Perceptions of parenting styles in South Africa: The effects of gender and ethnicity

Abstract Parenting style is an area of research which is emerging in South Africa. Previous research notes significant differences between ethnic groups in terms of parenting. In a country as diverse as South Africa, which has also evolved from an era of separatism and segregation, comparing parenting styles across ethnic groups is an interesting topic. This is a comparative study using a cross-sectional design. The Parenting Style and Dimension Questionnaire was used to collect the data. The final sample consisted of 746 participants with a mean age of 20.48 (SD = 1.96) years; 36% males and 64% females. The results found maternal authoritative parenting style to be the most prevalent across and within groups, but there were also significant differences between and within groups. Fathers’ parenting style was perceived as significantly different in three ethnic groups, but mothers’ parenting styles were not perceived as significantly different.

ABOUT THE AUTHOR Nicolette Vanessa Roman, PhD in Psychology, is nationally rated as an established researcher in South Africa and currently heads the Child and Family Studies Programme, in the Social Work Department at the University of the Western Cape. She has been a recipient of European Scholarships, received research funding from the National Research Foundation and provides continuous education for social workers and community-based organizations. She has published and presented her research both locally and internationally. She has also recently edited a book on parenting behaviours. This paper presents the first comparative information on parenting styles based on ethnicity and gender in South Africa. Trying to understand parenting is a continuous challenge and is an under-researched area in South Africa. However, this paper adds to the knowledge base in South Africa and internationally provides information on an underreported population.

PUBLIC INTEREST STATEMENT
South African families have a history of experiencing violence because black family members were separated by laws so mainly single women parented their children. Research on parenting is just over 10 years old in South Africa, but we do not know how different groups of mothers and fathers parent. Thus, we asked youth in different ethnic groups about the way their parents parent. We then compared male and female responses for the different groups and compared parenting of mothers and fathers. We found that mothers used more positive parenting than fathers for all the ethnic groups, but fathers parented differently in the different ethnic groups. Sons and daughters also had different views on their parents' parenting.

