Child Abuse & Neglect, Vol. t8, No. 3, pp. Z-269, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved 0145”2134/94 $6.00 + .a,
PARENTING STRESS AND SOCIAL SUPPORT OF MOTHERS WHO PHYSICALLY ABUSE THEIR CHILDREN IN HONG KONG Yurc CHUNG Department
Hong Kong Polytechnic,
forEnglish publicatiorz by Jane Gray, M.D., The Kempe
Hung Horn, Kowloon,
1205 Oneida, Denver, CO, USA
Abstract-Thirty-seven identified abusive mothers were matched on demographic and socioeconomic parameters with a known nonabusive comparison sample in order to examine the roie of parenting stress and maternal social support. The mothers were assessed using a personal (demographic) questionnaire, the Parenting Stress Index (PSI), and the Maternal Support Index (MSSI). Demographic data showed that the two groups were comparable on all variables except abusive mothers had significantly more children (p = .Ol). Abusive mothers showed significantly more stress on total PSI scores (p = .005), as well as in all three of the subsets: Child Domain (p = .007) Parent Domain (p = .OZ), and Life Stress (p = ,016). Abusive mothers scored lower in all seven items on the MSSI. The difference was significant on the MSSI as a whole (p = ,007) and on four subsets: number of people to count on in time of need (p = .02), perceived neighborhood support (p = .04), satisfaction with spousal relationship (p = .Ol), and degree of community involvement (p = .03). The greatest percentage (74.32%) of correct predictions of child abuse was achieved by combining the number of children, the Life Stress Scale and the MSSI. Implications for future research are discussed. Key Words-Parenting
stress, Social support,
INTRODUCTION A SIGNIFICANT AMOUNT of research shows that child abuse tends to correlate positively with the amount of stress perpetrators experience (Egeland, Breitenbucher, & Rosenbery, 1980; Newberger, Reed, Daniel, Hyde, & Kotelchuck, 1977; Straus, 1980). Additionally, a sizable amount of research also establishes a negative relationship between the incident of child abuse and the perpetrators’ social support (Fontana, 1968; Garbarino, 1977, 1981; Newberger et al., 1977; Steele & Pollock, 1968). In disentangling the relationship among stress, social support, and child abuse, references are repeatedly made to the stress-buffering role of social support. For example, Bronfenbrenner (1974) theorized that child abuse occurred as a function of the degree to which the human ecology enhanced or undermined parenting. Where the total human ecology provided adequate support, child abuse was minimized; where support was inadequate and the stress great, child abuse would likely occur (Bronfenbrenner, 1974). Garbarino and Sherman (1980) also found that it was the unmanageability of stress-that was the product of a mismatch between the level of stress and the availability and potency of social support that was the most crucial factor in child abuse. Specifically, Pascoe and Earp (1984) suggested that a prime function of social support might be to offer a lifeline to parents under stress. A Received
April 9. 1992; final revision
Requests for reprints should be sent to Yuk Chung Chan, Lecturer, Polytechnic, Hung Horn, Kowloon, Hong Kong. 261
8, 1992; accepted
Social Studies, Hong Kong
Y. C. Chan
lifeline might take the form of emergency child care services, help with inhibition of abusive impulses, reassurance of personal worth, daily task sharing among family members, or the provision of information concerning child behavior and appropriate child rearing practices. This study examines the role of parenting stress and maternal social support in physical child abuse in Hong Kong. Parenting stress is a specific kind of stress. It refers to the stress in the parent-child system as perceived by the parents (Abidin, 1983). On the other hand, maternal social support refers to the social support experienced by a mother in her discharge of mothering and housewife duties. Building on the findings of the previous studies, the present study aims to test the following three hypotheses: 1. Abusive 2. Abusive 3. Abusive
mothers mothers mothers
experience greater parenting stress than nonabusive mothers; experience more life stress than nonabusive mothers in Hong Kong; and have less social support than nonabusive mothers.
The study also intended to identify a discriminant function for abusive and nonabusive mothers using parenting stress and maternal social support as predicting variables.
