children are likely to outweighthe potential benets. A 2002 meta-analysis showed linkagesbetweenCPofchildren and risk for poor outcomes inchildhood, including aggressive and/orantisocial behavior, mental health prob-lems, and physical maltreatment.11

Whether CP causes aggression is ofparticular relevance for public healthinterests in disrupting the cycle of vio-lence. The meta-analysis performed byGershoff11 found the link between CPand aggression to be positive (d 0.36); however, most of those studieswere not longitudinal.12 To assertmorestrongly that CP is a causal determi-nant of aggression, it is necessary todemonstrate a statistically signicantand temporally specic link betweenCP and aggression and also to control

for the childs initial level of aggres-sion and key potential confounders.11,12

Other researchers have aimed to meetmost of these conditions.1322 However,the current study accounts for all 4conditions, has a larger sample sizeand therefore more statistical powerthan all except 2 of the aforementionedstudies,21,22 and controls for key poten-tial maternal parenting risk confound-ers that previously have not been ex-amined simultaneously. The currentstudy was designed to answer the fol-lowing question: is a mothers use ofCP on a 3 year-old child linked to riskfor that child beingmore aggressive at5 years of age, even after controllingfor the childs initial level of aggres-sion at age 3 and other important ma-ternal parenting risk factors and de-mographic features?

METHODS

Participants

The sample was obtained from theFragile Families and Child Well-beingStudy (FFCWS), a population-based, co-hort study of families from 20 large UScities. The original sample (N 4898)was obtained from 1998 to 2000 bysampling births within hospitals in cit-ies with populations of 200 000 in1994; a detailed description of theFFCWS design was published previous-ly.23 Four waves of data are available,that is, baseline (time of the indexchilds birth) and index child ages of 1,3, and 5 years. Two interviews wereconducted when the child was 3 and 5years of age, a core interview and aninterview conducted with a subsampleof mothers for the add-on In-Home Lon-gitudinal Study of Preschool-Aged Chil-dren. Questions about child aggres-sion and maltreatment were includedin the latter interview.

Mothers who met 1 of the followingcriteria were excluded from the studysample: did not participate in the3-year in-home interview (n 1610)

or, in the 5-year in-home interview(n 799), answered50%of the childaggression scale items, so that a validscore could not be calculated (missingitem values were imputed when50%of the scale items were missing) (n21), or did not report whether they hadspanked the index child at age 3 (n 7). Study participants (n 2461) weremore likely than nonparticipants (n2437) to have some college education,to be black, to have been born in theUnited States, and to be Christian (non-Catholic) or nonreligious (2 tests, P.05). Participants and nonparticipantsdid not differ according to childs gen-der, mothers age, household income,or marital status at childs birth.

The institutional review board of theTulane University Health Sciences Cen-ter reviewed this secondary data anal-ysis study and considered it exempt.Participant recruitment procedureswere approved by the institutional re-view boards at Columbia Universityand Princeton University, the aca-demic homes to the FFCWS. Additionaldetails pertaining to the original studywere published elsewhere.23

Measures

Data Collection

All data were provided through self-report from the mother. All child-related questions were asked with re-spect to the identied index child.

Predictor Variable: Mothers Use of CP

This variable reects how frequently themother spanked her 3-year-old child (apeak age for spanking use2,24) for mis-behaving or acting up in the month be-fore the interview. Responses werecodedasnever (scoreof 0), onceor twice(score of 1), or2 times (score of 2).

Dependent Variable: Index ChildsAggression at Age 5

This was assessed by using 12 itemsfrom the Child Behavior Checklist ver-

e1058 TAYLOR et al. Provided by Indonesia:AAP Sponsored on December 10, 2010 www.pediatrics.orgDownloaded from

sion for age 5,25 which asked whetherthe child argues a lot; is cruel, bullies,and shows meanness to others; de-stroys his or her own things; destroysthings belonging to family members orothers; is disobedient at home; is dis-obedient at school or in child care;gets in many ghts; physically attackspeople; screams a lot; teases a lot;threatens people; or is unusually loud( .82). Response options were nottrue (score of 0), somewhat or some-times true (score of 1), or very true oroften true (score of 2). An averagescore for the 12 items was obtained(mean: 0.40; median: 0.33; SD: 0.33). Be-cause the variable was highly skewed,it was dichotomized at the median andanalyzed as lower aggression (scoresof 00.32) versus higher aggression(scores of 0.331.83).

