Foster parents are generally more comfortable fostering younger children (Carter, 2013; Tunno, 2015; Zuniga, 2012), and when most adolescents enter foster care with severe emotional scars from early childhood trauma, they require more than just caring (Escobar et al., 2014).

In South Africa, research shows that there is an increase in the number of adolescents entering the foster care system (Mosimege, 2017; Van der Westhuizen et al., 2014). Adolescence is generally being characterised by difficulties and can be a daunting experience for fostered adolescents who have experienced a chaotic upbringing during their early years of life (Escobar et al., 2014). Being fostered can raise much deeper questions and concerns about identity, belonging, and attachment (Coleman et al., 2016). Further studies conducted in South Africa pertaining to adolescents in foster care, suggest that limited or lack of knowledge in addressing both foster parents’ and adolescents’ needs in foster care is one of the main challenges facing the welfare system (Bester, 2014; Carter, 2013; De Jager, 2011). Most of these studies support the notion that efforts should be made to create opportunities to increase foster parents’ capacity to handle adolescents in their care, specifically those presenting with risk behavior. Adolescents in foster care are more prone to presenting risk behavior, which, if not attended to, may have long-term, negative consequences for these adolescents and society at large (Farineau, 2016; White, 2015). Negative consequences can manifest externally in deviance, impulsiveness, or aggression, while internally in behaviors such as withdrawal, anxiety, suicidal ideation, or self-destruction (Sanders & Morawska, 2014).

Sanders and Morawska (2014) are of the opinion that parenting practices have an influence on several life domains of the development of adolescents. Positive attachment experiences can positively influence the development of adolescents with regard to self-reliance, emotional regulation, and pro-social behavior (Fisher, 2014; Sroufe, 2005). However, despite the large of empirical evidence available on foster parenting, few international studies have been published on understanding the experiences of parents fostering adolescents presenting with risky behavior (Brown et al., 2013; Mnisi & Botha, 2016). Even less evidence is available on the attitudes of foster parents towards risk behavior presented by adolescents in their care (Albertos et al., 2016). Studies by researchers, such as Blakey et al. (2012); Holtan et al. (2013); and Mnisi and Botha (2016) point out that foster parents’ attitude and lack of effective parenting skills in managing the risk behavior of adolescents is one of the contributing factors for most foster care placement disruptions or even breakdowns. Against this background, this study explored the parental attitude and behavior of foster parents in circumstances where risk behavior is experienced.

Literature Contextualization

The contextualization and theoretical underpinning of this study involved a substantial literature review during which the following key strands in the parenting of fostered adolescents presenting with risk behavior were identified and examined: attachment issues; risk behavior in fostered adolescents; parenting behavior and attitude; and parenting styles. Each of the strands is briefly discussed below.

Attachment

Attachment theory suggests that the quality of attachment in a relationship is influenced and shaped by the characteristics of the child, the caregiver, and the social environment in which the relationship is taking place, and is predictive of the child’s future psychological, social, emotional, and behavioral functioning (Sroufe, 2005; Uzaina & Srivastava, 2016; Zastrow et al., 2019). Studies by researchers, such as Escobar et al. (2014) and Garcia Quiroga and Hamilton-Giachitsis (2015), confirmed a correlation between attachment insecurity and behavior problems. The work of Lionetti et al. (2015) focused on attachment patterns between caregivers and children in institutional care, foster care, and adoption. Caregivers, such as foster parents, who experienced insecure attachments were found to be more likely to display disorganised attachment behavior and likely to exhibit maladaptive parenting styles. These foster parents may find it challenging to open up to emotions, and to manage the adolescent’s negative emotions or behavior constructively (Mosimege, 2017). In South Africa, there is an increase in families experiencing difficulties, such as child abuse, violence, gangs, severe poverty, substance abuse, and HIV/AIDS infections among adolescents. All these difficulties are likely to influence parents to become overly protective (Roman et al., 2016; Ward et al, 2015). These aspects may also lead parents to approach their parenting role in a punitive manner in an attempt to protect their adolescents from vulnerabilities perceived by parents as a threat. As a result, this might compromise their ability to form a healthy, secure attachments with the adolescents. Foster parents, particularly in South Africa, are typically grandparents who lack or have limited skills and knowledge in caring for these emotionally and psychologically wounded children (Van der Westhuizen et al., 2014). Bester (2014) points out that the emotional and psychological protection needs of children are neglected, by the foster care system, which leads to an increase in risk-behavior amongst adolescents in foster care.