METHOD Design The study employed a survey research method designed for static-group comparison. Two groups of mothers differing in whether or not they had physically abused their children, but similar in socioeconomic background were studied and compared with respect to their parenting stress and social support. By the design of the present study, respondents in both abuse and nonabuse groups were matched in age, educational level, total family income, and family composition so that the two groups were as similar as possible for comparison purposes. This allowed greater control over the extraneous variables, thereby identifying with greater certainty the relationships between child abuse on one hand and parenting stress and social support on the other. Sutnpling
Fifty mothers who abused their children were randomly sampled from 138 cases from the Child Protective Services Unit of the Social Welfare Department, Hong Kong. These mothers had been identified as perpetrators of physical child abuse during the multi-disciplinary case conferences that are convened according to the procedures for handling child abuse cases in Hong Kong. Of these 50 mothers, 37 consented to being interviewed and they were the subjects of the present study. Three nurseries and one kindergarten from different areas of Hong Kong were chosen to reach nonabusive mothers. After being approached, 66 mothers from these four schools agreed to take part in the study. It was confirmed that none of these mothers had any history of child abuse. Then the computer identified 37 of the mothers whose demographic and socioeconomic background were similar to the abusive mothers. These 37 mothers were the subjects in the Nonabuse Group. Instrument The instrument
for the present
study was a structured
Personal vuriables. Part I of the questionnaire measured the personal and socioeconomic status of the respondents, including their age, marital status, employment status, number and age of their children, their family structure and their total family income.
abuse their children
in Hong Kong
The parenting stress index. The Parenting Stress Index (PSI) was developed by Richard R. Abidin of the University of Virginia in 1977 as a result of his work in the area of parenting education. The PSI has a total of 101 items measuring parenting stress in two domains, namely the Child Domain and the Parent Domain. The Child Domain consists of the following subscales: adaptability, acceptability, demandingness, mood, distractibility/hyperactivity and child reinforces parent. The Parent Domain consists of: depression, attachment, restrictions of role, sense of competence, social isolation, relationship with spouse, and parent health. The PSI also included an optional 19-item Life Stress Subscale.
The maternal social support index. The Maternal Social Support Index (MSSI) was first developed by Pascoe, Loda, Jeffries, and Earp (1981) to quantitate those aspects of a mother’s social support network that may be related to her ability to provide her children with safe, warm, and stimulating environments (Pascoe et al., 1981). Subsequently, the MSSI has repeatedly been utilized in studies assessing social support of mothers (Adamakos, Ryan, & Ullman, 1986; Pascoe, Walsh-Clifford, & Earp, 1982; Pascoe & Earp, 1984). Essentially, the MSSI is composed of scales that measured the mother’s perception of daily task sharing among family members, satisfaction with relationships, availability of emergency help, and degree of community involvement. It has seven items that are broken down into 21 questions. The MSSI was administered to the subjects through standardized interviews. Data collection. Respondents who agreed to participate were interviewed by student interviewers who were social work students in the Hong Kong Polytechnic, one of the nine govemmentfunded tertiary educational institutions with degree-granting courses in Hong Kong. All interviewers had been briefed about the research design and the questionnaire before they formally collected information from respondents. In order to achieve validity of data collection, they were trained to solicit full cooperation from the respondents and to be sensitive to the possibility of misunderstanding and fatigue on the part of the respondents. Respondents in the Abuse Group were interviewed in their own homes while respondents in the Nonabuse Group were interviewed either in their home or in their child’s school. In the beginning of each interview, all the respondents were assured of confidentiality.
