Social Support and Personal Agency in At-Risk Mothers

Apoyo Social y Agencia Personal en Madres en Situación de Riesgo

María José Rodrigo y Sonia Byrne

Universidad de La Laguna, Spain

We would like to express our appreciation to all the social agents of the municipalities of the Autonomous Community of Castilla y León, Spain, and all the mothers who participated in this study. The implementation and evaluation of the parenting program, "Apoyo Personal y Familiar", for at-risk parents, was supported by the Junta de Castilla-León, the ECCA Foundation and the University of La Laguna, Spain. The study was supported by the Spanish Ministerio de Ciencia e Innovación (TRA2009_0145) to the first author and a research grant to the second author.

Correspondence

ABSTRACT

This study investigated: a) mothers' use and satisfaction with informal and formal supports in at-risk psychosocial contexts, and b) the relationships between satisfaction with help and the mothers' perception of their role (personal agency). Self-report data about the use and satisfaction with sources of help, and levels of internal control, self-efficacy, couple agreement, role difficulty and motivation for change were obtained from 519 mothers referred by Social Services and 519 non-referred mothers. Results indicated that at-risk mothers relied less upon close informal support and more on formal support than non at-risk mothers. They were also more satisfied with the formal sources of support and had lower levels of personal agency. There were beneficial effects of satisfaction with informal help and school support on several aspects of personal agency for both groups. However, satisfaction with school and social services support had a detrimental effect on couple agreement in the at-risk group. Implications of the results for providing social support to at-risk families are discussed.

Keywords: at-risk families, informal and formal use and source satisfaction, personal agency, social support.

RESUMEN

El presente estudio investiga: a) el uso y la satisfacción con las fuentes de apoyo informal y formal de las madres en contextos de riesgo psicosocial y b) la relación existente entre la satisfacción con el apoyo social y la percepción del rol parental (agencia personal). Se obtuvieron datos con respecto al uso y la satisfacción con las fuentes de apoyo social, así como los niveles de control interno, autoeficacia, acuerdo en la pareja, dificultad del rol y motivación para el cambio, de 519 madres referidas por los Servicios Sociales y 519 madres no referidas. Los resultados indicaron que las madres en situación de riesgo psicosocial dependen menos de los apoyos informales y más de los formales que las madres en situación de no riesgo. Asimismo las madres en riesgo se encontraban más satisfechas con las fuentes de apoyo formal y presentaban niveles más bajos de agencia personal. Se obtuvieron efectos beneficiosos de la satisfacción con las fuentes informales de apoyo y el apoyo de la escuela en diversos aspectos de la agencia personal en ambos grupos. Sin embargo, la satisfacción con el apoyo proporcionado por la escuela y los servicios sociales tuvieron un efecto perjudicial en el acuerdo en la pareja en el grupo en situación de riesgo psicosocial. Se discuten las implicaciones de estos resultados con respecto a la provisión de apoyo social en las familias en situación de riesgo psicosocial.

Palabras clave: agencia personal, apoyo social, familias en riesgo, uso y satisfacción con el apoyo informal y formal.

An interesting topic of research is the social network used by families described by social services as living in at-risk psychosocial conditions (e.g., marital violence, low educational background, poverty, substance abuse). Families receiving social assistance are usually multi-assisted, since the various problems they live with attract many forms of formal support provided by the social agencies (Matos & Sousa, 2004). Thus, they tend to interact with many help providers and are very good at seeking support from a variety of agencies. Likewise, it is likely that they interact with some members of the immediate family, the extended family and the community. This study investigated two questions. First, the use of informal and formal support and the level of satisfaction with the sources of help exhibited by at-risk mothers in comparison with those used by non at-risk mothers. Second, the relationships between satisfaction with certain sources of help and the mothers' perception of their role, that is, personal agency, according to their risk status.

Social networks of at-risk parents

Identification, development and protection of sources of social support for at-risk families are key tasks for social workers and other welfare professionals (Jack, 2000). Thus, it is important to know the particular use of informal and formal sources of help in at-risk families since they may have different impacts on family functioning. Informal support may have a positive impact as they consist of a rich system of private exchanges embedded in a natural framework of mutual assistance and obligations in the primary group, the neighborhood and the community. Individuals can be providers as well as recipients of help and this reciprocity ensures feelings of mutual respect that contribute to relational satisfaction (Gottlieb, 1983; 2000). By contrast, service delivery in formal support is embedded in a framework of unidirectional exchanges, guided by protocols and applied by experts in accordance with standards of quality. When help giving is unidirectional, it can make the recipient feel inferior and vulnerable, and this can quickly undermine the helping relationship (Shumaker & Brownell, 1984). Receiving help can be humiliating and stigmatizing, especially when the need for assistance derives from inadequacies in the recipient (such as poor parenting skills, substance abuse, or inadequate personal or financial management). Therefore, it is important not to rely exclusively upon formal sources of help to avoid recipients having feelings of inadequacy.