Index Childs Aggression at Age 3

This was assessed by using 19 itemsfrom the Child Behavior Checklist ver-sion for age 3,25 which asked whetherthe child is deant; has demands thatmust be met immediately; is disobedi-ent; does not seem to feel guilty aftermisbehaving; is easily frustrated; getsin many ghts; hits others; has angrymoods; shows behavior that punish-ment does not change; screams a lot;is selsh or will not share; is stubborn,sullen, or irritable; has temper tan-trums or hot temper; is uncooperative;wants a lot of attention; cannot standwaiting and wants everything now; de-stroys things belonging to family mem-bers or other children; hurts animalsor people without meaning to; or phys-ically attacks people ( .88). Re-sponse options were not true (score of0), somewhat or sometimes true(score of 1), or very true or often true(score of 2). An average score for the19 items was obtained (mean: 0.62;median: 0.58; SD: 0.36). Because thevariable was highly skewed, it was di-chotomized at the median value andanalyzed as lower aggression (scores

of 00.57) versus higher aggression(scores of 0.581.95).

Maternal Parenting Risks

Mothers use of CP is associated withuse of other harsh parenting (physicaland psychological maltreatment; theterm maltreatment is used through-out for these variables, rather than themore-commonly used term aggres-sion, so that these variables are notconfused with the outcome variable ofchild aggression), child neglect, inti-mate partner aggression and violence,and maternal parenting stress, de-pression, and consideration of abor-tion5; use of alcohol and/or drugs alsois linked with use of CP.26 Literaturendings showed links betweenmost ofthese variables (especially harsh par-enting, exposure to intimate partneraggression, parental depression, andparental stress) and childhood ag-gression.2732 Therefore, these vari-ables, which were assessed when thechild was age 3, for consistency withthe main predictor variable, mightconfound the association between CPand child aggression.

Child Maltreatment and IntimatePartner Aggression and Violence

Three child maltreatment proxieswere assessed with the Parent-ChildConict Tactics Scale,33 that is, physi-cal maltreatment (4 items), psycho-logical maltreatment (5 items), andneglect (5 items). (The physical mal-treatment scale usually contains a fthitem regarding spanking; however,this item was removed so that it wouldnot overlap with our main predictorvariable.) Intimate partner aggressionor violence experienced by the mothersince the index childs birth, eitherfrom the father or from a current part-ner, was assessed by using 7 items; 3items from the Conict Tactics Scale34

were adapted to assess physical ag-gression and 4 from the Spouse Obser-vation Checklist35 and the report by

Lloyd36 were adapted to assess psycho-logical aggression. Because childphysical and psychological maltreat-ment results were highly skewed, theywere dichotomized at their median val-ues (Table 1) for analysis. Child neglectand intimate partner aggression or vi-olence also were dichotomized (anyversus none).

Other Maternal Risks

Maternal parenting stress, major de-pression, use of alcohol and/or drugs,and unwantedness of the index childpregnancy were assessed. Stress wasmeasured by using 11 items from theParenting Stress Index37 ( .86). De-pression was measured on the basisof criteria from theDiagnostic and Sta-tistical Manual of Mental Disorders,Fourth Edition38 for major depression,by using section A of the CompositeInternational Diagnostic Interview-Short Form39; detailed scoring meth-ods were described previously.40 As-sessment of alcohol and other druguse also was derived from the Compos-ite International Diagnostic Interview-Short Form39 and was coded positivelyif the mother had had 4 drinks onone day, or had used any of nine differ-ent substances at least once, in thepast 12 months. Unwantedness ofthe index child pregnancy was approx-imated on the basis of the mothers re-sponse to the following baseline ques-tion: When you found out you werepregnant, did you think about havingan abortion? Parenting stress was an-alyzed as a continuous variable. All ofthe other maternal risk variables weredichotomized (yes versus no).

Family Demographic Features

Family demographic features were se-lected on the basis of their availabilityin the FFCWS data set and previous em-pirical evidence suggesting their asso-ciation with parenting risk and/or useof CP.

ARTICLES


Statistical Analyses

Descriptive and bivariate statisticalanalyses were conducted to examineassociations between all assessedma-ternal parenting risk factors/demo-graphic features and mothers use ofCP (Table 1) and child aggression atage 5 (Table 2). The Kruskal-Wallis test

was used for continuous variables be-cause the assumption of equal vari-ances generally was not met; 2 testswere used for binary and categoricalvariables.

Four multivariate logistic regressionmodels were used to examine predic-tion of child aggression at age 5 (Table

3). Each model controlled for parentsmarital status at the time of the indexchilds birth and interview city, be-cause these variables were part of thesampling design. Model 1 tested use ofCPwhen the child was 3 years of age asthe sole predictor. Model 2 added thechilds level of aggression at age 3.