Risk behavior in Fostered Adolescents

There is no single cause for risk behavior, and in most instances, the underlying issues surrounding risk behavior are often of a diverse nature (Fisher, 2014). Adolescents often do not have the skills to self-regulate their emotions, especially anger feelings, and are likely to present with risk behavior (Forkey et al., 2013; Jackson et al., 2012). Risk behavior is targeted because the total functioning and psychosocial well-being of adolescents are jeopardized by it (Bester, 2014; Landers et al., 2017; Ohara & Matsuura, 2016; Tunno, 2015). Even though a substantial amount of literature globally shows that parental intervention plays a critical role in helping foster parents modify their adolescents’ behavior and improve interaction, there is still a need in South Africa to enhance the cognitive, emotional, and behavioral responses of foster parents towards parenting adolescents, more so in instances where risk behavior is experienced. In South African studies conducted by Ward et al. (2015) as well as Wessels et al. (2016), group-based parenting programs have been identified as an effective way of intervention for parents in general. However, only few group-based parenting programs are available in South Africa. No evidence was found to support whether such programs target risk behavior presented by fostered adolescents. Moreover, these programs may not be sufficient to build the knowledge of foster parents with regards to risk behavior or shape their attitudes in terms of their responses to such behavior.

Parenting Behavior

Parents are highly influential and as a result, most skills adolescents acquire derive from their interactions with parents and the social environment in which they interact (Sanders & Morawska, 2014). The extent and quality of foster parents’ parenting knowledge and attitudes are vital to improve development and interaction between parents and adolescents in their care. Thus, insight into parenting behavioral patterns and processes shape the expectations of parents, and their response to the behavior of their children (Armenta & Huerta, 2015). A parenting management style and effective involvement are pivotal in helping adolescents develop pro-social skills, self-controlling responses, and to internalise behavior standards (Hiramura et al., 2010; Sanders & Morawska, 2014). Open, participatory communication; problem-centred coping; confidence; and flexibility are identified as potential indicators of parenting behavior that can play a significant role in the management of parental stress and can promote family well-being (Symons et al., 2016). Therefore, parenting behavior can either be a protective or a risk factor (Bester, 2014; Farineau, 2016). Parental behavior can either be protective or a risk factor (Farineau, 2016). Protective behavior refers to any factor that reduces the severity of behavior likely to cause potential harm (Healey & Fisher, 2011), whereas risk factors entail the likelihood of increasing behavior that usually has negative consequences (Bester, 2014; Sanders & Morawska, 2014). In South Africa, research on parenting behavior and attitudes is still emerging, and to date only a few studies were conducted on parenting practices and their implications. Even though there is training available designed for foster parents on managing fostered adolescents while in their care, such training is often superficial in their approach as they do not address the root cause of risk-behavior presented by adolescents and explore the attitudes that foster parents hold pertaining to risk behavior.

Baumrind (1971), the founder of three global typologies on parenting dimensions, identified three commonly used parenting styles, namely: authoritative, authoritarian, and permissive. According to this theory, ‘authoritative parents’ tend to have high expectations for good behavior and maturity and exhibit a high level of responsiveness and warmth. Such parents engage in open discussions and encourage adolescents to understand expectations and how to make reasonable demands. Adolescents who experience this type of parenting tend to adjust well and are likely to be resilient (Cain & Combs-Orme, 2013; Firoze & Sathar, 2018). ‘Authoritarian parents’ have high expectations and exhibit low warmth. Such parents use threats, intimidation, and. at times, physical punishment to instil fear in their adolescents in order for them to comply. Adolescents who experience this style of parenting tend to be anxious, withdrawn, and even unhappy (Firoze & Sathar, 2018). However, Latouf (2008) reported that the authoritarian parenting style was found to toughen adolescents. The ‘permissive’ parenting style refers to parents who show warmth, are nurturing, and exhibit reasonably responsive behavior in meeting the physical and emotional needs of their adolescents, but without clear expectations and boundaries (Firoze & Sathar, 2018). Adolescents that grow up with this parenting style are typically laid-back, emotionally immature, and lacking in self-confidence (Norman, 2017).