In a recent study, Rodrigo, Martín, Máiquez, and Rodríguez (2007) examined the differences between the social networks called on to solve problems of at-risk and non at-risk mothers referred by the Social Services. Results indicated that, comparatively with the at-risk mothers, the social support for non at-risk families relied heavily on the partner and the school. At-risk families have to open their personal and family space to the external sources of support (social services, voluntary associations, neighbors, friends, police, other supporting people and the child protection agency), which makes the well-being of vulnerable families more dependent on the quality of the neighborhood (Garbarino & Kostelny, 1992). They are also more in danger of suffering the negative consequences of a poor coordination among different agencies and social providers (Matos & Sousa, 2004). The existence of multiple help providers from outside the natural networks may also interfere in the everyday life of the family, diminishing the sense of privacy its member have and promoting a lack of responsibility for their own lives (Evans & Harris, 2004). As at-risk mothers show a greater reliance on formal support than non at-risk mothers, this would be potentially damaging to perceptions of their role.

Parental social support and personal agency

According to Azar (1998), effective parenting involves a set of personal competencies related to the performing of the parenting task. One of these competences is the parental perception of their role, or personal agency. Personal agency encompasses concepts such as Control and Self-efficacy (Turner & Roszell, 1994). Agency is improved when parents think they are able to take control of their lives (Internal Control) by acting upon everyday events with the potential goal of changing or keeping things the same (a mixture of internal locus of control: Koeske & Koeske, 1992; and coping skills: Folkman & Lazarus, 1988). Parents who feel less in control, show depression and less effective interactions with their children than parents with higher levels of control over their lives (Donovan & Leavitt, 1989). Personal agency is also improved by increasing a sense of Self-efficacy, defined as an individual judgment about how well a person can carry out the necessary steps to deal with a specific task or challenge (Bandura, 1982; 1997). Women who feel more competent in their role as mothers have been found to parent in ways that promote developmental opportunities and minimize risk (e.g., Elder, Eccles, Ardelt, & Lord, 1995). Parents with low self-efficacy may struggle to use positive parenting strategies and give up easily when challenges arise (Ardelt & Eccles, 2001). Increasing Couple Agreement on child education is also an important aspect of parental competence (e.g., the "co-parenting alliance" suggested by McHale & Cowan, 1996). Agency is also improved when parents have a better understanding of what it takes to be a parent, or perceived Role Difficulty (McLoyd, Jayaratne, Ceballo, & Borquez, 1994). Finally, parental Motivation for change refers to their readiness to participate and collaborate in activities with the social agents aiming to improve their functioning (Littell & Girvin, 2005).

Researchers and practitioners are only starting to discover the influence of different sources of help on parental functioning in at-risk families. For example, it is not yet evident what kind of help would best promote personal agency in those families. As mentioned before, it could be the case that satisfaction with a formal source of help may undermine the parental feelings of adequacy and confidence in their role (Doherty & Beaton, 2000). Among the studies which have shown effects of social support on parenting according to the risk status two compelling hypotheses seem to be at work: the buffering and the depleting hypotheses that lead to opposite results. The buffering hypothesis suggests that support has greater effects among individuals currently facing stressors and challenges, although supportive functions are not irrelevant for generally non-stressed populations as well (Cochran & Niego, 1995; Cohen & Wills, 1985; Cutrona, 1996). Social resources may reduce the parental stress experienced through living under negative conditions by contributing to strategies for coping with stressors (Schwarzer & Knoll, 2007; Wills & Shinar, 2000). Among high-risk parents, social support appears to contribute to more maternal enjoyment of the child, and to greater parental self-efficacy (Corse, Schmid, & Trickett, 1990). In turn, mothers from disadvantaged communities or environments with a lack of social support are unlikely to report high levels of parenting self-efficacy (Coleman & Karraker, 1997; Raver & Leadbeater, 1999). Less mobilization of spousal and social support has been related to couple distress and higher levels of conflicts and disagreements (Beach, Fincham, Katz, & Bradbury, 1996). Mothers with larger and more supportive networks perceive the tasks of helping their child develop socially and cognitively to be less difficult than mothers with smaller and less supportive networks (Melson, Ladd, & Hsu, 1993).

Our study tries to shed light on these hypotheses. According to the buffering hypothesis, social support might have greater or equal benefits on personal agency in the at-risk group of mothers as compared to the non at-risk group. According to the depleting hypothesis, social support might have less benefit for personal agency in the at-risk group of mothers as compared to the non at-risk group. Moreover, we want to know whether the source of support (informal or formal) also matters. For instance, it might be that the buffering hypothesis may hold for informal sources but not for formal sources.