Projiles of the Respondents
The mean age of respondents in the Abuse Group was 33.24 years. Seventy-five percent of whom were below 35 years of age. These mothers had an average of 2.92 children. An overwhelming majority of them (92%) lived in nuclear families with an average total income per month of 5,905 Hong Kong Dollars (approximately $740 in the United States). Furthermore, 57% of them had received only a primary education (about Grade 6 in the US) that is low according to the Hong Kong standard. Just over two-thirds (67.6%) of them were housewives, 19% of them had full-time employment and 13.5% were employed part time. As far as comparability is concerned, mothers in the Abuse Group matched those in the Nonabuse Group in terms of their demographic and socioeconomic characteristics except that the former had a significantly larger number of children (Abuse Group = 2.92, Nonabuse Group = 2.35, t = 2.61, p = .01)
Y. C. Chan Table 1. Respondents’
Score on the PSI Abuse Group Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean
Child Domain a. Adaptability b. Acceptability c. Demandingness d. Mood e. Distractability f. Reinforce
Parent Domain a. Depression h. Attachment c
e. Social Isolation f. Spouse Relation g. Health Parenting Stress Excluding (1 + 2 only) Life Stress Parenting (3 + 4)
Stress Index Including
Life Stress Scale
Life Stress Scale
Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD =
= 139.54 19.03 = 31.27 5.82 = 22.49 3.91 = 25.97 5.11 = 14.76 3.29 = 28.18 4.86 = 16.95 4.50 = 174.19 22.61 = 26.95 5.35 = 21.78 3.33 = 24.16 5.44 = 41.89 5.99 = 18.68 3.90 = 24.02 4.45 = 16.70 3.50 = 313.73 35.47 = 9.19 7.04 = 322.92 37.31
Nonabuse Group Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD = Mean SD =
= 126.95 19.92 = 28.89 5.53 = 19.43 4.91 = 25.22 4.85 = 12.97 3.30 = 27.70 4.95 = 12.73 3.33 = 162.95 18.61 = 24.97 5.87 = 19.27 3.50 = 24.00 4.31 = 39.57 6.10 = 17.00 3.92 = 22.16 4.73 = 15.97 2.66 = 289.89 34.52 = 5.57 5.44 = 295.46 35.55
Parenting stress of respondents. Respondents’ parenting stress and life stress are presented in Table 1. The mean score for the Abuse Group on the PSI was 313.73, while that for the Nonabuse Group was 289.89, (t = 2.93, p = .00.5). Respondents in the Abuse Group had significantly greater stress in all three areas, the Child Domain, Parent Domain, and Life Stress than did those in the Nonabuse Group. Parenting stress in the child domain. The mean score for the Abuse Group was 139.54 while that for the Nonabuse Group was 126.95, (t = 2.78, p = .007). Results also showed that the mean scores for the Abuse Group were higher than those of the Nonabuse Group in all six subscales of the Child Characteristics Domain. In three of these six subscales. Child Acceptability, Child Mood, and Child Reinforces Parent, the difference was statistically significant. Parenting stress in the parent domain. The mean score for the Abuse Group while that for the Nonabuse Group was 162.95, (t = 2.34, p = .02). Results also the mean scores for the Abuse Group were higher than those of the Nonabuse seven subscales of the Parent Domain. However, only the difference in the Parent subscale was significant statistically (t = 3.17, p = .002).
was 174.19, showed that Group in all Attachment
abuse their children
in Hong Kong
Table 2. Respondents’
Score on the MSSI
Nonabuse Group N
5.4 2.7 0.0 5.4 10.8 10.8 10.8 18.9 13.5 10.8 2.7 5.4 I 2.7 0 0.0 0 0.0 100.0 37 Mean = 7.405 SD = 2.920 2 1 0 2 4 4 4 7 5 4 1 2
1 2 3 4 5 6 I 8 9 10 11 12 14 15 17
0.0 2 5.4 1 2.1 0 0.0 0 0.0 2 5.4 10.8 4 4 10.8 1 2.1 11 29.7 4 10.8 5 13.5 0 0.0 1 2.7 1 2.7 37 100.0 Mean = 9.378 SD = 3.139
t = -2.80. p = 0.007.
Life stress. The mean score for the Abuse Group in the optional Life Stress Scale was 9.19 while that for the Nonabuse Group was 5.57 (t = 2.48, p = .016), indicating the Abuse Group experienced far more life stress as compared with the Nonabuse Group. MSSZ. The distribution of respondents’ overall scores in the MSSI are presented in Table 2. Respondents’ scores on the individual items of the MSSI are presented in Table 3. As shown in Table 2, the mean of the Abuse Group was 7.4, while that of the Nonabuse Group was 9.38, (t = -2.80, p = .007). Respondents in the Abuse Group scored lower in all seven items of the MSSI when compared with the Nonabuse Group (see Table 3). For the MSSI as a
Table 3. Resuondents’
Score on Individual Items of MSSI
Item Item Item Item Item Item Item
1 2 3 4 5 6 I
Item Item Item Item Item Item Item
1 2 3 4 5 6 7
= = = = = = =
1.95 1.05 1.30 0.41 1.35 1.22 0.14
1.31 0.89 0.10 0.50 0.54 0.89 0.54
2.27 1.08 1.81 0.65 1.68 1.32 0.58
1.10 0.92 0.91 0.48 0.48 0.92 1.04
Perception of daily task sharing among family Satisfaction with relationship with relatives. Number of people count on in time of need. Perceived Neighbourhood support. Satisfaction with spousal relationship. Satisfaction with confidante talks. Degree of community involvement.
