Chdd Abuse & Neglecr. Vol 9, pp 159-163. Prmted m the U.S.A. All nghts reserved
0145.2134/85 $3.00 + .W Copyright C 1985 Pergamon Press Ltd
1985
SELF-ESTEEM OF ABUSED CHILDREN R. KIM OATES, M.D., M.H.P., F.R.A.C.M.A., F.R.A.C.P., DOUGLAS FORREST, B.Sc.
AND ANTHONY PEACOCK, B-SC. (APP.PsY.)
M.A.P.S.
Royal Alexandra Hospital for Children, Camperdown 2050, New South Wales, Australia Abstract-Thirty-seven children admitted to hospital with a diagnosis of abuse at an average of 5.5 years previously were compared with 37 non-abused children matched for age, sex, ethnic group, school and social class. The mean age of the children was 8.9 years (range 4.6-14.0 years). Self-esteem in the children was assessed by means of a structured interview with the child and the Piers-Harris Self-Concept Scale. The abused children saw themselves as having significantly fewer friends than the comparison children and they played with friends less often. They were less ambitious than the non-abused children with regard to the sort of occupation they would like as adults and were significantly lower in self-concept on the Piers-Harris Self-Concept Scale. These results show that, as well as providing treatment for abusive parents, a long-term program which aims at improving interpersonal relationship skills and self-esteem is required for abused children. Resume-On a hospital& 37 enfants avec le diagnostic de mauvais traitement. Ces 37 enfants ont Cte revus en moyenne 5.5 ans plus tard et compares avec 37 enfants du meme Age,du mtme sexe, du m&megroupe ethnique et de la meme categoric scolaire et sociale. Au moment de la comparaison, l’age moyen des enfants ttait de 8, 9 ans avec une variation de 4, 6 a 14 ans. L’estime de G-m&me a ete evaluke par le moyen d’une interview structurke avec l’enfant et en utilisant la “Piers-Harris Self-Concept Scale.” II est apparu que les enfants maltraitts estimaient avoir moins d’amis et il semblait bien qu’ils jouaient avec des amis moins souvent que les autres enfants. Leurs ambitions ttaient moms tlevees que celles des enfants non maltraitb, en particulier en ce qui conceme leur avenir professionel et en utilisant la Piers-Harris Self-Concept Scale, il etait evident que l’estime qu’ils avaient pour eux-m6mes ttait plus basse que chez les enfants ttmoins. Les resultats montrent done que non seulement il faut s’occuper des parents maltraitants mais qu’il faut egalement foumir aux enfants maltraitts un programme a long terme visant a amthorer leur capacite de communiquer avec autrui et s’effor9ant de leur redormer confiance en eux-m&mes. K~J Words-Self-esteem,
Parental expectations.
INTRODUCTION THERE IS CONFLICTING EVIDENCE about the self-esteem of abused children. It would to find that children who have been the subject of high and unrealistic expectations, and who have been abused and neglected, would show a diminution of selfconcept. This was shown by Martin and Beezley [ 1] who found that, based on a child interview, 62% of a group of 50 children followed for a mean of 4.5 years had low self-esteem. Similar findings where 39 abused children were interviewed by a psychologist at an average of 4 years after the presenting abusive incident, were described by Lynch and Roberts [2]. These authors reported that the most common problems reported by the psychologist were signs of anxiety. extreme shyness and fear of failure. In contrast, Elmer’s study [3] of 17 children where abuse had occurred eight years previously and 17 children matched for social class, did not show any difference in self-concept
not be surprising
This research was supported in part by a grant from the Children’s Medical Research Foundation, Royal Alexandra Hospital for Children, Camperdown 2050. 159
160
R. Kim Oates, Douglas Forrest and Anthony Peacock Table 1. Number of Friends Claimed by Abused Children A Lot
Average number
Few
Abuse group (N = 37)
17
4
16
Comparison group (N = 37)
26
7
4
Number of friends
x= = 9.9, df = 2,p
between the abused and comparison children. In that study, self-esteem was assessed using a formal test, the Piers-Harris Self-Concept Scale [4] rather than an interview. The only other reported study using this scale [5] also showed no difference between the mean scores of 30 abused and 30 non-abused children. However a factor analysis of these results showed that the abused children were below the comparison children on one of six specific aspects of selfconcept. The aim of this study was to clarify some of this conflicting information about self-concept by reviewing a group of abused children using interviews as well as the Piers-Harris SelfConcept Scale.