Method

Participants

Participants were 1038 mothers, 519 referred by the municipal social services and 519 non-referred mothers living in the Autonomous Community of Castilla-León, Spain. Both groups were recruited while participating in a parenting program called "Apoyo Personal y Familiar" (APF, Family and Personal Support). The program is aimed at increasing parental and personal competence, in order to improve the autonomous functioning of poorly-educated parents at psychosocial risk (Rodrigo, Correa, Máiquez, Martín, & Rodríguez, 2006). APF is implemented by the social services personnel through weekly group meetings in community centres and is offered to families with and without risk indicators. Parents at risk were particularly invited to participate whereas non at-risk parents attend the program on a more voluntarily basis. The referral criterion of the at-risk group was having an at-risk minor. A minor is considered to be at-risk when several psychosocial family and personal factors indicate a situation that is potentially harmful to his or her healthy development. Social services personnel in each municipality (psychologists, social workers and educators) specially trained for working with at-risk families were requested to perform this evaluation. Protocols for the assessment of risk to the minor have been designed to provide a homogeneous criterion for the definition of risk in two-parent and one-parent families (Rodríguez, Camacho, Rodrigo, Martín & Máiquez, 2006). The protocols analysed 42 indicators with a yes/no scoring in the following areas: sociodemographic factors (8 items), family social network (2 items), family organization and household conditions (7 items), caregiver's history of maltreatment and personal characteristics (7 items), quality of relationships within the family (7 items), inadequate child-rearing practices (5 items), and child adjustment problems (6 items). Mothers in the at-risk group had an average of 17.5 risk indicators with a 6.23 standard deviation.

Socio-demographic characteristics of the mothers who participated in this study are presented in Table 1 by risk status. Mothers ranged in age from 17 to 66 years old, with an average of 37.1 (SD = 6.9). Mothers had an average of 1.9 (SD = 0.8) children, ranging from 0 to 9 children. At-risk mothers in comparison to non at-risk mothers were more likely to be younger than 28 years old, and live in urban areas. Likewise at-risk mothers tended to live more in one-parent families, had a lower level of education, were jobless and on welfare more than non at-risk mothers. At-risk mothers also had a partner with a lower level of education and jobless.

Measures

Scale of Personal and Social Support (Escala de Apoyo Personal y Social), adapted from the Scale of Social Support in the Informal and Formal Systems (Escala de apoyo social en los sistemas informales y formales) by Gracia, Herrero, and Musitu (2002), and the Social Support Questionnaire (SSQ; Sarason, Levine, Basham, & Sarason, 1983). By means of eight questions, four when the mother has a problem with her child and four when she has a personal problem, it explores the sources of informal and formal support. In each case, the respondent is first asked a general question (yes/no answer): Do you seek personal support when you are facing a child problem? Second, the respondent is provided with a list of potential sources of informal support: partner, older sibling, father, mother, brother/sister, aunt/uncle, friend, neighbor, others, and she has to mark the ones she uses. Personal sources are named from the point of view of the mother. There is no limitation to sources of support or to the number of supporters who can be listed. Third, the respondent is asked a general question about seeking sources of support from the institutions when they face a child problem (yes/no answer). Fourth, a list of options is given: school, social services, parochial support, Caritas, police, neighbourhood association, child protection agency, and other institutions with similar purposes. The same sequence of four items is given for when the mother has to solve a personal problem. Ratings of support satisfaction (1-5 rating scale) are obtained for each of the informal and formal sources of support actually used in each situation.

Parental Questionnaire on Personal Agency (Máiquez, Rodrigo, Capote, & Vermaes, 2000) includes five scales (7-point agreement scale): Internal Control (5 items; Cronbach alpha reliability = .82; sample item: "I think that things can be changed if I try"); Self-efficacy (4 items; Cronbach alpha reliability = .80; sample item: "I believe in my capacity to solve family problems"); Couple Agreement (2 items; Cronbach alpha reliability = .75; sample items: "We have reached consensus about educational matters"); Role Difficulty (5 items; Cronbach alpha reliability = .84; sample item: "The task of parenting implies much effort"); and Motivation for change (3 items; Cronbach alpha reliability = .72; sample item: "I think that this meeting could give me more information about educational matters"). We computed five average scores for each mother. A high score in internal control means that a mother endorses the idea that she has control over her life and that things can be changed. A high score in self-efficacy means that a mother believes in her own capacity to solve family problems and has confidence in her role. A high score in marital agreement means that a mother talks to her partner often about child education and tries to reach consensus about important family matters. A high score in role difficulty means that a mother thinks that the task of parenting is very complicated, time consuming and requires much effort. A high score in motivation for change means that a mother thinks that she is ready to improve her educational practices and to make the most of a parenting program.

Procedure

Prior to the initial session of the parenting program, the social services personnel in each municipality who were in charge of the referred mothers filled in the protocol of risk assessment. For the non-referred attendees no protocol was used. Risk and non at-risk mothers were similarly distributed among the provinces. The groups were formed following the criteria of mixing risk status to facilitate the exchange of a variety of parenting experiences during the program sessions. The first session of the parenting program was used to fill in the two questionnaires by the mothers. The sample of at-risk and non at-risk mothers was drawn from 148 groups (about seven people per group) monitored by 73 social agents (about one monitor for every two groups), but the conditions of administration of the questionnaires were the same.