TABLE 1 Descriptive and Bivariate Statistical Analyses of Maternal Characteristics According to Mothers Use of CP in Month Before Interview WhenChild Was 3 Years of Age

Total Sample(N 2461)

Did Not Spank(n 1123)

Spanked 1 or 2Times (n 686)

Spanked3Times (n 652)

P

Maternal parenting risksPsychological maltreatment of child, no. of

incidents in previous year, median(range: 0115 incidents)

25 16 25 33 .001

Physical maltreatment of child (spankingnot included), no. of incidents inprevious year, median (range: 0108incidents)

12 4 16 26 .001

Any neglect of child in previous year, % 11.3 8.5 13.3 14.3 .001Victim of intimate partner aggression

and/or violence since birth of indexchild, %

53.2 47.4 57.9 58.4 .001

Parenting Stress Index score, median(range: 044)

12 11 13 13 .001

Major depression, % 21.3 18.2 21.4 26.5 .001Use of drugs and/or alcohol, % 16.1 13.0 16.6 20.9 .001Considered aborting this child, % 27.7 24.8 29.3 30.8 .05

Mother and family demographic featuresMale child, % 51.9 48.9 53.4 55.5 .05Mothers age, median, y (range: 1650 y) 27 28 27 26 .001Mothers education, % .01Less than high school 32.7 34.9 34.1 27.5 .01High school 30.6 29.1 29.2 34.8 .05Some college 25.6 23.9 26.2 27.8College graduate 11.0 12.1 10.2 9.8

Mothers race/ethnicity, % .001Black 50.7 45.3 53.6 56.8 .001Hispanic 24.2 29.2 23.3 16.4 .001White 21.9 21.9 19.7 24.1Other 3.1 3.2 3.2 2.6

Mother foreign-born, % 12.6 17.5 10.5 6.1 .001Mothers religion, % .001Catholic 26.4 32.3 26.8 15.8 .001Protestant 39.9 36.0 40.2 46.3 .001Other Christian 12.7 11.0 12.4 16.1 .01No religious preference 17.0 16.6 16.8 17.9Other 3.1 3.5 2.9 2.8

Mother-father relationship status, % NSMarried 32.1 33.6 31.5 30.4Cohabiting 27.7 28.0 26.7 28.2Visiting 29.6 27.8 30.9 31.3No relationship 10.6 10.6 10.9 10.1

Mothers annual household income,median, ln $ (range: 013.8; actualmedian: $23 721)

10 10 10 10 NS

NS indicates not statistically signicant. Kruskal-Wallis tests were used for continuous variables, because equal-variance assumptions generally were not met; 2 tests were used for binaryand categorical variables. Missing data for each variable equaled1%.


Model 3 added the assessed maternalparenting risk factors. Model 4 addedall of the assessed demographic fea-tures.

RESULTS

Almost one-half (45.6%) of themothersreported no use of spanking in the pre-vious month, 27.9% reported spanking1 or 2 times, and 26.5% reported

spanking2 times. All of the examinedrisk factors and demographic fea-tures, except for parental relationshipstatus and income, were signicantlyassociated with CP in bivariate analy-ses (Table 1). More-frequent use of CPwas signicantly associated with in-creased maternal parenting risks. Re-spondents who had a male index childor a high school education or who

were younger, black, or Christian (non-Catholic) were at higher-than-averagerisk for using CP, whereas those whowere Hispanic, foreign-born, or Catho-lic were at lower-than-average risk.

As with CP, higher levels of each of theassessed maternal parenting riskswere associated with risk for higherlevels of child aggression (Table 2).These ndings conrmed that theseparenting risk factors might indeedconfound the link between CP and childaggression and therefore should becontrolled for in the nal analysis. Inaddition, children who were male orwhose mothers were younger, hadlower education levels or householdincome, had no religious preference,or had just a visiting relationshipwith the father were at risk for higherlevels of aggression. Race/ethnicityand nativity were not associated withchild aggression.

Table 3 presents results from 4 logisticregression models predicting higherlevels of child aggression at age 5.Across all 4 models, mothers more-frequent use of CP (2 times in theprevious month) when the child wasage 3 was a statistically signicantpredictor of higher levels of aggres-sion when the child was age 5. At thebivariate level (model 1), more-frequent use of CP more than doubledthe odds of higher aggression levels,and less-frequent use of CP (1 or 2times in the previous month) in-creased the odds by almost 40%. Whenthe childs level of aggression at age 3was included (model 2), the impact ofCP use on subsequent aggression wasdecreased almost in half; this was be-cause, as expected, having a higherlevel of aggression at age 3 was astrong predictor of having a higherlevel of aggression at age 5. When theassessed maternal parenting riskswere included (model 3), the impact ofmore-frequent CP use was decreasedby another 27% and less-frequent CP

TABLE 2 Descriptive and Bivariate Statistical Analyses of Maternal Characteristics According toChild Aggressive Behavior When Child Was 5 Years of Age

Total Sample(N 2461)

LowerAggression(n 1137)

HigherAggression(n 1324)