Results of a South African study by Latouf (2008) show that the authoritative parenting style appeared to be beneficial for the emotional regulation of adolescents, as this style is associated with enhancing their ability to think independently and promoting a sense of resilience. In another study by LoveLife (2010) in which 122 parents of adolescents between the ages of 12 to19 years from across different provinces in South Africa participated, it was established that authoritarian parenting is the main parenting style used by most parents across different cultures and ethnic groups. However, these results are in contrast to similar studies conducted by Latouf (2008), Moyo (2012), and Makwakwa (2011) where the authoritative parenting style was identified as the predominant parenting style of South African parents. Roman et al., (2016, p.5) conducted a study on parenting techniques in South Africa, with a bigger sample of 746 respondents (36% males and 64% females) and found that parenting approaches vary significantly between males and females. The parenting styles of fathers were found to be different in the three ethnic groups, whereas the authoritative style was predominantly used by females, and within the different ethnic groups (Roman et al., 2016, p.5). The result implies that more research on parental behaviors and attitudes in South Africa is needed, particularly in cases where risk behavior is evident.

There has been an increase in recent decades of adolescents entering the South African welfare system, particularly within the North West Province. SASSA statistics (2018) show that, as of September 2018, the North West Province had a total of 34, 636 children in legal foster care. Locations where the increased number became prominent include Dr. Kenneth Kaunda and Dr Ruth Mompati municipal districts. The identified districts consist of both semi-urban and rural areas, where problems associated with risk behavior amongst fostered adolescents have been identified as some of the most common problems experienced in these 00communities. This study was therefore interested in assessing parents’ attitudes and behavior in caring for adolescents in their foster care.

Objective of the Study

The objective of this study was to assess respondents’ parental behavior and attitudes based on their self-reported responses.

Method

Research Approach and Design

This study adopted a quantitative approach with the use of a cross-sectional descriptive design (Creswell, 2013) to study the phenomenon of risk behavior of fostered adolescents against the behavior, attitudes, and parental styles of a specific group of foster parents.

Respondents and Sample Size

The respondents of this study included parents fostering adolescents in the North West Province, South Africa, under the supervision of designated welfare organizations. Purposive sampling was used to ensure those specific characteristics from the demarcated population were isolated (Maree, 2016). In the recruitment of respondents, the following criteria were applied for sample inclusion:

  • Respondents had to be registered on the caseload of the designated welfare organizations

  • Respondents had to be fostering adolescents between the ages of 14 and 17 years.

  • Respondents who reported that the adolescents in their care were presenting with risk behaviors.

  • Respondents had to have been fostering the same adolescent(s) for at least two years.

Sample Recruitment and Informed Consent

A list of designated welfare organizations rendering foster care services in the afore-mentioned districts was obtained, and a total of 29 welfare organizations were identified. During initial contact with representatives of these organizations, it was projected that there would be more than 100 cases in their caseload likely to meet the sample inclusion criteria. This was important in terms of the viability of the study to produce statistically meaningful results. The authors aimed to recruit as close as possible to 100 respondents (Creswell, 2013). A response rate of 50% is considered adequate for analysis and reporting (Rubin & Babbie, 2011). A total of 150 questionnaires were issued to respondents who met the sample criteria, and 97 were completed, resulting in a response rate of 64.6%. The results were statistically analysed.

The managers of the identified welfare organizations were contacted to request permission to conduct this study at their organizations, as well as negotiations to recruit possible respondents from their clients. The administrative officers of several organizations were asked to serve as mediators. Even though they work for targeted organizations, they have limited contact with potential respondents and are hence the ideal candidates to serve as mediators. They were told to approach possible respondents and explain the study’s purpose, as well as post advertisements in all participating organizations where foster parents may see them. The advertisement provided foster parents with information regarding participation in the study as well as what was expected of them. An independent person was assigned to get written informed consent from respondents interested to participate in the study.