t -1.15 -0.13 -2.32 -2.13 -2.75 -0.52 -2.25
P 0.25 0.87 0.02 0.04 0.01 0.61 0.03
Y. C. Chan Table 4. Discriminant Predicting I. 2. 3. 4. 5.
Ability of the Predicting
Factors Percentage of Cases Correctlv Classified
Number of children alone PSI alone Life Stress Scale alone MSSI alone Number of children, PSI, Life Stress Scale and MSSI
60.81% 62.16% 63.51% 64.29% 72.29%
whole and in four of its items (Number of people to count on in time of need, perceived neighborhood support, satisfaction with spousal relationship, and degree of community involvement), the difference was significant. This suggests that the Abuse Group had less social support than the Nonabuse Group.
of the Fuctnrs Associating
with Child Abuse
Discriminant analysis of the factors differentiating the Abuse and Nonabuse Groups are presented in Table 4. It can be seen from Table 4 that predicting child abuse only on the number of children alone results in only 60.8 1% of the cases being correctly identified. Parenting stress alone predicted 62.16% of the cases correctly. Results for Life Stress and Maternal Social Support Index were 63.51% and 64.29% respectively. When the four variables were taken together, the correct prediction rate was 72.29%. A stepwise analysis for a discriminant model based on these four factors showed that the inclusion of the PSI made no difference in the significance of the model. Hence another discriminant analysis using Number of Children, Life Stress Scale and the MSSI as predictors for Child Abuse was done. The percentage of correct identification for this model increased to 74.32%. This model and its significance were presented in Tables 5 and 6.
Basically, the results are consistent with what has been suggested in the literature. These abusive mothers were relatively young and poorly educated. They thus probably lack parenting competence. They had more children to parent and were most likely to be living in a nuclear family. Under the circumstances, they were likely to have difficulty coping with their parental
Table 5. Stepwise Analysis
of the Discriminant Model Using PSI, MSSI, Life Stress Scale, and the Number Children as Predictors Action
I 2 3 4 5
PSI MSSI Life Stress Scale Number of Children
Variables in the Model
I 2 3 4 3
0.8955 0.844 I 0.7857 0.7232 0.7279
0.0050 0.0024 0.0007 0.0001 0.0001
abuse their children
in Hong Kong
Model Using Number of Children, Life Stress Scale, and MSSI as Predictors for Child Abuse
Unstandardized Canonical Discriminant Coefficient
Number of Children Life Stress Scale MSSI
0.6755 0.1009 -0.2287
Wilk’s Lambda 0.9138 0.92 15 0.9005
6.790 6.133 7.954
0.0110 0.0156 0.0062
role. Their lower socioeconomic status and their vulnerability to the recurring socioeconomic stress left little doubt that these mothers would be an at-risk group for abuse. Abusive mothers had a mean score on the Parenting Stress Index of 313.73, with 94% scoring 261 or above. According to Abidin (1983), parents scoring 260 or above should be referred for professional consultation and intervention. The fact that these mothers had all abused their children and were receiving professional attention suggests that the PSI is a valid and reliable tool for screening problems in the parent-child system. The study’s hypothesis, that abusive mothers had a significantly greater amount of parenting stress than nonabusive mothers, was confirmed. This corresponds with the results of a previous study that parenting stress was a significantly discriminant factor for child abuse (Belsky, 1980; Mash, Johnston, & Kovitz, 1983). The second hypothesis, that mothers in the Abuse Group had greater life stress than mothers in the Nonabusive Group, was also confirmed. Mothers in the Abuse Group had an average score of 9.19 on the optional Life Stress Scale of the PSI, while in the Nonabuse Group averaged 5.57 on the same scale. In other words, abusive mothers are in general experiencing more life stress than their nonabusive counterparts. Again, this is consistent with findings of previous studies (Egeland, Breitenbucher, & Rosenberg, 1980; Howze & Ketch, 1984). The third hypothesis, that the mothers in the Abuse Group had significantly less social support as compared with the mothers in the Nonabuse Group, was also confirmed. Abusive mothers scored lower on all dimensions of MSSI than nonabusive mothers. In four of these dimensions, namely, emergency help, neighborhood support, spousal support and degree of community involvement, the difference was significant. Again, this result agrees with previous research findings (Garbarino, 1977; Polansky, Chalmers, Buttenwieser, & William, 1979; Pascoe et al., 1982). A significant positive relationship was found between the number of children and the amount of parenting stress. The relationship was particularly strong between the number of children and stress in the Child Characteristics Domain. This is an interesting finding, yet it needs to be interpreted with caution. Possibly, an increased number of children is a causative factor in parenting stress. However, this relationship may be due to another factor, and needs to be further investigated. In the absence of other information, the present study found that scores on the MSSI are the best single predictor for child abuse. It correctly identified 64% of the cases. This was followed by the optional Life Stress Scale (63.57%), then by the Parenting Stress Index (62.16%), and lastly by the number of children (60.81%). As mentioned earlier in this paper, the Parenting Stress Index was not a significant discriminating factor when taken together with the number of children, the optional Life Stress Scale and the Maternal Social Support Index. A discriminant model incorporating the latter three factors correctly identifies 74.32% of the child abuse cases. This result possibly suggests that there may be some interesting interacting effects between the Parenting Stress Index with the other three variables, and further research may unveil their relationships.