P

Maternal parenting risksPsychological maltreatment of child, no. of incidents

in previous year, median (range: 0115incidents)

25 18 27 .001

Physical maltreatment of child (spanking notincluded), no. of incidents in previous year,median (range: 0108 incidents)

12 9 16 .001

Any neglect of child in previous year, % 11.3 8.1 14.1 .001Victim of intimate partner aggression and/or

violence since birth of index child, %53.2 47.9 57.8 .001

Parenting Stress Index score, median (range: 044) 12 11 13 .001Major depression, % 21.3 17.5 24.6 .001Use of drugs and/or alcohol, % 16.1 13.9 18.0 .01Considered aborting this child, % 27.7 25.0 30.0 .01

Mother and family demographic featuresMale child, % 51.9 47.5 55.7 .001Mothers age, median, y (range: 1650 y) 28 28 26 .001Mothers education, % .001Less than high school 32.7 29.4 35.6 .01High school 30.6 28.5 32.5 .05Some college 25.6 28.0 23.5 .05College graduate 11.0 14.1 8.3 .001

Mothers race/ethnicity, % NSBlack 50.7 49.3 51.9Hispanic 24.2 24.4 24.0White 21.9 23.3 20.6Other 3.1 2.9 3.2

Mother foreign-born, % 12.6 13.5 11.7 NSMothers religion, % .05Catholic 26.4 27.7 25.3Protestant 39.9 41.3 38.7Other Christian 12.7 12.7 12.8No religious preference 17.0 14.5 19.1 .01Other 3.1 3.1 3.2

Mother-father relationship status, % .001Married 32.1 36.6 28.3 .001Cohabiting 27.7 27.1 28.2Visiting 29.6 25.9 32.8 .001No relationship 10.6 10.4 10.7

Mothers annual household income, ln $ (range:013.8; actual median: $23 721)

10.1 10.1 10.0 .001

NS indicates not statistically signicant. Kruskal-Wallis tests were used for continuous variables, because equal-varianceassumptions generally were notmet;2 tests were used for binary and categorical variables. Missing data for each variableequaled1%.

ARTICLES


use was no longer statistically signi-cant. The nal model (model 4), whichalso included all assessed demographicfeatures, suggests that the odds of thechild having a higher level of aggressionat age 5 were increased by 49% withmore-frequent use of CP at age 3.

DISCUSSION

Our study accounted for 8 maternalparenting risks for child aggression,

including other forms of harsh parent-ing besides CP, child neglect, intimatepartner aggression or violence, andmaternal parenting stress, depres-sion, use of substances, and consider-ation of abortion. As anticipated, all ofthese factors were found to be associ-ated both with CP use and with childaggression and therefore had the po-tential to be important confounders ofthis association. Although previous

studies on this topic accounted forparenting risks such as maternal psy-chopathological conditions,22 parentalmarital adjustment or conict,14,15

and/or relevant demographic fea-tures, no other studies, to our knowl-edge, accounted simultaneously forall of the confounders addressed inthis study while also addressing theother key conditions (statistical sig-nicance, temporality, and initial lev-

TABLE 3 ORs of CP and Other Family Characteristics Predicting Higher Levels of Child Aggression at Age 5

Model 1 (N 2461) Model 2 (N 2461) Model 3 (N 2432) Model 4 (N 2432)

OR (95% CI) P OR (95% CI) P OR (95% CI) P OR (95% CI) P

Mothers use of CPMother spanked2 times 2.03 (1.822.26) .0001 1.59 (1.421.79) .0001 1.43 (1.221.69) .0001 1.49 (1.241.78) .0001Mother spanked 1 or 2 times 1.37 (1.171.61) .0001 1.21 (1.021.45) .032 1.15 (0.931.40) NS 1.17 (0.941.44) NS

Higher level of child aggression at age 3 3.79 (3.434.19) .0001 3.34 (3.053.66) .0001 3.35 (3.063.67) .0001Maternal parenting risksPsychological maltreatment of child 0.98 (0.811.18) NS 0.97 (0.801.19) NSPhysical maltreatment of child (spanking not

included)1.11 (0.951.28) NS 1.10 (0.931.29) NS

Neglect of child 1.14 (0.841.55) NS 1.13 (0.831.53) NSVictim of intimate partner aggression and/or

violence1.14 (0.941.38) NS 1.15 (0.941.39) NS

Parenting stress 1.02 (1.011.04) .001 1.02 (1.011.04) .006Major depression 1.07 (0.911.26) NS 1.07 (0.921.25) NSUse of drugs and/or alcohol 1.14 (0.871.50) NS 1.15 (0.881.51) NSConsidered aborting this child 1.04 (0.791.37) NS 1.04 (0.771.40) NS