Process of Data Collection and Measuring Instrument

The Parenting Style Dimension Questionnaire (PSDQ) (original version), developed by Robinson et al. (1995) was used to collect data via a survey method involving a self-reported measuring instrument. This study is explorative in nature. Field workers were trained to administer the measuring instrument during regular group sessions with foster parents in the selected welfare organizations. Sealed boxes were made available at the organizations where field workers could deposit the completed questionnaires for collection by the research team.

The PSDQ focuses on important aspects related to parenting skills, including warmth and involvement, reasoning, democratic participation, good-natured, verbal hostility, corporal punishment, being punitive, directness, lack of following through, ignoring misbehavior, and self-confidence (Kimble, 2014; Robinson et al., 1995). In the development of the questionnaire, Robinson et al. (1995) were guided by Baumrid’s global parenting typologies (i.e., authoritarian, authoritative and permissive parenting styles).The original version of the PSDQ consists of 62 items in which it is expected of respondents to evaluate their own parenting behavior, and choose the most applicable answer on a 5-point Likert scale, ranging from ‘never’ to ‘always’. The original, longer version of the PSDQ was favoured over the shorter version consisting of 32 items, because of its comprehensiveness and better fit to a South African context. The items are divided into three factors to measure authoritarian, authoritative, and permissive parenting styles. ‘Factor 1’ focuses on the authoritative parenting style and consists of 27 items divided into four sub-factors labelled as: a) Warmth and involvement (11 items); b) Reasoning/induction (7 items); c) Democratic participation (5 items); and d) Good-natured (4 items). ‘Factor 2’ involves the authoritarian parenting style and consists of 16 items with four sub-factors labelled as: a) Verbal hostility (4 items); b) Corporal punishment (6 items); c) Non-reasoning, punitive strategies (6 items); and d) Directness (4 items). ‘Factor 3’ is about the permissive parenting style consisting of 19 items, and three sub-factors labelled as: a) Lack of following through (6 items); b) Ignoring misbehavior (4 items); and c) Self-confidence (5 items). There are no negative items in the questionnaire, and high scores in each of the factors are indicative of a particular parenting style.

The PSDQ was originally designed in English, but Setswana was added to the English version for this study to allow field workers to explain the questions to respondents who did not speak English fluently. The PSDQ was translated into Setswana by independent, professional language experts because the study was conducted in a province where Setswana is the most commonly spoken language. After finishing the translation, it was translated from Setswana back into English to ensure maximum content similarity. A simple, linguistic meaning was maintained, and no items were substituted or eliminated.

The PSDQ claims good content validity, assesses important components in a relationship, and its internal consistency was assessed with Cronbach’s alpha (Kimble, 2014; Robinson et al., 2001). The construct validity of the PSDQ in this study was confirmed with confirmatory factor analysis (CFA), and the reliability was confirmed with Cronbach’s alpha. Internal consistency reliability in the current study was assessed with Cronbach’s alpha, and the outcomes were as follows: authoritative (consists of 20 questions: Cronbach’s alpha of 0. 86; authoritarian (consists of 20 questions: Cronbach’s alpha of 0,70); and permissive (consists of 15 questions: Cronbach’s alpha of 0,39). The results of this study is consistent with the results of studies conducted by Robinson et al. (1995), Ismaili, (2015) and Kimble (2014). It is however, of importance to highlight that our study results show very weak dimensions on authoritarian directiveness and permissive self.

Data were captured on a computer program (IBM SPSS Statistics Version 25, release 25.0) by a statistician for verification, and were discussed with the researchers (Field, 2013).

Statistical Analysis

The researchers followed the basic process of quantitative data analysis described by Creswell (2013) and implemented it as follows:

  • Data were captured and verified using EPI INFO stat C.S. (trends identification in data sets).

  • PSDQ analyses are conducted through IBM SPSS Statistics Version 25, Release 25.0 (Field, 2013).

  • Scale properties were established by means of a CFA and a reliability analysis by means of Cronbach’s alpha.

  • Descriptive statistics were calculated to represent the data in figures and tables.

  • T-Tests, ANOVAs, and Spearman’s correlations were used to determine if the sub-scale scores could be associated with the biographical data. Because a random sample was not used, the results cannot be generalized, and p-values are therefore irrelevant. Instead of only reporting p-values, Cohen’s D was used to calculate effect sizes for the tests in order to identify statistically significant differences. Effect sizes of 0.2 and higher were interpreted as having a minor but significant effect.