METHODS Each of the 37 children who had been abused at an average of 5.5 years previously was individually matched with a child not known to have been abused who attended the same school and who was of the same age (to within six months), sex and ethnic group, and whose parents were of the same social class. Each child participated in a short, structured interview and was given the Piers-Harris Self-Concept Scale. This test consists of 80 sentences to which the child is asked to answer yes or no, indicating whether the statement is true about him. The psychologist conducting the test was not informed as to which children were in the study or comparison groups. As this was part of a larger study, the results were analyzed using the analysis of variance program of the Statistical Package for the Social Sciences [6].
RESULTS The study group sample consisted of 23 boys and 14 girls. Their mean age was 8.9 years (range 4.6 to 14.4 years). Eight were from social class C and 29 were from class D on a fourpoint scale [7]. When the children were asked how many friends they saw themselves as having, there was a significant difference between the abused and comparison children. Sixteen (43%) of 37 abused children said that they did not have many friends, compared with four (11 W) of the non-abused children admitting this (p < 0.01, Table 1). The result was confirmed when the children answered the question about how often they played with or went out with friends. Fourteen of 37 abused children played with friends less than once each week compared with 6 of 37 comparison children. While 20 (54%) comparison children played with friends daily, this was so for only 7 (19%) of the abused children (p < 0.02, Table 2). When asked about the sort of job they would like to have, only 7 of the 37 abused children nominated occupations in the higher social classes, A or B. This is in contrast with 17 of the comparison children nominating jobs in class A or B (p < 0.05, Table 3). As well as being asked about what sort of job they would like to have, the children were also asked what sort of job they realistically thought they would be able to obtain. Only five of the abused children
161
Self-esteem of abused children Table 2. Freuuency with Which Abused Children Plav or Go Out with Friends Frequency of play with friends
TWiCe weekly
Daily 7
6
WeekfY 10
20
6
5
Abuse group (N = 37) Comparison group (N = 37)
Less than weekly I4 6
x2 = 11.13, df = 3,~ < 0.02.
nominated an occupation in social class A or B compared with 16 comparison children (p < 0.02, Table 4). The large groups in the “don’t know” category in this table represented the younger children in the groups, although there were approximately twice as many abused as non-abused children whose answers were in this category. Average scores for the Piers-Harris Self-Concept Scale are between the 31st and 70th percentile or between raw scores of 46 to 60 ]4]. High scores suggest high self-concept with low scores suggesting the reverse. The comparison children had slightly higher scores than the abused children, The raw score for the abused children (51.4, SD = 13.2) feIl towards the lower end of the normal range, while that of the comparison group (60.9, SD 9.2) was just above the normal range (Table 5).
DISCUSSION A potential problem with a study such as this, where a structured interview is used for a group of children whose ages range from 4.6 to 14.4 years, is that different-aged children are likely to give different responses. Matching each study child with a child of the same age, to within 6 months, has gone some way to ensure that any bias due to the age range is equally reflected in both groups. This study shows that, compared with non-abused children of similar social class, the abused children have fewer friends, lower ambitions and lower selfesteem. The finding that abused children see themselves as having fewer friends than other children and that they play with other children less often suggests that some aspects of the apathetic, withdrawn behavior described at the time of presentation [8,9] and the inability to develop basic trust [lo] persist long after the initial incident. What is not certain is to what Table 3.
Abused CIdidreu’s idea of the Sort of Occupation Tbey Would Like to Have as au AduIt
Social class of job chosen Abuse group (N = 37) Comparison group (N = 37)
C or D
Don’t know
7
19
11
17
14
6
A or 3
x2 = 6.39, df = 2, p < 0.05.
Table 4. Abmed Children’s Idea of the Sort of Occupation They Realistically ‘lMnk They Will Be Able to Have as an Adult Social class of job chosen Abuse group (N = 37) Comparison group (N = 37) x2 = 8.81, df = 2,p < 0.02.