Demographic featuresMale child 1.31 (1.091.58) .004Mothers age 1.00 (0.971.03) NSMothers educationLess than high school (reference) 1.00High school 0.97 (0.801.19) NSSome college 0.78 (0.650.94) .009College graduate 0.66 (0.440.99) .043

Mothers race/ethnicityBlack (reference) 1.00Hispanic 0.99 (0.801.22) NSWhite 1.07 (0.851.34) NSOther 1.13 (0.721.80) NS

Mother foreign-born 1.04 (0.771.41) NSMothers religionProtestant (reference) 1.00Catholic 1.10 (0.751.60) NSOther Christian 0.99 (0.641.55) NSNo religious preference 1.24 (1.021.50) .030Other 1.32 (0.772.26) NS

Mother-father relationship statusMarried (reference) 1.00Cohabiting 0.99 (0.701.39) NSVisiting 1.18 (0.891.56) NSNo relationship 1.00 (0.651.54) NS

Mothers annual household income 0.96 (0.901.04) NS

OR indicates odds ratio; CI, condence interval. Missing data for each variable equaled1.2%. All models were adjusted for 2 key variables used in the sampling design, that is, parentsmarital status at time of index childs birth (married or unmarried) and city. In model 4, a Bonferroni correction for multiple tests suggested that only ndings with P .002 should beconsidered.


els of child aggression) that must bemet to assert more strongly that useof CP leads to higher levels of aggres-sion in children.

We found that, even when all of thosematernal parenting risks were con-trolled for, mothers more-frequentuse of CP with their 3-year-old childrenincreased the odds of those childrenbeing more aggressive at age 5. Thisnding is consistent with dozens ofother studies that showed a signicantstatistical link between the use of CPand child aggression, including stud-ies17,19,4151 summarized by Gershoff11

and other studies22,5255 conductedsince the time of that meta-analysis; italso is consistent with studies thatsimilarly controlled for childrens ini-tial level of aggression.1422 In our nalmodel, CP was the only parenting riskfactor examined that remained statis-tically linked (after Bonferroni correc-tion) with subsequent child aggres-sion. This nding seems to support asocial learning approach to under-standing the cycle of violence,56

whereby the child learns to be aggres-sive by being treated directly withaggression.

One may wonder, then, why child phys-ical maltreatment by the mother wasnot related to child aggression. Thephysical maltreatment subscale of theParent-Child Conict Tactics Scale con-tained 5 items (shook; hit on the bot-tom with something like a belt, hair-brush, a stick, or some other hardobject; slapped on the hand, arm, orleg; pinched; and spanked on the bot-tom with a bare hand). When the latteritemwas removed, however, therewasa substantial decrease in the reliabilitycoefcient for this subscale (from.63 to .48). Furthermore, 2 of theremaining 4 items were reported veryrarely (shook, 5%; pinched, 8%);spanking was much more common.Therefore, the lack of association be-

tween child physical maltreatment bythe mother and subsequent childaggression may be an issue of statisti-cal power rather than theoreticalinconsistency.

There are several limitations to ourstudy. First, this study focused only onmothers use of CP and did not accountfor fathers or other caregivers use ofCP. Furthermore, all variables in thisstudy were based on mothers self-re-ports; there were no observationaldata, and reports might be subject tobiases related to recall and/or socialdesirability. Also, there is always con-cern in observational studies that un-measured confounders may explainthe associations found; however, whenthis concern was addressed to someextent in a previous study by usinghierarchical linear modeling, the linkbetween CP and child aggressionremained.21

Given the problem of potential unmea-sured confounders, it is not possible toassert causality between CP and childaggression in observational studiessuch as this. As with other studies ofrisk behaviors (eg, smoking), it wouldbe unethical to assign parents ran-domly to use CP or not to use CP, giventhe existing evidence linking CP withassociated harm in children. There-fore, we must rely largely on evidencefrom observational studies, such asthe current one, that aim to accountfor as many other possible explana-tions of the association between CPand subsequent child aggression aspossible.

CONCLUSIONS

This study adds to the growing body ofliterature suggesting that parental useof CP may lead to increased child ag-gression. This evidence base suggeststhat primary prevention of violencecan start with efforts to prevent theuse of CP against children. Pediatri-cians and others concerned with chil-

drens well-being know that CP is not anecessary form of child discipline andthat other, more or equally effective,nonphysical forms of discipline exist.Reductions in parents use of CP (dem-onstrated in randomized, clinical trialsof parenting interventions designed totreat conduct disorder in children)have been shown to reduce childrenssubsequent aggression57; additionalstudies of this nature could aid in ad-dressing the issue of CP as a causalagent in subsequent aggression. How-ever, efforts to teach nonphysical dis-cipline strategies to parents in generalpediatric ofce visits have met withmixed success.58,59 Research to furthersuch efforts is needed, given that par-ents cite pediatricians as the profes-sionals from whom they are mostlikely to seek advice regardingchild discipline.60 In addition, broaderpopulation-based efforts, such as so-cial marketing campaigns, are neededto shift normative beliefs and ex-pected outcomes regarding CP60 andto strengthen the message of theAmerican Academy of Pediatrics thatother child discipline strategies thatare effective and less risky should beused instead of CP.