Biographical Information of Respondents

The researchers anticipated that certain respondents’ biographical information would have an impact on the foster parent-adolescent relationship and how they deal with certain life challenges. This context includes their ethnicity, religious beliefs, age, level of education, employment, marital status, and the number of foster children they care for. Furthermore, it was necessary to gain insight into the fostered adolescents' identifying information, such as gender and ethnicity, as this can provide insight into how foster parents navigate their relationships with these adolescents. Table 1 summarizes respondents' biographical information.

Table 1 Biographical information of respondents

Results on the PSDQ

Scale Reliability

In this study, Cronbach’s alphas between 0.5 and 0.9 were reliable (Mohajan, 2017). Two scale dimensions, Authoritarian directness, and Permissive self-confidence rendered low reliability coefficients (0.039 and 0.18) and were discarded from further analyses. The descriptive statistics indicated that standard deviations on these two dimensions were 0.99 which indicates significant variations in responding.

Table 2 provides comparisons between parental styles of different ethnic groups. The different ethnic groups of the respondents were grouped together in Table 2. The ethnicity of respondents was deemed important for the purposes of this study because it is likely to influence foster parents’ attitudes toward raising and disciplining adolescents. The hypothesis that ethnic groups differed regarding the dimensions of the Parental Styles and Dimensions Questionnaire was examined by means of one-way ANOVA based on the four ethnic groups, Group 1: Tswana, Group 2: Sotho, Group 3: Coloured and Group 4: other, and the Brown Forsythe post-hoc test for evaluating the equality of group variances. The Brown Forsythe test shows good robustness against many types of non-normal data while it still retains good statistical power. This test further relies on the median as measure for the underlying distribution of the data. The Tukey B post hoc test was used to investigate the exact differences between groups. This test is suitable for unequal group comparison and uses the Harmonic mean instead of the arithmetic mean. α = 0.05 was used in this instance. Eta squared effect sizes were also calculated to further interpret between group differences. In these instances Cohen classifies effect sizes of 0.1 as small effect, 0.6 as medium effect and 0.14 as a large effect (Pallant, 2016).

Table 2 Descriptive statistics on parental styles among different ethnic groups

A one-way between-groups analysis of variance was conducted to explore the impact of ethnic groups on parental styles, as measured by the Parental Styles Dimensions Questionnaire (PSDQ). The ANOVA indicated statistically significant differences at the p < 0.05 level for the four ethnic groups on the authoritarian–verbal hostility dimension: F (3, 96) = 2.818, p = 0.043) and authoritarian–non-reasoning: F (3, 96) = 3.333, p = 0.023. According to the above Table, ETA Effect size calculations indicated consistently that Group 3: Coloured Effect sizes were larger than any of the other groups with respect to all the dimensions of the Parental styles scale. In all these instances the Mean scores for the Coloured group was higher than the other groups although the difference was small. Regarding authoritative-reasoning dimension, the Coloured group was followed by Group 1: Tswana ethnic group. The overall lowest Mean scores were obtained by Group 2: Sotho ethnic group. In contrast, and consistent with the ANOVA result, the practical value in terms of ETA effect size of the Coloured group on the corporal punishment, verbal hostility, and non-reasoning dimensions compared to the other groups, were large. (0.71, 0.71 and 0.56). The Sotho group (Group 2) on average scored the lowest of all the groups on both these dimensions. As a result, it is concluded that the Coloured group appeared to have overall high levels of positive, warm and participatory parenting styles while also being very strict, hostile and unreasonable towards the foster children. The Coloured parents had generally positive attitudes towards the fostered adolescents in their care, but when things went wrong, they were more likely to become enraged with difficult children.

As shown in Table 3 the results of the independent t-test that was used to determine and compare the responses of respondents according to their marital status. There was a significant difference in responses between the married and single foster parents. The married foster parents showed a higher mean between (4,17 to 4,36) on two of the factors, namely authoritative warmth, authoritative reasoning whereas, the single foster parents showed a higher mean score (2,08 to 2,86) on authoritarian non-reasoning. The effect size differences of (0,36 to 0,48) showed no practical significance. There was no significance difference in terms of the geographical location of respondents.