Results

Differences in the use of social networks and support satisfaction according to the risk status

Responses to the four general questions concerning the use of informal and formal support when facing either child problems or personal problems were averaged for the non at-risk and at-risk groups (see Figure 1). An ANOVA with risk status as a between-subjects variable, and repeated measures for social network (informal/formal) and type of problem (child/personal) were all performed on the use of social support. A main effect was obtained for the risk status, F(1, 1036) = 30.8, p < .001, indicating that at-risk mothers sought support significantly more (72%) than non-at risk mothers (62%). Seeking support in the informal network was more frequent (88%) than in the formal network (46%) in both groups, F(1, 1036) = 928.4, p < .001. Seeking support also differed according to the type of problem in both groups, more for child problems (73%) than for personal problems (61%), F(1, 1036) = 159, p < .001. However, patterns differed according to the risk status. The two groups did not differ in their use of informal help (88%, 87%) but at-risk mothers used more formal help (57%) than non at-risk mothers (35%), F(1, 1036) = 73.1, p < .001. In turn, atrisk mothers looked for help, both for child problems (76%) and for personal problems (68%), whereas the non at-risk mothers tended to use help mostly for child problems (69%) and much less for personal problems (44%), F(1, 1036) = 14.6, p < .001. Even after adding the socio-demographic variables, the previously significant findings were still significant, indicating that this pattern of results was quite stable.

Overall, satisfaction with the support source varied according to the risk status (Tables 2 & 3). Satisfaction with the partner (child problem: F(1, 807) = 17.52, p < .001; personal problem: F(1, 756) = 11.92, p < .001) and mother supports (personal problem: F(1, 487) = 5.53, p < .001) were higher in the non at-risk than in at-risk mothers. Satisfaction with the social services (child problem: F(1, 298) = 8.78, p < .001; personal problem: F(1, 254) = 6.23, p < .001), Caritas (child problem: F(1, 57) = 5.25, p < .05), and neighbourhood association supports (personal problem: F(1, 35) = 4.97, p < .05) were higher in the at-risk mothers. These results were not accounted for by socio-demographic differences.

Relations between support satisfaction and personal agency

Means and standard deviations of personal agency were calculated by risk status. One-way ANOVAs were performed for the group comparison and results are shown in Figure 2. As expected, perceived personal agency was higher across all the factors in the non at-risk mothers than in the at-risk group: internal control, F(1, 1036) = 13.97, p < .001, self-efficacy, F(1, 1036) = 31.19, p < .001, couple agreement, F(1, 1036) = 124.13, p < .001, perception of parental role difficulty, F(1, 1036) = 8.21, p < .01, and motivation for change F(1, 1036) = 34.79, p < .001. When socio-demographic variables were entered in the analyses as covariables, only the level of education yielded significant results though risk status remained significant. Mothers with higher levels of education had more internal control, F(1, 967) = 11.62, p < .001, perceived more parental role difficulty, F(1, 967) = 7.026, p < .001, and had more motivation for change, F(1, 967) = 4.16, p < .001, than mothers with lower levels.

To highlight the possible evidence of risk status moderation, regression analyses were performed using the level of education, support satisfaction with each informal and formal source averaged across type of problems, and support satisfaction x risk status as predictors of the factors of personal agency. In this way, we would see whether higher or lower levels of personal agency are predicted either by support satisfaction plus the interaction term (weak moderation effect) or just by the interaction term (strong moderation effect). Only significant results and standardized beta are reported.

Concerning internal control, mothers' satisfaction with partner, F(3, 761) = 11.19, p < .001; adjusted R2 = .07, mother, F(3, 486) = 4.78, p < .01; adjusted R2 = .04, brother/sister, F(3, 431) = 5.22, p < .001; adjusted R2 = .04, and school supports, F(3, 489) = 3.52, p < .01; adjusted R2 = .04, significantly predicted the level of internal control. Higher levels of education significantly predicted more internal control when considering satisfaction with partner, β = .167, p < .001, mother, β = .095, p < .05, brother/sister, β = .147, p < .01, and school supports, β = .171, p < .05. Higher satisfaction for partner, β = .140, p < .01, mother, β = .166, p < .05, brother/sister, β = .206, p < .05, and school supports, β = .185, p < .01, predicted higher levels of internal control. However, the interaction term was not significant with any source of support, indicating a direct link between support satisfaction and higher levels of internal control whatever the risk status.

Risk status played a more prominent role in the influence of social support on couple agreement for educational matters. Mothers' higher satisfaction with partner, mother, brother/sister, school and social services support significantly predicted higher levels of couple agreement, even when controlling the effects of level of education (Table 4). The interaction term was not significant for mother support, indicating a direct link with higher levels of couple agreement whatever the risk status. However, for the rest of the sources of help the interaction term significantly predicted the decreasing level of couple agreement, showing evidence of the weak (couple, brother/sister and social services) and strong (school) moderation effects of risk status.