ACKNOWLEDGMENTSThis research was supported by theCenters for Disease Control and Pre-vention (grant R49 CE000915-02) andthe National Institute of Mental Healthand the National Institute of ChildHealth and Human Development (grantR01 HD41141-02).

We thankMili Duggal and Ransome Ekefor their assistance with literature re-view and table construction, as well asElizabeth Gershoff for sharing reviewsof key literature. We also thank the2461 study participants who gave theirvaluable time and information to thisstudy, along with the anonymous re-viewers for their helpful commentsand critiques of this article.

ARTICLES


REFERENCES

1. Durrant JE. Physical punishment, culture,and rights: current issues for profession-als. J Dev Behav Pediatr. 2008;29(1):5566

2. Straus MA, Stewart JH. Corporal punish-ment by American parents: national data onprevalence, chronicity, severity, and dura-tion, in relation to child and family charac-teristics. Clin Child Family Psychol Rev.1999;2(2):5570

3. Daro D. Public Opinion and Behaviors Re-garding Child Abuse Prevention: 1999 Sur-vey. Chicago, IL: Prevent Child AbuseAmerica; 1999

4. Day RD, Peterson GW, McCracken C. Predict-ing spanking of younger and older childrenby mothers and fathers. J Marriage Fam.1998;60(1):7994

5. Taylor CA, Guterman NB, Lee SJ, RathouzPJ. Intimate partner violence, maternalstress, nativity, and risk for maternal mal-treatment of young children. Am J PublicHealth. 2009;99(1):175183

6. SurveyUSA. Disciplining a child 08/24/05. Available at: www.surveyusa.com/50StateDisciplineChild0805SortedbyTeacher.htm. Accessed July 3, 2009

7. Board of the Canadian Psychological Asso-ciation. Policy statement on physical pun-ishment of children and youth. Available at:www.docstoc.com/docs/5250464/Board-of-the-Canadian-Psychological-Association-Policy-Statement-on. Accessed September10, 2009

8. National Association of School Psychologists.Position statement on corporal punishmentin schools. Available at: www.nasponline.org/aboutnasp/pospapercorppunish.aspx. Ac-cessed September 10, 2009

9. Durrant JE, Ensom R. Joint Statement onPhysical Punishment of Children and Youth.Ottawa, Canada: Coalition on Physical Pun-ishment of Children and Youth; 2004

10. Wolraich ML, Aceves J, Feldman HM, et al.Guidance for effective discipline. Pediatrics.1998;101(4):723728

11. Gershoff ET. Corporal punishment by par-ents and associated child behaviors andexperiences: a meta-analytic and theoreti-cal review. Psychol Bull. 2002;128(4):539579

12. Baumrind D, Larzelere RE, Cowan PA. Ordi-nary physical punishment: is it harmful?Comment on Gershoff (2002). Psychol Bull.2002;128(4):580589

13. Lansford JE, Criss MM, Dodge KA, Shaw DS,Pettit GS, Bates JE. Trajectories of physicaldiscipline: early childhood antecedents and

developmental outcomes. Child Dev. 2009;80(5):13851402

14. Campbell SB, Pierce EW, Moore G, Marako-vitz S, Newby K. Boys externalizing prob-lems at elementary school age: pathwaysfrom early behavior problems, maternalcontrol, and family stress. Dev Psycho-pathol. 1996;8(4):701719

15. Cohen P, Brook JS. The reciprocal inuenceof punishment and child behavior disorder.In: McCord J, ed. Coercion and Punishmentin Long-Term Perspectives. New York, NY:Cambridge University Press; 1995:154164

16. Grogan-Kaylor A. The effect of corporal pun-ishment on antisocial behavior in children.Soc Work Res. 2004;28(3):153162

17. Gunnoe ML, Mariner CL. Toward adevelopmental-contextual model of the ef-fects of parental spanking on childrens ag-gression. Arch Pediatr Adolesc Med. 1997;151(8):768775

18. Kandel DB, Wu P. Disentangling mother-child effects in the development of antiso-cial behavior. In: McCord J, ed. Coercion andPunishment in Long-Term Perspectives.New York, NY: Cambridge University Press;1995:106123

19. Singer JL, Singer DG, Rapaczynski WS. Fam-ily patterns and television viewing as pre-dictors of childrens beliefs and aggression.J Commun. 1984;34(2):7389