Table 3 Independent t-test regarding marital status, gender, and geographical

Table 4 shows that there was a strong correlation between the age of respondents with regards to authoritative warmth and a negative correlation with authoritative verbal hostility. These findings suggest that the older the foster parents are, the more authoritative warmth, authoritative reasoning and authoritative participating they view themselves, and the less verbal hostile, non-reasoning and permissive self-confidence. Permissive self-confidence dimension was very low with an effect size of 0,39. Results further showed that the higher educated foster parents are the lesser authoritative reasoning and are more hostile. The correlation between the number of adolescents in foster parents’ care is high between authoritative reasoning and authoritative participation, and negative in correlation with permissive self-confidence. This suggests that the more fostered children in a parent’s care, the more authoritative reasoning and participative they regard themselves. Foster parents’ period of fostering an adolescent show a high correlation between authoritative reasoning and authoritative participation. These results suggest that respondents’ views on the longer they care for an adolescent, the more authoritative reasoning and participating and less permissive they are. Respondents with older biological children (those over the age of 18 and living independently) had a higher mean (3.58 to 3.62) on the authoritative parenting style. According to these results, respondents are of the view that, the older their children were, the less authoritarian they were. Respondents with their own younger biological children scored a high mean on the authoritarian parenting style (2.46 to 2.86). The difference in effect size was significant (0.59 to 0.74). The effect size difference for permissive self-confidence was 0.39, which was statistically significant.

Table 4 The Responses of foster parents on factors: (foster parents’ age, level of education, number of biological children, and period of fostering an adolescent)

Table 5 shows that the relationship between authoritative and authoritarian factors was statistically significant. The authoritarian parenting style and the permissive parenting style were found to have a significantly high correlation. The findings also revealed a negative relationship between authoritative parenting style and self-confidence. Verbal hostility correlated positively with non-reasoning, authoritarian directness, and corporal punishment, and negatively with permissive self-confidence. Corporal punishment was found to be associated with permissive self-confidence and verbal hostility. There was a link discovered between non-reasoning and verbal hostility, corporal punishment, and permissive self-confidence. Corporal punishment, self-confidence, verbal hostility, and non-reasoning all had a positive correlation with authoritarian directness (Table 6).

Table 5 Independent t-test on the ages of the respondents’ biological children
Table 6 Parenting practice dimension questionnaire (PSDQ)—measuring parenting styles: authoritarian, authoritative and permissive

Discussion

Contemporary research in the fields of adolescence and foster care pays attention to the significance of parenting by focusing on aspects, such as parenting attitudes, skills, values, and a good understanding of the developmental milestones of adolescents (Farineau, 2016; Landers et al., 2017). However, few studies have reported on parental attitude and behavior towards adolescents presenting with risk behavior particularly those in foster care (Sanders & Morawska, 2014; Firoze & Sathar, 2018). This study was interested in exploring foster parents’ attitudes and behavior in response to risk behavior presented by adolescents in their care. The results showed significant differences in parenting attitudes, while socio-economic factors and contextual factors appeared to have an influence on parents’ attitudes and behavior towards the presence of risk behavior in fostered adolescents.

This study’s results corroborate what has been emphasized by a number of researchers that parenting in general needs to be viewed in the broader context in which the relationship takes place since these aspects can have an impact on parents and subsequently shape their attitudes towards parenting. From the biographical information the respondents provided, it was deducted that all the respondents in this study were females (foster mothers), and most of them were older than 50 years. Most of them reported to be single, and fostered adolescents on their own with no other means of support. Most respondents reported to be single, parenting adolescents on their own with no other means of support. The results of our study show that when the foster parent is single whether with the help of family scored low on authoritative warmth whereas those with spouse are relatively good regarding warmth. Given the complexities associated with parenting in general, having to care for an adolescent while being a single parent with limited resources is likely to have an enormous impact in parenting behavior and subsequently on family relations (Zuniga, 2012). This suggest that having a spouse is likely to have moderate effect in helping a foster parent care for the fostered adolescent. In addition, there was small effect size on the foster parents’ state of being employed or unemployed. The result, however, show that foster parents who are state grant recipient and self-employed were better with participation, warmth, less punitive towards the adolescents, not hostile and punitive. These results are consistent with those reported by (Roman et al., 2016; Ward et al., 2015).