To illustrate the above interactive effects, high and low levels of satisfaction of each source, defined by a mean split, and the mean reported level of couple agreement were plotted for the non at-risk and at-risk groups. ANCOVAs were calculated using risk status and high/low levels of satisfaction as between-subject variables and couple agreement as a dependent variable, and controlling the level of education as a covariate. Interactive effects were significant for partner, F(1, 805) = 6.017, p < .01, brother/sister, F(1, 427) = 3.910, p < .05, school, F(1, 168) = 4.464, p < .05, and social services support, F(1, 296) = 5.542, p < .01. Atrisk mothers who were highly satisfied with couple support reported the highest level of couple agreement, and those with low satisfaction reported the lowest level of couple agreement. The pattern was similar but not so extreme in non at-risk mothers. At-risk mothers' level of couple agreement was not affected by the satisfaction with brother/sister support. Non at-risk mothers who were highly satisfied with brother/sister support reported the highest level of couple agreement, and those who had low satisfaction reported the lowest level. However, for the two formal sources the pattern strongly differed by risk status. At-risk mothers who were highly satisfied with school and social services support reported the lowest level of couple agreement, and those with low satisfaction reported the highest level of couple agreement. The results were totally reversed for both sources in non at-risk mothers, that is, those mothers who were highly satisfied with school and social services support reported the highest level of couple agreement (see Figure 3).

Concerning the perception of parental role difficulty, mothers' satisfaction with father, F(3, 145) = 3.5, p < .01; adjusted R2 = .05, mother, F(3, 459) = 2.77, p < .05; adjusted R2 = .04, and school supports, F(3, 489) = 7.298, p < .001; adjusted R2 = .07, predicted the level of parental role difficulty. Higher levels of education significantly predicted less perceived difficulty in the parental role when considering satisfaction with school, β = -.186, p < .001. Higher levels of support satisfaction with father, β = -.260, p < .05, and mother, β = -.156, p < .05, also predicted less perceived difficulty in the parental role. The interaction term was not significant for father and mother indicating a direct link of satisfaction with both sources and lower levels of perceived role difficulty whatever the risk status. However, the interaction term significantly predicted higher levels of parental role difficulty, F(1, 168) = 5.35, p < .05, showing evidence of a strong moderation effect of risk status when considering satisfaction with school, β = .133, p < .05. At-risk mothers who were highly satisfied with school support reported the highest level of parental role difficulty, and those who with low satisfaction reported the lowest level of parental role difficulty. The reverse was found for non at-risk mothers (see Figure 4).

As for motivation for change, mothers' satisfaction with partner, F(3, 761) = 11.19, p < .001; adjusted R2 = .07, and school support, F(3, 489) = 3.169, p < .024; adjusted R2 = .04, significantly predicted the level of motivation for change. Higher levels of education significantly predicted more motivation for change considering satisfaction with partner, β = .167, p < .001. Higher levels of satisfaction for partner support β = .140, p < .01, predicted higher levels of motivation for change. The interaction term was not significant for partner support, indicating a direct link with higher levels of motivation for change whatever the risk status. However, the interaction term significantly predicted higher levels of motivation for change showing evidence of a strong moderation effect of risk status when considering satisfaction with school, β = .176, p < .05. A significant effect for school support was found, F(1, 485) = 6.739, p < .01, indicating that at-risk mothers who were highly satisfied with school support reported higher levels of motivation for change, and those with low satisfaction reported lower levels of motivation for change. Non at-risk mothers' motivation for change was independent of the satisfaction with school support (see Figure 4).

Discussion

Concerning the first research question, all groups sought more support in the informal than in the formal sources, whatever the type of problem. Going to the proximal sources for help, where people can find natural, reciprocal and more satisfactory support is a general tendency (Gottlieb, 1983; 2000). However, as expected, at-risk mothers look for more formal help than non at-risk mothers, given that these mothers belong to multi-assisted families (Matos & Sousa, 2004). Both groups sought more support when facing child problems than when facing personal problems. In particular, at-risk mothers looked for help both for solving child problems and personal problems, whereas the non at-risk mothers tended to use help mostly for child problems (Rodrigo et al., 2007). These results are not surprising given that vulnerable mothers who turn to social services lack personal resources to solve their everyday problems by themselves (Azar, 1998).

The use of informal and formal sources of help differed according to risk status, regardless of the type of problem. The social support of non at-risk mothers relied comparatively more on the partner and on the school, both of which are resources available in the micro-system (Rodrigo et al., 2007). By contrast, at-risk mothers relied comparatively less upon couple support and more on less close individuals (brother/ sister and others) and formal sources of help such as social services, Caritas and Police. The formal sources are mainly in the exo-system, meaning that the at-risk mothers' well-being is heavily dependent on the quality of the resources in the neighborhood (Garbarino & Kostelny, 1992) and on the quality of the help providers from the social agencies (Matos & Sousa, 2004; Evans & Harris, 2004).

