Personal characteristics that predict South Africans ’ participation in activities of their children ’ s schools

We investigated the extent to which personal characteristics such as age, marital status, education level, living standard measure (LSM), environmental milieu, race, gender and employment status predict parents’ participation in the activities of their children’s schools. The data used for analysis were drawn from 5,734 South Africans aged 16+ years who participated in the Human Sciences Research Council’s (HSRC) annual South African Social Attitudes Survey (SASAS). Respondents (n = 1,364) who had at least one child in school were selected for the analysis. The logistic regression model was used to estimate the extent to which personal characteristics would predict parents’ level of participation. Results showed that personal characteristics such as age, marital status, gender, and living standard measure (LSM) had a significant influence on parental participation. Amongst these, gender was the most significant factor. Respondents aged 35–49 years were 1.5 times more likely than those aged 50+ years to participate in their children’s school activities. Respondents who were married were more likely than those who had never married to participate. Mothers were more likely than fathers to participate in their children’s school activities. The literature suggests that American fathers show some interest in participating in activities of their children’s schools, at least in the first few years, but discontinue that participation in time. The results confirm a common belief in some communities in South Africa that fathers do not take seriously their responsibility to provide for their children and their partners in that regard from the onset. Even though the results on personal characteristics support existing theory and empirical literature on parental participation, a multidimensional perspective is needed that would incorporate other factors into the study.


Introduction
Parents' participation in activities of their children's schools is one of the most important facets of successful education.Children gain better scores and improve in academic achievement when their parents work hand-in-hand with the school (Grolnick, Benjet, Kurowski & Apostoleris, 1997;Goldring & Shapira, 1996).Besides improving their academic skills, parental participation also enhances children's social and interpersonal relations with the entire schooling system that includes their classmates and school personnel.Corner and Haynes (1991) have also asserted that the educational process of children is enhanced when parents get involved in the education of their children while at the same time a good personal relation is developed with their parents and other stakeholders outside the school.In this way, parental involvement creates a partnership between the school and the home of the child, and makes the education of children sufficiently benefiting to all stakeholders involved in the education of the children.Our main objective in this study was to investigate the extent to which personal characteristics such as age, marital status, education level, living standard measure (LSM), environmental milieu, race, gender and employment status can predict level of parents' participation in activities of their children's schools.In the next section a review of studies on parental participation in activities of their children's schools will be presented.
According to Heystek and Louw (1999), parents as clients of the school have little say in the school governance, and as such a partnership-oriented approach is necessary to allow them to be proactive in their children's education.Parents, as the most important actors in the child's family, play an important role as mediators between the child and other parts of society like the school.Thus, their involvement in the child's education would help accomplish the educational responsibilities of transmitting the culture and social interests of the child to other parts of the community (Agabrian, 2007).Specifically, their participation in activities of the child's school would elevate not just the cultural identity of their children, but would also facilitate their socialisation, attitudes and behaviours towards those around them.
It is likely that age, marital status and gender of the parent interact with other factors to influence participation levels of parents in their children's school activities.Internationally, research suggests that personal characteristics such as age, environmental milieu, gender, race, ethnicity, marital status, level of parents schooling, personal monthly income, and employment status determine the level of parental participation in the activities of their children's school (Epstein, 1990;Ho Sui-Chu & Willms, 1996).