Important results in line with the different ethnic groups suggest that even though the Coloured respondents show good participation, nurture, warmth and reasoning, but in a small proportion of the good Coloured foster parents’ cases because they are dealing with difficult children, they are likely to become hostile. Given this result, it is therefore necessary that the Coloured foster parents be helped to learn not to snap under difficult circumstances and be able to deal with conflict and difficulties while they remain calm and warm. Thus, characteristics of the parents, the child developmental stage, and the social environment in which the interactions and relationship take place influence parenting attitude and behavior (Cain & Combs-Orme, 2013). Understanding these these aspects can help social work practitioners on how to best design, support and provide needed intervention for both the fostered adolescents and parents.

The ability to meet the needs of emotionally wounded adolescents in foster care appears to be one of the most difficult tasks that foster parents might have to master. Continual stress may lead to a decline in parenting abilities, which may affect the mental and emotional well-being of parents and can lead to a maladaptive way of coping. Negative coping mechanisms can have lasting detrimental effects on the nature of the relationship that exists between adolescents and their foster parents. Thus, a need for program to assist foster parents deal with adolescents presenting with risk behavior is of paramount importance. Foster parents need to be equipped with necessary skills and knowledge to help them not to snap when they find themselves under difficult situation enable them under challenging circumstances they may encounter while caring for the adolescents. It is therefore imperative that parents’ attitude and behavior towards foster parenting is thoroughly explored to determine necessary interventions that are context specific to meet the needs of both foster parents and adolescents in their care (Sanders & Morawska, 2014).

Another important result was that most of the fostered adolescents came from the Sotho ethnic group, while the majority of foster parents were from the Tswana ethnic group. These results suggest that foster parents do not necessarily foster adolescents who come from a similar cultural or ethnic background language as theirs. This may be attributed to the South African philosophy of “Ubuntu” (I am, because of you”). This philosophy encourages the spirit of caring for each other as human beings irrespective of our cultural or genetic heritage (Roman et al., 2015). It is rather more important given the fact that South Africa continue to experience a need of adolescents who require placement in foster care. This philosophy, therefore, has the potential to influence foster parents to assess and re-think their attitudes towards parenting.

Furthermore, the results showed a correlation between the authoritative and authoritarian parenting styles with regard to most of the respondents who participated in the study. The report on a permissive parenting style was extremely low. These findings are consistent with those reported by Roman et al. (2016). Most parenting studies show that female parents tend to practice an authoritative parenting style characterised with warmth, nurture, and a strive to create stability (Kimble, 2014; Latouf, 2008; Makwakwa, 2011, Moyo, 2012; Roman et al., 2016).

Conclusion

The results of our study suggest that parenting attitudes and behaviors toward fostered adolescents play a significant role and may have a negative or positive impact on the well-being of both the adolescents and the parents. Parenting practices play an important role in the development of adolescents, necessitating the use of effective research-based interventions. When foster parents feel empowered to deal with risky behavior displayed by adolescents in their care, they will gain confidence and develop positive coping strategies.

Limitations

The fact that the study was conducted with a small sample can be viewed as a major limitation. It makes it, therefore, impossible to generalise the results to the entire population of foster parents in the North West Province. It is also important to note that even though the study was not limited to only female foster parents, respondents turned out to be females only. Thus, the gender of respondents can also be viewed as an additional limitation of this study. Despite its limitation, the outcomes of this study make a significant contribution to the existing literature within a South African foster care setting.

Future Research

Future research recommend hypothesis testing and replication of the exploratory results of this study. The sample that participated in this study consisted only foster mothers, it will therefore be more beneficial for future studies to include foster fathers. In addition, South Africa is a very diverse society that consist of different cultural practices, ethnicity, language and socio-economic status. All these factors may have influence on parents’ attitudes and behavior towards parenting and therefore require further exploration. Furthermore, future similar studies to be conducted with adolescents in foster care to ascertain their views on their parent’s attitudes and behavior.

Practice Recommendations

Results of the current study have implications for child welfare practices. With an increase in older children entering foster care, there is a need for social work practitioners to explore the views and attitudes of foster parents on risk behaviors. Obtaining an in-depth understanding, will guide practitioners on developing research-based intervention that is effective and responsive.