Introduction
Parenting and associated variables have been well documented as a developmental process and socialization tool. Over five decades of parenting research provides vast evidence of different approaches of parents (and the effects of these) in raising their children (Carr & Pike, 2012;Cheah, Leung, Tahseen, & Schultz, 2009;Fuemmeler et al., 2012;Kerr, Stattin, & Özdemir, 2012). In particular, Baumrind (1966Baumrind ( , 1967Baumrind ( , 1991 identified different parenting styles and the effects these parenting styles have on children's behaviour and to a certain extent well-being and subsequent adult adjustment. In her early research, Baumrind (1967) identified three common parenting styles known as authoritative, authoritarian and permissive parenting styles. These parenting styles reflect the patterns of parental values, practices and behaviours as well as a distinct balance of responsiveness and demandingness presenting different outcomes for children. The cornerstones of these parenting styles are warmth and control (Maccoby & Martin, 1983). The "warmth" element aims to foster the self-reliance, individuality and self-assertiveness of a child (Hart, Newell, & Olsen, 2003). The "control" element integrates children into the whole family, through disciplinary and supervision methods applied by parents (Baumrind, 1991).
In terms of understanding the parenting styles and the effects on children, authoritarian parents are low on warmth and high on control. They attempt to shape, control and evaluate the behaviour and attitudes of the child in terms of a set of standards. Obedience is all-important and imposed with harsh punishment. This style does not allow for individuation. There is very little communication between the parent and the child. Children of authoritarian parents are withdrawn, discontented and distrustful, and present with maladaptive behaviour (Baumrind, 1967).
Authoritative parents, while high on control, are also high on warmth. They demonstrate autonomy, self-will and discipline towards their children. One of the most apparent characteristics is the manner in which discipline is achieved. Authoritative parents use both power and reason (Baumrind, 1967(Baumrind, , 1978(Baumrind, , 1991, which allow children to be aware of what they have done wrong in order to amend their behaviours. This parental approach is followed by setting structured guidelines and allowing the child some freedom at the same time as conforming to discipline. The outcomes for children raised by authoritative parents are social responsibility, independence, self-confidence, higher self-esteem and adaptive behaviour (Baumrind, 1991;Steinberg, Lamborn, Dornbusch, & Darling, 1992).
Permissive parents show more warmth and less control. They use less discipline, and they appear to be more compliant with the desires and actions of their children. This form of parenting lacks structure for children and instead portrays more responsiveness and indulgence with their children's needs and wants without setting proper boundaries. Outcomes for children include high self-esteem; however, they are at risk of using drugs and alcohol, and often display problematic behaviours (Scaramella & Leve, 2004;Weis, 2002). Children raised by permissive parents show a lack of self-control, are egocentric and demanding and have difficulty in forming relationships (Baumrind, 1997).
Thus, in summary, the approach of parents has an effect on the child, which is further evidenced by a recent systematic review conducted by Davids, Roman and Leach (2015) who found that negative parenting approaches were associated with maladaptive behaviour of children and adolescents.
Although parenting has been found to be linked to child outcomes, research also suggests that gender plays a role in parenting. For example, research mainly conducted in North America and Europe found that mothers are more authoritative in their parenting than fathers (Conrade & Ho, 2001;Russell et al., 1998). This was similarly found in research conducted in South Africa (Roman, 2014;Roman et al., 2015), a comparative study of parenting in the USA and Australia (Russell, Hart, Robinson, & Olsen, 2003) and a meta-analysis in which gender affected the parenting approach of the parents (Hoeve et al., 2009).
In terms of the predictive aspects of parenting styles, Rinaldi and Howe (2012) compared mothers' and fathers' unique parenting styles in terms of internalizing, externalizing and adaptive behaviour in their children, and found that maternal permissive parenting and paternal authoritarian parenting significantly predict externalizing behaviour, while paternal authoritative parenting predicts adaptive behaviours. Furthermore, Simons and Conger (2007) and McKinney and Renk (2008) found that when both parents adopt an authoritative parenting style, children will have very positive outcomes. In addition, if one parent is authoritative, this may act as a buffer for children in problematic homes with the other parent being less optimal in parenting. The effects of gender were also present in studies which focused on parent-child dyads. This effect was found in a meta-analysis in which same-sex parent-child pairs significantly predicted delinquent behaviour (Hoeve et al., 2009).
Research examining and comparing parenting strategies in terms of ethnicity or race highlights the disparities found in different groups. For example, in the USA, Hofferth (2003) found differences in fathering in black, Hispanic and white groups, indicating that black children's fathers exhibit less warmth but monitor their children more; Hispanic fathers monitor their children less; and both groups exhibit more responsibility for child rearing than white fathers do. In another study, Varela et al. (2004) compared two-parent families of Mexican descent and Caucasian-non-Hispanic parents, and found that parents used more authoritative parenting styles in both groups. Parents of Mexican descent used more authoritarian parenting than Caucasian-non-Hispanic parents. In a study conducted in the UK, significant differences in parenting were noted between groups with minority groups reporting harsher parenting approaches than in white groups (Maynard & Harding, 2010). More recently, a study comparing different migrant and native ethinic groups in Germany found significant differences in parenting styles with migrant groups using more neglectful and authoritarian parenting styles (Nauck & Lotter, 2015).