Rates of support satisfaction for each source used also differ by risk status. Satisfaction with partner and mother support was higher in the non at-risk mothers than in at-risk mothers, whereas satisfaction with social services, Caritas and neighbourhood association were higher in the at-risk mothers. At-risk mothers might fail to cultivate mutually supportive relationships with the partner and mother and these working models of attachment may shape their low satisfaction perceptions with those sources of support (Collins & Feeney, 2000). In turn, their negative psychosocial risk conditions lead them to ask for assistance from institutions in many situations, which may increase their level of satisfaction with the help providers as a sign of personal gratitude (Matos & Sousa, 2004).

The second research question also examined the pattern of associations between support satisfaction and perceived personal agency by risk status. With the present data set it is not possible to test the causal direction of the relationship between support and personal agency. This is the case for a good deal of other research on personal agency. For instance, one aspect of personal agency, parental self-efficacy has been posited as an antecedent, a consequence, a mediator, and a transactional variable with respect to parent and child adjustment (Jones & Prinz, 2005). It is also a limitation that there is a lack of information concerning types of support and recipients' feelings or motivations for seeking support. This might help to understand better the different natures of social support revealed in our study. However, it is possible to examine the systematic variation in the pattern of associations depending on the type of source and risk status.

Overall, we have found more evidence in favor of the buffering than the depleting hypotheses and the buffering hypothesis held, always for informal sources but also in most cases for the formal sources. Direct links were found between satisfaction with informal sources of help and the five aspects of personal agency, indicating that the nature of these associations was not affected by the risk status. In both groups of mothers, satisfaction with the partner, mother (grandmother), and brother/sister supports represent a buffering factor associated with an increase in the reported level of control over their lives and in their perceived capacity to solve family problems and to have confidence in their role. Direct links were also found for satisfaction with mother support and couple agreement, satisfaction with mother and father support and role difficulty, and satisfaction with partner support and motivation for change. Therefore, satisfaction with the help provided by close relatives represents a buffering factor for all the facets of personal agency in at-risk and non at-risk mothers. There are at least two explanations of these results, and which could be complementary. Firstly, seeking relief care from close relatives is a strategy that may increase the chances of experiencing a less stressful parenting processe. Many studies have shown that social support may reduce the parental stress experienced through living under negative conditions by contributing to strategies for coping with stressors (Cochran & Niego, 1995; Coleman & Karraker, 1997; Cutrona, 1996; Schwarzer & Knoll, 2007; Wills & Shinar, 2000). Secondly, the help is embedded in a natural and mutual network of affectionate exchanges (Collins & Feeney, 2000), which makes the mothers feel valuable, effective and capable of the parenting task. In line with this interpretation, beneficial effects of satisfaction with mother and father support have been found for positive and negative parental behavior regardless of the risk status (Rodrigo et al., 2007).

Concerning the formal sources of help, satisfaction with school support was a buffering factor directly associated with higher internal control and self-efficacy no matter the risk status. However, the risk status played a more important role in those aspects of personal agency more closely related to the parenting role: couple agreement on educational matters, role difficulty and motivation for change. It seems that living in a stressful environment modifies the influence of formal support on parental functioning (Doherty & Beaton, 2000), specifically on those aspects of parental agency more related to educational matters. A buffering effect of the school support, specially tailored to the at-risk and non at-risk mothers' perception of the parental role, was found. As expected, in at-risk mothers, satisfaction with school support was positively related to higher perceived role difficulty and motivation for change. At-risk mothers who completed a parenting support program experienced more difficulty in their role as a positive result of the program, while at the beginning they tend to underestimate their importance (Rodrigo, et al., 2006). By contrast, in non at-risk mothers support satisfaction with the school and other institutions were associated to lower perceived difficulty in the parental role. Non at-risk mothers with larger and more supportive networks should normally perceive the tasks of parenting to be less difficult than mothers with smaller and less supportive networks (Melson, Ladd, & Hsu, 1993). School support is an interesting case of a buffering factor adjusted to the specific needs of the at-risk and non at-risk mothers (Ghate & Hazel, 2002).

The only depleting effect of risk status was found for the links between school and social services support and couple agreement. This means that, for this particular aspect, the quality of school and social service influences are clearly mediated by the family´s psychosocial conditions. At-risk mothers who were highly satisfied with these types of professional help reported the lowest level of couple agreement, and those who were least satisfied reported the highest level of couple agreement. In contrast, higher satisfaction with professional help was related to higher couple agreement on educational matters in non at-risk mothers. In all probability, non at-risk mothers might consider professional help to be a complementary support whereas this type of help plays a more substitutive role for at-risk mothers who frequently have problems with their partners. This might be especially true for social service professionals and school teachers, who might have the tendency to substitute the role of vulnerable parents or criticize them, thus taking responsibility away from the family (Woodcock, 2003). In addition, parental agreement may be lower in at-risk mothers because mothers (but not usually fathers, who rarely attend meetings with these professionals) are under the positive influence of the school and social services. However, professionals should know that this is a depleting effect since the educational message that the couple conveys to the children would be totally inconsistent.