Y. C. Chan
CONCLUSIONS The present study suggests that prediction of mothers who may abuse their children may be possible by paying attention to significant factors such as the number of children they have, their parenting stress, their life stress and their social support. Early identification of these at risk mothers allows for timely intervention and service to them and their family that will hopefully prevent child abuse. In connection with issues associated with early identification, the PSI alone was only a modest tool for predicting child abuse (62.16%). However, as an instrument for identification of stressful parent-child relationship and a screening test for referral for professional consultation, it was both a valid and reliable instrument. As the present study has shown, it could have identified and referred 95% of those in the Abuse Group for professional consultation and intervention. In view of the PSI’s strong diagnostic value, it is recommended that this instrument be considered as one for identifying and screening for stressful parent-child relationships. Hopefully, this will make early identification possible, thus helping to prevent child abuse. As far as parenting stress is concerned, the difference between the Abuse and the Nonabuse groups is more significant in the Child Characteristics Domain than that in the Parent Characteristics Domain. This possibly suggests that victims of child abuse may in general be more difficult to handle. In view of this, it is suggested that more training in parenting skills may be necessary to help strengthening parenting ability and coping skills. Results of the present study correspond with the findings of the research from other cultures in that the number of children is a significant factor for child abuse. If fewer children in a family is associated with nonabuse, it seems that more public education on family planning is important. Particularly, efforts should be focused on population sectors such as families from the lower socioeconomic strata of our society, the new immigrants from China, and families who are likely to have a large number of children. Definitely, the government and the voluntary sector, the Hong Kong Family Planning Association in particular, will have an important part to play. Needless to say, mere intervention in the parent-child system is not sufficient. Though not being a panacea for all ills, social support in a number of aspects is important in preventing child abuse. Hence, intervention in the ecological environment of the abusers is necessary. Specifically, more attention should be directed towards fostering a more favorable spousal relationship, encouraging neighborhood support, and creating a social network in which emergency and community support is available. Hopefully, this will help to ease parental stress.
Abidin, R. R. (1983). Parenting Stress Index (Form #6). Charlottesville, NC: Pediatric Psychology Press. Adamakos, H., Ryan, K., & Ullman, D. G. (1986). Maternal social support as a predictor of mother-child stress and stimulation. Child Abuse & Neglect, 10,463-470. Belsky, J. (1980). Children maltreatment: An ecological integration. American Psychologist, 35(4), 320-335. Bronfenbrenner, U. (1974). Developmental research. public policy, and the ecology of childhood. Child Development, 45, 1-5. Egeland, B., Breitenbucher, M., & Rosenbery, D. (1980). Prospective study of the significance of life stress in the etiology of child abuse. Journal ofCounselling and Clinical Psychology, 48(3), 195-205. Fontana, V. J. (1968). Further reflections on maltreatment of children. New York State Journal ofMedicine, Vo1.68. pp. 2214-2215. Garbarino, J. (1977).The human ecology of child maltreatment: A conceptual model for research. Journal ofMarriage and the Family, 33, 721-735. Garbarino, J. (1981). An ecological approach to child maltreatment. In L. H. Pelton (Ed.), The social context of child abuse and neglect. (pp.228-267). New York: Human Sciences Press.