The South African context of parenting
In South Africa, research focusing on parenting is limited to parenting styles and parental behaviour and their effects on child outcomes (Kritzas & Grobler, 2007;Latouf, 2008;Makwakwa, 2011). Studies, which specifically focused on the associational qualities between parenting styles and child outcomes, have shown that in early childhood, an authoritative parenting style was associated with acceptable behaviour of children in a sample of five-year olds (Latouf, 2008); during adolescence, authoritative parenting explained the variance in resilience for black and white adolescents, while an authoritarian parenting style was linked to emotion-focused coping strategies of adolescents (Kritzas & Grobler, 2007). An authoritative parenting style was also associated with thoughtful decision-making in young adults (Makwakwa, 2011) and the goals and aspirations of youth (Moyo, 2012). The studies conducted by Latouf (2008), Makwakwa (2011) and Moyo (2012) showed that parents used mainly an authoritative parenting style across ethnic groups. The studies which focused on parental behaviour and its effects on child outcomes suggest that parental behavioural control, monitoring and limit setting predicted high rates of substance abuse during adolescence (Amoateng, Barber, & Erickson, 2006); fathers' overprotection resulted in less intimacy and less assertive conflict resolution/communication, while mothers' care resulted in assertive conflict resolution/communication and independence in adult relationships, with mothers' overprotection resulting in less independence (Lowe, 2005).
Studies comparing parenting in terms of race found either similar or significantly different results across the race groups. For example, Kritzas and Grobler (2005) found that authoritative parenting significantly accounted for the variance in resilience for black and white adolescents. In addition, a paternal authoritarian parenting style was associated with emotion-focused coping strategies in white adolescents, which is in contrast to international research associating authoritarian parenting and harsh parenting with maladaptive behaviour. De Jager (2011) compared white and coloured 1 adolescents' relationships with their mothers, fathers, best friends and romantic partners.
Adolescents had stronger relationships with their mother than with their father, as indicated by nurturance, support, satisfaction, affection and intimacy. Bomester (2012) examined the parentadolescent relationship in a coloured community, and found strong mother-daughter and fatherson relationships. In general, mothers spent more time with the adolescent child than fathers did.
In summary, the parenting-child behaviour outcomes in South Africa are inconsistent with international research in terms of parenting styles, gender and ethnicity. In terms of ethnicity, the minority groups presented internationally are the majority in South Africa. These majority groups have been historically socio-politically subjected to violence, separation and segregation, but for the past two decades, there has been a shift to a more humanitarian and democratic society with new child protection laws and family policy. In a country as culturally diverse as South Africa, and given its socio-political history, it would be of interest to examine if international findings of parenting styles are replicated in terms of gender and ethnicity between native ethnic groups in South Africa, as the outcomes may have implications for practitioners in South Africa and add to the international body of knowledge on parenting differences in an under-reported population. This study therefore compared the parenting styles (i.e. authoritative, authoritarian and permissive) of mothers and fathers in different ethnic groups (i.e. black, coloured and white) in South Africa and proposed that there would be significant differences between groups in terms of fathers' parenting styles. This may be the first known study to be conducted in South Africa comparing parenting styles across ethnic groups.

Methods
This study is comparative with a cross-sectional research design. Participants who identified themselves as Indian/Asian (N = 8) and "other" (N = 19) were not included. Of the remaining 785 participants, 531 (67.6%) identified themselves as black Africans, 232 (29.6%) as coloured and 22 (2.8%) as white. The university where this sample was located has a predominantly black student population and this explains the small number of white (N = 22) and the relatively large number of black participants (N = 531). In the total group of 785 participants, 35.7% were male and 64.3% female. Among the black (38.4%) and especially the white (40.9%) participants, there were relatively more boys than girls. In the coloured group, there were relatively more girls (71.1%) and fewer boys (28.9%). Participants identified their parents as married (54%) or single (46%).
As Table 3 shows, for paternal parenting styles, we had 749 participants, and 782 for maternal parenting styles. The final sample of participants who answered both the maternal and paternal parental styles questionnaire consisted of 746 voluntary university students (36% males and 64% females; 69% black, 28.3% coloured and 2.7% white) with a mean age of 20.48 years (SD = 1.96).

Measurements
Participants were asked to retrospectively perceive the parenting styles of their father and mother when they were about 16 years old and still living at home. All participants completed the English 32-item Parenting Style and Dimension Questionnaire developed by Robinson, Mandleco, Oslen and Hart (2001). Participants responded on a five-point Likert scale with 1 = Never and 5 = Always. The items measured the (remembered) three parenting styles of fathers and mothers separately: authoritative (e.g. Gave praise when I was good), authoritarian (e.g. Used physical punishment as a way of disciplining me) and permissive (e.g. My parent found it difficult to discipline me). alphas for mothers' and fathers' parenting styles in the three ethnic groups and for all three groups combined. As we can see in Table 1, the internal consistencies are satisfactory, except for the scale "mothers' permissive parenting style" among black and coloured participants. In addition, the alphas for coloured fathers' authoritative and authoritarian parenting styles are almost similar which could be due to redundancies on certain items possibly based on cultural interpretations.