20. Weiss B, Dodge KA, Bates JE, Pettit GS. Someconsequences of early harsh discipline:child aggression and a maladaptive socialinformation processing style. Child Dev.1992;63(6):13211335

21. Grogan-Kaylor A. Corporal punishment andthe growth trajectory of childrens antiso-cial behavior. Child Maltreat. 2005;10(3):283292

22. Berlin LJ, Ispa JM, Fine MA, et al. Correlatesand consequences of spanking and verbalpunishment for low-income white, AfricanAmerican, and Mexican American toddlers.Child Dev. 2009;80(5):14031420

23. Reichman NE, Teitler JO, Garnkel I, McLana-han SS. Fragile Families: sample and de-sign. Child Youth Serv Rev. 2001;23(4/5):303326

24. Paolucci EO, Violato C. Ameta-analysis of thepublished research on the affective, cogni-tive, and behavioral effects of corporal pun-ishment. J Psychol. 2004;138(3):197221

25. Achenbach TM, Rescorla LA. Manual for theASEBA Preschool Forms and Proles. Burl-ington, VT: University of Vermont, ResearchCenter for Children, Youth, and Families;2000

26. Lee SJ, Perron BE, Taylor CA, Guterman NB.Paternal psychosocial characteristics andcorporal punishment of their 3-year oldchildren. J Interpers Violence. In press

27. Brennan PA, Hall J, Bor W, Najman JM, Wil-liams G. Integrating biological and socialprocesses in relation to early-onset persis-tent aggression in boys and girls. Dev Psy-chol. 2003;39(2):309323

28. Koblinsky SA, Kuvalanka KA, Randolph SM.Social skills and behavior problems of ur-ban, African American preschoolers: role ofparenting practices, family conict, andmaternal depression. Am J Orthopsychia-try. 2006;76(4):554563

29. McFarlane JM, Groff JY, OBrien JA, WatsonK. Behaviors of children who are exposedand not exposed to intimate partnerviolence: an analysis of 330 black, white,and Hispanic children. Pediatrics. 2003;112(3). Available at: www.pediatrics.org/cgi/content/full/112/3/e202

30. Meadows SO, McLanahan SS, Brooks-GunnJ. Parental depression and anxiety andearly childhood behavior problems acrossfamily types. J Marriage Fam. 2007;69(5):11621177

31. Owens EB, Shaw DS. Predicting growthcurves of externalizing behavior across thepreschool years. J Abnorm Child Psychol.2003;31(6):575590

32. DeMulder EK, Denham S, Schmidt M, Mitch-ell J. Q-sort assessment of attachment se-curity during the preschool years: linksfrom home to school. Dev Psychol. 2000;36(2):274282

33. Straus MA, Hamby SL, Finkelhor D, MooreDW, Runyan D. Identication of child mal-treatment with the Parent-Child ConictTactics Scales: development and psycho-metric data for a national sample of Amer-ican parents. Child Abuse Negl. 1998;22(4):249270

34. Straus MA, Hamby S, Boney-McCoy S, Sugar-man D. The Revised Conict Tactics Scale(CTS2): development and preliminary psy-chometric data. J Fam Issues. 1996;17(3):283316

35. Weiss RL, Margolin G. Assessment ofmaritalconict and accord. In: Ciminero AR, Cal-houn KD, Adams HE, eds. Handbook of Be-havioral Assessment. New York, NY: Wiley;1977:555602

36. Lloyd SA. Physical aggression, distress, andeveryday marital interactions. In: Cahn DD,Lloyd SA, eds. Family Violence From a Com-munication Perspective. Thousand Oaks,CA: Sage; 1996:177198

37. Abidin RR. Parenting Stress Index. 3rd ed.


Odessa, FL: Psychological AssessmentResources; 1995

38. American Psychiatric Association. Diagnos-tic and Statistical Manual of Mental Disor-ders. 4th ed. Washington, DC: American Psy-chiatric Association; 1994

39. Kessler RC, Andrews G, Mroczek D, Ustun B,Wittchen HU. The World Health OrganizationComposite International Diagnostic Inter-view Short-Form (CIDI-SF). Int J MethodsPsychiatr Res. 1998;7(4):171185

40. Fragile Families Study Group. Fragile Families:scales documentation andquestion sources forthree-year questionnaires: revised 4/26/06.Available at: www.fragilefamilies.princeton.edu/documentation/core/scales/ff3yrscales.pdf.Accessed July 3, 2009

41. BeckerWC, Peterson DR, Luria Z, ShoemakerDJ, Hellmer LA. Relations of factors derivedfrom parent-interview ratings to behaviorproblems of ve-year-olds. Child Dev. 1962;33(2):509535