abuse their children
in Hong Kong
Garbarino, J. & Sherman, D. (1980). High-risk neighbourhoods and high-risk families: The human ecology of child maltreatment. Child Developmenr, 51, 188 198. Howze, D. C. & Ketch, J. B. (1984) Disentangling life events, stress and social support: Implications for the primary prevention of child abuse and neglect. Child Abuse & Neglecf, 8, 401-409. Mash, E. J., Johnston, C., & Kovitz, K. (1983). A comparison of the mother-child interactions of physical abused and nonabused children during play and task situations. Journal of Clinical Child Psychology, 12, 337-346. Newberger, E., Reed, R. B., Daniel, J. H., Hyde, J. N., & Kotelchuck, M. (1977). Pediatric social illness: Toward an etiological classification. Pediatrics, 60, l78- 185. Pascoe, J. M., Loda, F. A., Jeffties, V., & Earp, J. A. (1981). The association between mothers’ social support and provision of stimulation to their children. Developmental and Behavioral Pediatrics, 2(l), l5- 19. Pascoe, J. M., & Earp, J. A. (1984). The effect of mother’s social support and life changes on the stimulation of their children in the home. American Journal of Public Healrh, 74, 358-360. Pascoe, J. M., Walsh-Clifford, N., & Earp, J. A. (1982). Construct validity of a maternal social support scale. Developmental
N. A., Chalmers,
M. A., Buttenwieser,
E., & William, D. P. (1979). Isolation of the neglectful families. American Journal of Orthopsychiatry. 49, l49- 152. Steele, B., & Pollock, C. (1968). A psychiatric study of parents who abuse infants and small children. In R. Helfer and C. Kempe (Eds.), The battered child (pp.89- 134). Chicago, IL: University of Chicago Press. Straus, M. A. (1980). Stress and physical child abuse. Child Abuse & Neglect, 4, 75-88.
RCsum&Un bon nombre de recherches indiquent que les mauvais traitements tendent a se produire lorsque les agresseurs vivent des experiences stressantes. Par ailleurs, bon nombre de recherches suggerent que le manque de supports sociaux contribuent aussi 1 la maltraitance. On fait souvent mention du fait que les supports sociaux servent a tamponner les situations de stress. Par example, Bronfenbrenner propose une theorie de l’ecologie humaine oti la maltraitance survient lorsque le milieu social mine le role du parent. Lorsque l’environnement en entier foumit les supports adtquats, les mauvais traitements diminuent; a l’inverse, lorsque les supports sont inadtquats et le stress est grand, il est probable qu’il y aura maltraitance. Garbarino et Sherman ont aussi dtcouvert qu’il devient difficile de toltrer le stress lorsqu’il y a un desequilibre entre le niveau de stress et la qualitt du support social. Plus particulierement, Pascoe et Earp proposent que le role principal de l’appui social serait celui d’offrir aux parents une boute de sauvetage lorsque ceux-ci vivent des periodes de stress, c-8-d par le biais de services de garde d’enfant sur demande, en les aidant a controler leurs tendances agressives, en les rassurant sur leur valeur en tant que personne, en encourageant le partage des tlches au foyer ou en fournissant des connaissances sur le comportemente en enfance et sur les bonnes pratiques d’education des enfants. Resumen-Treinta y siete madres abusivas identificadas fueron comparadas con otras semejantes en parametros demogdficos y socioeconomicos de una muestra de madres no abusivas para estudiar el papel de1 stress parental y el apoyo social matemo. Las madres fueron evaluadas utilizando un cuestionario personal (demografico), el “Parenting Stress Index” (PSI), y el “Maternal Stress Index” (MSSI). Los datos demograficos demostraron que 10s dos grupos eran comparables en todas las variables except0 en que las madres abusivas tenian signiticativamente mas niiios (p = .Ol). Las madres abusivas mostraron significativamente mas stress en 10s puntajes totales de1 PSI (p = .005), asi coma en las tres subpruebas: Dominio Infantil (p = ,007) Dominio Parental (p = .02), y Stress de Vida (p = ,016). Las madres abusivas obtuvieron puntajes mas bajos en 10s siete items de1 MSSI. La diferencia fue significativa en el MSSI coma un todo (p = .007) y en las cuatro subpruebas; nlimero de personas en quien contar en momentos de necesidad (p =.02), apoyo percibido de 10s vecinos (p = .04), satisfaction en la relation marital (p = .Ol), y grado de compromise de la comunidad (p = .03). El porciento mayor (74.32%) de predicciones correctas de abuso a 10s nitios se obtuvo al combinar el nlimero de nitios, la Escala de Stress de Vida y el MSSI. Se discuten las implicaciones para investigaciones futuras.