Parental participation
Low parental participation in activities of the school has been detected in South African black schools in recent years.Though some of the causes are attributed to various factors including negative attitudes and inferior feelings of parents (Heystek, 2003), most research findings have linked the causes to socio-economic status of families.Existing research evidence indicates that socio-economic status of parents has more influence on parental participation than other variables such as age, gender and marital status (Astone & McLanahan, 1991;Epstein, 1986;Fehrmann, Keith & Reimers, 1987;Grolnick et al., 1997;Stevenson & Baker, 1987).Others, such as Ho Sui-Chu and Willms (1996), note that the education level of parents themselves is more important than other personal characteristics variables.Due to changing attitudes and behaviours, growing economies and advancing socio-political positions today, it is highly likely that parents' behaviours towards education of their children could be influenced by multiple socio-personal characteristics than a unitary aspect such as the parents' social class.
In education there are different ways in which the notion of parental participation is understood; the term participation appears to be used interchangeably with the term involvement.Authors, such as Sy, Rowley and Schulenberg (2007), view parental involvement as primarily parents' participation in school activities.These school activities include volunteering in the classroom, attending school events, open houses, and back-to-school nights.Unlike this international view of participation, the South African approach to participation follows from the stipulations of the South African Schools Act (SASA) whose view is more aligned with school governance and representation.For example, South African researchers like Heystek (2003), Heystek and Louw (1999), and Lewis and Naidoo (2004) emphasise the need for parents to actively participate in school governance.International literature seems to use the notion of "involvement" whereas South African literature uses the notion of "participation".In our study, parental participation implied that parents get involved, participate and attend their children's school meetings, sport activities, leisure activities and social activities.In this context the term participation is used synonymously with involvement.In broader terms, parental participation implies that parents fulfill the mandate of being responsible for their children and participating in activities of their children's schools.
Parental participation is used across different fields.For example, in the field of medical sciences the term is simply understood as the act of parents being involved in their children's care (Eckle & MacLean, 2001).This kind of involvement includes parents who participate in their children's medical check-ups where they may need to consult with medical personnel about their children's health.In the context of education, the shift is evident from a client-oriented type to a more interactive partnership between various stakeholders within the education sector.Though other stakeholders such as the school district or education officials can play a role in determining the child's education, the immediate stakeholders like teachers, the principal, parents or legal guardian of the child are more important as they are directly involved with the individual child.Smit, Moerel and Sleegers (1999) assert that parental participation in the educational context is when parents actively participate in their children's education.'Active participation' denotes that parents work closely with the school as partners in their children's education rather than as passive clients who only have contact with the school when they come to pay school fees or called on by a teacher regarding progress of the child.

A multidimensional understanding of parental participation
Factors that influence parental participation are multidimensional; however, few studies have made the effort to integrate these factors, or explain how they affect parental participation (Grolnick et al., 1997).Although personal characteristics play a vital role in predicting levels of participation in parents, they are nonetheless only indicators of situations where participation can be high or low.There are other factors such as psychological characteristics, daily psychosocial stressors, attitudes, and behaviour towards schooling that explain participation.
Interaction between different overlapping factors that affect parental participation can be explained using an ecological model.According to Merrick (1995), an ecological model stresses the importance of the interaction between people and their changing environments.While the relationship between people themselves within the family context is something family systematic theorists are striving to understand, the larger picture in the schooling environment is more complex.With the ecological model, relationships are understood from the standpoint of reciprocity rather than that of linearity, and that they are not solely placed on unitary sources, but combined into various sources of influence (Lubbers, Van der Werf, Kuyper & Offringa, 2006).The school and community should be understood as a single body rather than as separate bodies with a different agenda in society.This would prevent further research finding discrepancies within a single body.It does not necessarily help much if some researchers find problems to be within the teachers of a particular schooling system, while others find fault within families alone.In this case, to understand problems associated with schooling, the problems facing the community should be understood as well.The incorporation of multiple factors, systems approach and their multilevel analysis into existing and known personal characteristics can add value to the understanding of the phenomenon of parental participation.

Psychological characteristics
Psychological factors such as stress, depression, anxiety, unstable moods, beliefs, attitudes and personality characteristics of parents could be linked to parents' participation levels.Some analysts (Sy et al., 2007;Grolnick et al., 1997) have included psychological characteristics to explain and understand the aetiology of parental participation.Even though a parent's and a child's distinct psychological characteristics in some ways affect each other, research has only concentrated on the psychology of the parents and paid little attention to the psychology of the child (Guner, Cabrera, & Shannon, 2005;Ambert, 2001).The psychological status of the child as an individual who interacts with the parents in the family should also be considered.Children can have an effect on why parents have a low level of participation in the activities of their children's school.Children might have contact disorders or can act disruptively for the attention of their parents.In the process the parents' belief in their children's educational attainment can be hampered.
In South Africa, like in high income countries such as Australia and Britain, most parents have faith in education and wish their children to be educated (Kivilu & Morrow, 2006).However, everyday reality conflicts with their educational faith for their children and survival of their families (Samuel, 2005).For example, as high as 48.5% per adult equivalent is below the poverty line of R354 per month (Frye, 2006).The reality of poverty and unemployment amongst most South Africans precipitates strong psychological problems that follow from the history of unequal education and economic opportunities in the country.As stated in Belle (1990), there is a strong correlation between some serious and common psychological problems, such as depression, anxiety and substance abuse, and negative psychosocial stressors such as poverty and unemployment.To add to such psychosocial challenges, parents' daily distresses of school fees, poor school environment, violence within the schools and poor relationship with teachers and the entire school system are other factors that contribute to a complicated interplay between psychological displacements and the parents' responses to demands of education today.These psychological characteristics are part of the complex interplay of multiple factors influencing levels of parental participation.