With regard to the practical implications, this study showed that enrolling informal support for at-risk and non at-risk mothers is the most straightforward way of reinforcing their personal agency. It is also important to increase satisfaction with school support which is less used by at-risk mothers and may also have positive consequences on their parenting role. However, higher satisfaction with professional support may be associated with an increasing distance between the couple on educational matters in at-risk parents. Social support is an essential component of any multifaceted effort to prevent psychological difficulties in at-risk families and children. The challenge is to craft well-designed interventions that improve the social support afforded by natural helpers and enlist the assistance of formal helpers, specially trained to overcome the negative side of social support by paying special attention to reinforcing mothers' personal agency.

References

Ardelt, M. & Eccles, J. S. (2001). Effects of mothers' parental efficacy beliefs and promotive parenting strategies on inner-city youth. Journal of Family Issues, 22, 944-972.         [ Links ]

Azar, S. T. (1998). Parenting and child maltreatment. In W. Damon, R. M. Lerner, K. A. Renninger, and I. E. Sigel (Eds), Handbook of child psychology. Child psychology in practice (Volume 4, 5th edition) (pp. 361-388). New York: John Wiley & Sons.         [ Links ]

Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147.         [ Links ]

Bandura, A. (1997). Social learning theory. New York, Englewood Cliffs: Prentice Hall.         [ Links ]

Beach, S. R., Fincham, F. D., Katz, J., & Bradbury, T. (1996). Social support in marriage: A cognitive perspective. In G. R. Pierce, B. R. Sarason, & I. G. Sarason (Eds.), Handbook of social support and the family (pp. 43-66). New York: Plenum.         [ Links ]

Bolger, N., Zuckerman, A., & Kessler, R. C. (2000). Invisible support and adjustment to stress. Journal of Personality and Social Psychology, 79, 953-961.         [ Links ]

Ceballo, R. & McLoyd, V. C. (2002). Social support and parenting in poor, dangerous neighbourhoods. Child Development, 73, 1310-1321.         [ Links ]

Cochran, M. & Niego, S. (1995). Parenting and Social Networks. In M. H. Bornstein (Ed.), Handbook of parenting, Volume 3: Status and social conditions of parenting (pp. 393-418). Mahwah, NJ: Lawrence Erlbaum Associates.         [ Links ]

Cohen, S. & Wills, T. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310-357.         [ Links ]

Coleman, P. K. & Karraker, K. H. (1997). Self-efficacy and parenting quality: Findings and future applications. Developmental Review, 18, 47-85.         [ Links ]

Collins, N. L. & Feeney, B. C. (2000). A safe heaven: An attachment theory perspective on support-seeking and caregiving in intimate relationships. Journal of Personality and Social Psychology, 78, 1053-1073.         [ Links ]

Corse, S. J., Schmid, K., & Trickett, P. K. (1990). Social network characteristics of mothers in abusing and nonabusing families and their relationships to parenting beliefs. Journal of Community Psychology, 18, 44-59.         [ Links ]

Cutrona, C. E. (1996). Social support as determinant of marital quality: The interplay of negative and supportive behaviours. In G. R. Pierce, B. R. Sarason, & I. G. Sarason (Eds.), Handbook of social support and the family (pp.173-194). New York: Plenum Press.         [ Links ]

Doherty, W. & Beaton, J. (2000). Family therapists, community, and civic renewal. Family Process, 39, 149-161.         [ Links ]

Donovan, W. L. & Leavitt, L. A. (1989). Maternal self-efficacy and infant attachment: Integrating physiology, perception and behavior. Child Development, 60, 460-472.         [ Links ]

Elder, G. H., Eccles, J. S., Ardelt, M., & Lord, S. (1995). Inner-city parents under economic pressure: Perspectives on the strategies of parenting. Journal of Marriage and the Family, 57, 771-784.         [ Links ]

Evans, T. & Harris, J. (2004). Citizenship, social inclusion and confidentiality. British Journal of Social Work, 34, 69-91.         [ Links ]

Folkman, S. & Lazarus, R. S. (1988). Coping as a mediator of emotion. Journal of Personality and Social Psychology, 48, 150-170.         [ Links ]

Garbarino, J. & Kostelny, K. (1992). Child maltreatment as a community problem. Child Abuse & Neglect, 16, 455-464.         [ Links ]

Ghate, D. & Hazel, N. (2002). Parenting in poor environments. Stress, support and coping. London: Jessica Kingsley.         [ Links ]

Gottlieb, B. H. (1983). Social support strategies: Guidelines for mental health practice. London: Sage.         [ Links ]

Gottlieb, B. H. (2000). Selecting and planning support interventions. In S. Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement and intervention: A guide for health and social scientists (pp. 195-220). New York: Oxford University Press.         [ Links ]