Procedure
Permission to conduct the study was granted by the University Review Board. The study was explained to the participants before the questionnaire was administered. Students were also informed that they could choose not to participate in the study or withdraw from it at any time during the data collection process. In addition, students were informed that their confidentiality and anonymity would be maintained at all times. Participants completed consent forms before completing the questionnaire.

Data analysis
Means scores on parenting styles were compared using analysis of variance (MANOVA and ANOVA) and within groups using repeated measures analysis of variance. Bonferroni correction was applied to account for multiple testing, and obtained adjusted p-values. The level of significance for all statistics tests was 0.05. All analyses were done with SPSS version 22.

Results
In Table 2, the intercorrelations for the total group are presented. These intercorrelations present similar patterns between the six parenting styles in the three ethnic groups. The correlations were positive and high between the three styles for fathers. For mothers, the correlation was significantly negative between the authoritative and the authoritarian parenting styles and rather low, but positively significant, between the authoritative and permissive parenting styles. The correlations between fathers' and mothers' parenting styles indicate rather high (around 0.30) significantly positive correlations between the corresponding parenting styles. Table 3 provides the mean scores and standard deviations (in parentheses) for the remembered parenting styles of black African, coloured and white groups as a function of ethnicity and gender. A MANOVA showed a significant main effect of ethnicity (Wiks' λ = 0.97 F [12, 1,470] = 1.89, p = 0.031, multivariate 2 p = 0.015). There was no main effect for gender or for the interaction of ethnicity by gender. Table 4 presents the same mean scores as a function of ethnicity.
A follow-up ANOVA with a Bonferroni-adjusted alpha level was conducted. The results suggest that there is a significant main effect of ethnicity for the authoritative parenting style for fathers (F (2, 743) = 3.45, p < 0.032, Within each ethnic group, an ANOVA for repeated measures was conducted to see if there were significant differences between the six perceived parental styles. Table 4 shows that there were significant differences in each group. Based on a posteriori comparison (Bonferroni), the following significant findings were found:  In the black African group (F (1, 748) = 160.60, p < 0.000, 2 p = 0.238), mothers scored significantly higher than fathers for all three parenting styles. The score for mothers' authoritative parenting style (M = 3.42, SD = 0.74) was significantly higher than all other scores. There were no significant differences between the permissive (M = 2.33, SD = 0.93) and authoritarian (M = 2.24, SD = 0.94) parenting styles of fathers, but they were both significantly lower than all other parenting styles of both fathers and mothers. In addition, in this group, the score for mothers was significantly higher than fathers for each of the three parenting styles.
In the white group (F (1, 748) = 6.70, p < 0.000, For mothers, the score for the authoritative parenting (M = 3.49, SD = 0.84) was significantly higher than the permissive and authoritarian parenting styles. In the white group, fathers and mothers were not significantly different for each of the three parenting styles. Table 5 rank orders the six parental styles in each ethnic group. The rank order is almost identical in the black African and the coloured groups. In the white group, the two highest scores are for fathers' and mothers' authoritative parenting style. However, compared to the other two groups, fathers' authoritarian parenting style becomes more important and mothers' less important.