Personal characteristics
While most research studies have concentrated on the socio-economic status of parents as key contributors to parental participation (Ho Sui-Chu & Willms, 1996;Stevenson & Baker, 1987), other variables may be involved, or even play a major and more significant role than is understood.
South Africa has been experiencing an increase in the number of teenage mothers in recent years (Swartz, 2002).By virtue of their young age, lack of proper formal education, their vulnerability to life-threatening illnesses, poor living conditions and daily distresses, these young mothers may not be ready compared to their older counterparts in terms of participation in the educational processes of their children.Whereas older mothers are employed, have support of a male partner, have more experience in parenting and are selfsufficient in making their own decisions, young mothers are more likely to have low parental efficacy mainly because of single parenthood and other psychosocial and psychological challenges.Various researchers have linked teenage motherhood to poverty (Rich-Edwards, 2002), poor mental health (Thompson & Peebles-Wilkins, 1992) and low education as compared to older mothers (Kalil & Danziger, 2000).Other research has disputed a mother's age in itself as a predictor of school-related problems of their children (Hollander, 2001).The blame is attributable to the socio-economic and personal characteristics that place the mother at a disadvantage.Other personal characteristics considered relevant here are marital status, a mother's level of education and race.
With regard to marital status, there is a dearth of literature on marriage patterns in South Africa due to difficulties in obtaining and interpreting appropriate data (Budlender, Chobokoane & Semelane, 2004).For this reason, few studies support the marital status of parents as a prominent predictor of parental participation in the activities of their children's school.Little is associated with the education system and how parents in different family patterns respond to the education of their children, except education used here as a variable indicating level of parental schooling.

Sample
Data from the HSRC's South African Social Attitudes Survey (SASAS) 2005 project were analysed.SASAS is conducted annually using a stratified national household sample.The respondents were sampled regardless of their nationality or citizenship using the HSRC's Master Sample.According to Pillay et al. (2006) the HSRC's Master Sample consist of 1,000 census enumeration areas (EAs) with 11 visiting points being selected in each area so as to give a total of 11,000 visiting areas.A sample of 500 EAs was drawn stratified by socio-personal characteristics domains of the province, geographical sub-type and four racial groups found in South Africa.Fourteen (14) households were drawn from each EA using the systematic sampling technique and one adult aged 16 years and above was selected from each household for an interview.Although the original sample was n = 5,734, only 1,364 respondents who indicated that they had children in school (any level) were used for the analysis.

Data collection and analysis
Questionnaires developed to measure respondents' attitudes, beliefs and opinions about participation in activities of their children's school were piloted, translated and administered face-to-face to the respondents to obtain the pertinent information.Apart from biographical information respondents were asked to respond on a 4-point scale question about 'how often do you participate or did participate in the activities of your children's schools?'The scale consisted of very often=1, often=2, sometimes=3, and almost never=4.For the purposes of logistic regression, we further dichotomised the 4 points into high participation (very often and often=1) and low participation (almost never and sometimes=0).Forty-two percent reported low participation while 58% reported high participation.
All independent variables under study (age, marital status, level of education, living standard measure, environmental milieu, race, gender and employment status) were ordinal categorical data, except for age, which was continuous.Age was categorised into four groups depicting the developmental phases measured in years completed.Descriptive statistics were used to describe the sample and key variables.

Results
Of the 1,364 selected for analysis, 30% (n = 954) were female while 70% (n = 410) were male respondents.Only 4% were in the 16-24 years age category, 20% in the 25-34 years, 43% in the 35-49 years and 33% in the 50+ year age category.Fifty-six percent were black Africans, 15% Coloureds, 15% were Indian/Asian and 14% white.Respondents were equally distributed among the three categories of LSM with 33% being in the low and medium, respectively, and 34% in the high LSM.Details of the descriptive statistics for the personal characteristics and those of level of parental participation are summarised in Table 1.