42. Deater-Deckard K, Dodge KA, Bates JE, PettitGS. Physical discipline among African Amer-ican and European Americanmothers: linksto childrens externalizing behaviors. DevPsychol. 1996;32(6):10651072

43. Engfer A, Schneewind KA. Causes and con-sequences of harsh parental punishment:an empirical investigation in a representa-tive sample of 570 German families. ChildAbuse Negl. 1982;6(2):129139

44. Eron LD. Parent-child interaction, televisionviolence, and aggression of children. AmPsychol. 1982;37(2):197211

45. Kandel DB. Parenting styles, drug use, andchildrens adjustment in families of youngadults. J Marriage Fam. 1990;52(1):183196

46. Levin H, Sears RR. Identication with par-ents as a determinant of doll play aggres-sion. Child Dev. 1956;27(2):135153

47. Mahoney A, Donnelly WO, Lewis T, MaynardC. Mother and father self-reports of corpo-ral punishment and severe physical aggres-sion toward clinic-referred youth. J ClinChild Psychol. 2000;29(2):266281

48. McCabe KM, Clark R. Family protective fac-tors among urban African American youth.J Clin Child Psychol. 1999;28(2):137150

49. McLeod JD, Shanahan MJ. Poverty, parent-ing, and childrens mental health. Am SociolRev. 1993;58(3):351366

50. Sears RR. Relation of early socialization ex-periences to aggression in middle child-hood. J Abnormal Soc Psychol. 1961;63(3):466492

51. Simons RL, Lin K-H, Gordon LC. Socializationin the family of origin and male datingviolence: a prospective study. J MarriageFam. 1998;60(2):467478

52. Tang CS-K. Corporal punishment andphysical maltreatment against children: acommunity study on Chinese parents inHong Kong. Child Abuse Negl. 2006;30(8):893907

53. Pagani LS, Tremblay RE, Nagin D, ZoccolilloM, Vitaro F, McDuff P. Risk factor models foradolescent verbal and physical aggressiontoward mothers. Int J Behav Dev. 2004;28(6):528537

54. Nelson DA, Hart CH, Chongming Y, Olsen JA,Shenghua J. Aversive parenting in China: as-sociationswith child physical and relationalaggression. Child Dev. 2006;77(3):554572

55. Sim TN, Ong LP. Parent physical punishmentand child aggression in a Singapore Chi-nese preschool sample. J Marriage Fam.2005;67(1):8599

56. Bandura A. Aggression: A Social LearningAnalysis. Englewood Cliffs, NJ: Prentice-Hall;1973

57. Beauchaine TP, Webster-Stratton C, ReidMJ. Mediators, moderators, and predictorsof 1-year outcomes among children treatedfor early-onset conduct problems: a latentgrowth curve analysis. J Consult Clin Psy-chol. 2005;73(3):371388

58. Sege RD, Perry C, Stigol L, et al. Short-termeffectiveness of anticipatory guidance to re-duce early childhood risks for subsequentviolence. Arch Pediatr Adolesc Med. 1997;151(4):392397

59. Barkin SL, Finch SA, Ip EH, et al. Is ofce-based counseling about media use, time-outs, and rearm storage effective? Resultsfrom a cluster-randomized, controlled trial.Pediatrics. 2008;122(1). Available at:www.pediatrics.org/cgi/content/full/122/1/e15

60. Taylor CA, Hamvas L, DeJong W, Rice JC.Normative Beliefs, Expectations, and Use ofCorporal Punishment: Findings From anRDD Community Survey. Philadelphia, PA:American Public Health Association; 2009

ARTICLES


DOI: 10.1542/peds.2009-2678 2010;125;e1057-e1065; originally published online Apr 12, 2010; Pediatrics

Catherine A. Taylor, Jennifer A. Manganello, Shawna J. Lee and Janet C. Rice Aggressive Behavior

Mothers' Spanking of 3-Year-Old Children and Subsequent Risk of Children's

& ServicesUpdated Information

http://www.pediatrics.org/cgi/content/full/125/5/e1057including high-resolution figures, can be found at:

References

http://www.pediatrics.org/cgi/content/full/125/5/e1057#BIBLat: This article cites 42 articles, 5 of which you can access for free

Rs)3Peer Reviews (PPost-Publication

http://www.pediatrics.org/cgi/eletters/125/5/e1057R has been posted to this article: 3One P

Permissions & Licensing

http://www.pediatrics.org/misc/Permissions.shtmltables) or in its entirety can be found online at: Information about reproducing this article in parts (figures,

Reprints http://www.pediatrics.org/misc/reprints.shtml

Information about ordering reprints can be found online:

. Provided by Indonesia:AAP Sponsored on December 10, 2010 www.pediatrics.orgDownloaded from

anak di pukul

Documents