A or B
C or D
Don’t know
5
19
13
16
15
6
162
R. Kim Oates, Douglas Forrest and Anthony Peacock Table 5. Piers-Harris Self-Concept Scale Scores for Abused and Comparison Children Piers-Harris Score
Raw Score
Percentile
Study group (N = 37)
51.4 (SD = 13.2)
47.2 (SD = 28.9)
Comparison group (N = 37)
60.9 (SD = 9.2)
69.7 (SD = 23.2)
F = 12.9, p < .OOl
F = 13.6, p < ,001
Significance
degree these personality characteristics were present before the abuse commenced and whether these features may have put the children at greater risk for abuse. It could be thought that abused children might have quite fanciful and unrealistic employment ambitions, saying that they wanted glamorous and prestigious jobs as a way of escape. However this study showed the opposite. The self-esteem of the abused children was such that they were less likely than the comparison children to want, or to feel that they would be able to obtain, jobs in the higher socioeconomic groups. The Piers-Harris Self-Concept Scale clearly distinguished between the abused and comparison children. This is in contrast to Elmer’s study [3] where no differences were found and to Kinard’s study [5] where there were no significant differences on the raw scores, although after further statistical analysis a difference was found on one of six variables related to selfesteem. In Elmer’s study the group was small, 17 matched pairs, thereby reducing the possibility of detecting differences. Also, the abused children and the matched controls came from quite deprived backgrounds and so are unlikely to be representative of the general spectrum of child abuse. Lynch and Roberts [2] in a discussion of problems associated with control groups, point out that Elmer’s study, which attempted to match for all factors, apart from conflrrned inflicted injuries, may have simply obtained a control group of children who were abused but not actually identified as such. This would explain the lack of difference in selfesteem between the study and control children. Kinard’s group consisted of 30 pairs, carefully matched for age, sex, race, socioeconomic class, neighborhood and family size. Apart from the possibility of differences in matching, the reason for the clear difference between the groups in the present study and the other two studies where the Piers-Harris Self-Concept Scale has been used is not readily apparent. In the present study, the comparison children scored at the upper normal range of the test although they were carefully matched for social class and other appropriate demographic factors. However the results of the present study are consistent with the findings of the interviews held with the children and with the findings of others who have looked at selfconcept without using formal psychological tests [l, 21. It is likely that the high and unrealistic expectations that abusive parents hold for their children, and which their children cannot meet, play an important part in lowering the child’s self-esteem. It is therefore ironic that these high expectations become counterproductive with the children becoming failure-oriented and lacking in ambition and feelings of self-worth. In a review of nine studies published between 1971 and 1978 on the emotional development of children, Kinard [ 1l] concludes that abused children manifest a general air of depression, unhappiness and reduced self-concept.
IMPLICATIONS Although considerable emphasis is placed on supporting abusive parents to prevent further episodes from occurring, it is essential that the future emotional development of the abused
Self-esteem of abused children
163
child should also be emphasized. As the majority of abused children remain in their natural families, they also require treatment while support is being given to their parents. This should first include a careful assessment of the child’s developmental and emotional status followed by a long-term program for the child which aims at improving the child’s skills in interpersonal relationships and in building up self-esteem.
REFERENCES 1. MARTIN, H. P. and BEEZLEY, P. Behavioral observations on abused children. DevelopmentalMedicine and Child Neuroloav19373-378 (1977). 2. LYNCH, M. x. and ROBERTS, j. Consequences of ChildAbe, pp. 98-l 14. Academic Press, London, (1982). 3. ELMER, E. A follow-up study of traumatized children. Pediatticr59:273-279 (1977). 4. PIERS, E. V. The Piers-Harris children’s self-concept scale. Counsellor Recordings and Tests, Nashville, TN (1976). 5. KINARD, E. M. Emotional development in physically abused children. American Journal of Orthopsychiatry 50:686-696 (1980). 6. NIE, N. H. StatisticalPackagefor the Social Sciences. McGraw-Hill, New York (1975). 7. CONGALTON. A. A. Social standing of occuuations in Australia. Studies in Socioloav. -- No. 3. School of Scciology, University bf New South Wales,-Sydney,Australia (1963). 8. GALDSTON, R Observations on children who have been physically abused and their parents. AmericanJournal of Psychiatry12Z440-443 (1965). 9. OWNSTED, C. V. and LINDSAY, J. Aspects of bonding failure. DevelopmentalMedicine and Child Neurologv 16:447-456 (1974). 10. GREEN, A. H. Psychopathology of abused children. Journal of the American Academy of Child Pgxhiatv 17:92-103 (1978). 11. KINARD, E. M. Mental health needs of abused children. Child Werfore59~451-462 (1980).