Gracia, E., Herrero, J., & Musitu, G. (2002). Evaluación de estresores psicosociales en la comunidad. Madrid: Síntesis.         [ Links ]

Jack, G. (2000). Ecological influences on parenting and child development. British Journal of Social Work, 30, 703-720.         [ Links ]

Jones, T. L & Prinz, R. J. (2005). Potential roles of parental self-efficacy in parent and child adjustment: A review. Clinical Psychology Review, 25, 341-363.         [ Links ]

Koeske, G. F. & Koeske, R. D. (1992). Parenting locus of control: Measurement, construct validation, and a proposed conceptual model. Social Work Research and Abstracts, 28, 37-46.         [ Links ]

Lin, N. & Ensel, W. M. (1989). Life stress and health: Stressors and resources. American Sociological Review, 54, 382-399.         [ Links ]

Littell, J. H. & Girvin, H. (2005). Caregivers' readiness for change: predictive validity in a child welfare sample. Child Abuse & Neglect, 29, 59-80.         [ Links ]

Máiquez, M. L., Rodrigo, M. J., Capote, C., & Vermaes, I. (2000). Aprender en la vida cotidiana: un programa experiencial para padres. Madrid: Visor.         [ Links ]

Matos, A. R. & Sousa, L. M. (2004). How multiproblem families try to find support in Social Services. Journal of Social Work Practice, 18, 65-80.         [ Links ]

McHale, J. P. & Cowan, P. A (1996). Understanding how family-level dynamics affects children's development: Studies of two-parent families. San Francisco, CA, US: Jossey-Bass.         [ Links ]

McLoyd, V. C., Jayaratne, T. E., Ceballo, R., & Borquez, J. (1994). Unemployment and work interruption among African-American single mothers: Effects on parenting and adolescent socioemotional functioning. Child Development, 65, 562-589.         [ Links ]

Melson, G. F., Ladd, G. W., & Hsu, H. C. (1993). Maternal support networks, maternal cognitions, and young children's social and cognitive development. Child Development, 64, 1401-1417.         [ Links ]

Pierce, G. R., Sarason, I. G., & Sarason, B. R. (1996). The role of social support in coping. In M. Zeidner & N. Endler (Eds.), Handbook of coping: Theory, research, and applications (pp. 434-451). New York: Wiley.         [ Links ]

Raver, C. C. & Leadbeater, B. J. (1999). Mothering under pressure: Environmental, child and dyadic correlates of maternal self-efficacy among low-income women. Journal of Family Psychology, 13, 523-534.         [ Links ]

Rodrigo, M. J., Correa, A. D., Máiquez, M. L., Martín, J. C., & Rodríguez, G. (2006). Family preservation services on the Canary Islands: Predictors of the efficacy of a parenting program for families at-risk of social exclusion. European Psychologist, 11, 57-70.         [ Links ]

Rodrigo, M. J., Martín, J. C., Máiquez, M. L., & Rodríguez, G. (2007). Informal and formal supports and maternal child-rearing practices in at-risk and non at-risk psychosocial contexts. Children and Youth Services Review, 29, 329-347.         [ Links ]

Rodríguez, G., Camacho, J., Rodrigo, M. J., Martín, J. C., & Máiquez, M. L. (2006). La evaluación del riesgo psicosocial en las familias usuarias de los servicios sociales municipales. Psicothema, 18, 200-206.         [ Links ]

Sarason, I. G., Levine, H. M., Basham, R. B., & Sarason, B. R. (1983). Assessing social support: The Social Support Questionnaire. Journal of Personality and Social Psychology, 44, 127-139.         [ Links ]

Schwarzer, R. & Knoll, N. (2007). Functional roles of social support within the stress and coping process: A theoretical and empirical overview. International Journal of Psychology, 42, 243-252.         [ Links ]

Shumaker, S. A. & Brownell, A. (1984). Toward a theory of social support: Closing conceptual gaps. Journal of Social Issues, 40, 11-36.         [ Links ]

Turner, R. J. & Roszell, P. (1994). Personal resources and the stress process. In W. Avison & I. Gotlib (Eds.), Stress and mental health: Contemporary issues and prospects for the future (pp. 179-210). New York: Plenum.         [ Links ]

Wills, T. A. & Shinar, O. (2000). Measuring perceived and received social support. In S. Cohen, L. Gordon, & B. Gottlieb (Eds.), Social support measurement and intervention: A guide for health and social scientists (pp. 86-135). New York: Oxford University Press.         [ Links ]

Woodcock, J. (2003). The social work assessment of parenting: An exploration. British Journal of Social Work, 33, 87-106.         [ Links ]

Correspondence:
María José Rodrigo ]]> Faculty of Psychology. Campus of Guajara. University of La Laguna
La Laguna. Tenerife, Spain.
E-mail: mjrodri@ull.es

Manuscript received: 29/09/2010
Review received: 11/01/2011
Accepted: 12/01/2011