Discussion
This is the first-known study to compare remembered parenting styles (i.e. authoritative, authoritarian and permissive) of mothers and fathers in different ethnic groups (i.e. black, coloured and white) as perceived by a non-representative sample of South African university students. In the current study, a main effect was only found for ethnicity, but not for gender or interaction effects for ethnicity by gender.
South Africans come from a history of violence steeped in a dominant patriarchal society (Ginwala, 1990). Historically and currently, the majority of the South African population, that is black Africans, continue to live in poverty, with additional social problems such as unemployment, substance abuse and crime (Peltzer, Ramlagan, Johnson, & Phaswana-Mafuya, 2011;South African Human Rights Commission (SAHRC) and UNICEF, 2014). These parents could therefore be considered at-risk for harsh and punitive parenting (Brooks, 2011;Phillips, Erkanli, Costello, & Angold, 2006). After almost two decades of democracy, with the acquisition of new child and family laws as well as the abolition of corporal punishment, there is a covert assumption that parenting should be more positive. However, the South African Institute of Race Relations (Holborn & Eddy, 2011) indicates that there are serious problems within the families, and parents are the focus, as there is mention of a lack of parental responsibility.
The results of this study suggest that parenting styles of mothers and fathers were perceived as significantly different between and within ethnic groups. For between groups, only fathers were perceived as significantly different in their parenting styles, and not mothers. It could be that mothers were perceived as being more involved than fathers, as suggested in previous research (Bomester, 2012;Conrade & Ho, 2001;Makwakwa, 2011;Moyo, 2012;Russell et al., 1998). In the present research, as in previous research, perceptions of mothers' authoritative parenting styles were more prevalent than any other parenting style (Makwakwa, 2011;Moyo, 2012). Previous South African research suggests that there are stronger relationships between mothers and their children than between fathers and their children (De Jager, 2011).
The differences between fathers of different ethnicities in the present study showed that black African fathers scored significantly lower on both authoritative and authoritarian parenting styles than whites. This finding was consistent within the groups as well. This significant difference could be due to the fact that historically and culturally, black African fathers were less involved with their children, possibly due to the experiences of apartheid with fathers working away from their families (Bernstein, 1985;Ginwala, 1990;Posel, 1991). In addition, gender ideologies could be very prevalent, as found in previous research in the USA (Hofferth, 2003) where less warmth was experienced between black fathers and their children than Hispanic and white groups. Furthermore, the parenting style of white parents could be linked to a more Westernized approach to parenting with more father involvement than in the other ethnic groups, as well as a more frequent authoritative parenting style (Brooks-Gunn & Markman, 2007). These findings were similarly found in the UK and Germany for minority groups (Nauck & Lotter, 2015).
Research suggests that, within the family, different relationships exist and different interactions occur based on the gender of the child and the gender of the parent (Bohanek et al., 2009;Hastings, Rubin, & DeRose, 2005;Lambert & Cashwell, 2004). In the present study, boys scored significantly higher than girls for mothers' authoritarian style. However, girls scored significantly higher than boys for mothers' authoritative parenting style. This finding is in line with previous research, which highlights distinct gendered parenting occurring between a mother and her son and a mother and her daughter (Bomester, 2012;Georgiou, 2008;Lipps et al., 2012). In the three groups, mothers seem to use a more authoritative parenting style with their daughters than with sons (see Table 2). This may have implications for the same-sex parent-child relationship (Hoeve et al., 2009).

Limitations
This study has limitations. The sample is non-representative and therefore the results cannot be generalized to the population at large. In addition, this sample of students also reported their parents' parenting styles retrospectively. This reporting may pose a challenge for interpreting the results as the students may have a different current relationship with their parents than in the past. This could mean that different results would emerge if a study was conducted on current parenting styles rather than past parenting styles.
The benefits of this sample and retrospective parenting are that the results of this study may be relevant to practitioners who may need to use a cultural perspective in designing or adapting their intentions to address the parent-child relationship. Furthermore, future studies regarding this topic could consider including a more representative sample of participants. Although there were significant differences between and within groups, the study may have different results if a larger sample size of white and Indian participants had participated. In the current study, the small number of white participants is a limitation.
As a cross-sectional design, this study only provides a snapshot, and does not offer any causal effects. The intention of the study was to provide significant differences between and within groups, but future research may consider either longitudinal studies or ethnographic types of studies.
Another limitation of the current study is the focus on young adults who remembered the parenting styles of their parents. The perceptions of current and past experiences of their parents' parenting styles may differ. However, the current study provides first insights into group differences in a South African context for parenting styles and therefore future research could focus on a longitudinal developmental study in this regard.

Conclusion
This is the first-known study to compare the parenting styles across different ethnic groups in South Africa and more specifically in post-apartheid South Africa. There were significant differences found between and within the different groups, even though authoritative parenting of mothers was the most prevalent parenting style across the groups. This study also highlights parenting style of fathers in comparison to mothers. Although the study focuses on ethnicity, it would seem that the role of children's gender is intertwined with the parenting of mothers and fathers.