Binary logistic model
To estimate the extent to which independent variables could predict parental participation in the activities of their children's schools, a binary logistic regression model was developed and tested using levels of participation as a dependent variable.Results of a goodness of fit test presented in Table 2 show that the chi-square value is not significant and conclude that the model adequately fitted the data and that at least one of the predictors is significantly related to the measure of parental participation.Table 3 presents a summary of the model in terms of the amount of variance in the parental participation variable that is explained by the predictor variables.The predictor variables accounted for 11.3% of the amount of variance in the dependent variable (parental participation).The results show that other variables not included in the model are responsible in the remaining proportion of the variance.A summary of results from the binary logistic regression model are presented in Table 4.
Overall, in order of the magnitude of the Wald coefficient, respondents' age category, marital status, gender and LSM were the only predictors that had a significant influence on levels of participation in activities of their children's school.Respondents in the middle age category of 35-49 years were 1.5 times more likely than those in the 50+ category to participate in their children's schooling while respondents in the 25-34 years category were 1.3 times more likely than those in the 50+ category to participate.Respondents in the 16-24 years age category had the lowest chance of participating in their children's schooling than those in the other three age categories.
Respondents who were married were 1.4 times more likely than those who were 'never married' to participate in the activities of their children's school.Married respondents were also more likely to participate than a widower or widow and the divorced or separated.Female respondents compared to their .393male counterparts, were more likely to participate in the activities of their children's school.Respondents in the low and medium LSM were less likely than those in high LSM to participate activities of their children's school.This is even more apparent in low LSM categories.

Discussion
The results of this study are consistent with international research findings.Grolnick et al. (1997) have found that mothers are more involved in their children's lives than fathers.Fathers in two-parent families are less likely to be involved in their children's education than fathers in single-parent families (Nord, Brimhall & West, 1997).In studies by Lewis, Maka and Papacosta (1997) the discontinuity hypothesis also confirmed that fathers are most likely to withdraw involvement in their children when they separate from spouses, while those by Hardesty and Ganong (2006) blamed such discontinuity on contact avoidance between spouses.Though some of the fathers measured in this study showed permanent involvement, the tendency for low participation in other groups still showed father involvement as a problem.Richter (2005) supports the notion that most young fathers in South Africa lack responsibility to make financial provisions at least to their children and their partners.Adding to their lack of responsibility, cases of physical and sexual abuse have been recorded as some of the most significant reasons that lead them to be absent in their children's lives (Richter, 2005).
Though in this study we did not formulate a hypothesis to test differences between age groups, we found that parents who were within the ages 35-49 were more likely than the younger age group to participate.It is difficult to locate the reason in this context to cultural differences because race did not prove to affect level of participation.Time spent at work by fathers can be blamed in this regard.Men are more likely to be employed than women in South Africa (Hassim, 2005).Younger men are less likely to be employed than their older counterparts.As stated by Davies (2000), families of a young mother can force the young father to be involved in the life of his child through customs where monies for reparation and damage can be charged.For non-African young fathers, it is usually the support from their family that makes them responsible for their children.
The results in this study show that respondents in the high LSM are more likely to participate than those in the low and medium LSM.This is consistent with existing literature which suggests that socio-economic class determines parental participation.Ho Sui-Chu & Willms, (1996) argue that parents from low socio-economic background tend to place less emphasis on schooling than those from middle social class.Likewise, South Africans from lower social class have shown less determination towards participation in their children's school than those from higher class.
The study has shown that married couples are more likely to participate in their children's school activities than those who were never married.Marital status of parents signifies not only changes in the fertility and marriage patterns among South Africans, but has come to be the most significant predictor of level of parental participation, a finding consistent with research elsewhere.Ho Sui-Chu and Willms (1996) found that marital status had an impact on parental participation.Astone and McLanahan (1991) note that single parents compared to married parents are less inclined to participate in their children's schools.The analysts agree that being a single parent comes with more responsibility that not only affects the parent's well-being, but also affects the child and the family.In the United States, mothers who have just divorced, separated or have lost a spouse are determined to economically make it through life in spite of being without their spouses.This is because they are more active than men in the labour market than before (Bradbury & Katz, 2002).Unlike the economically empowered women in the United States, South African women are still among the poorest of the poor and largely depend on their spouses for basic needs.Forty-nine percent of woman-headed households were found to be among the poorest in 1995 (Baden, Hassim & Meintjes, 1999).
Whereas the focus was traditionally put on socio-economic status rather than any other personal characteristics as the most significant predictor variable (Ho Sui-Chu & Willms, 1996), the results in our study show that the multiple interaction and interplay of factors have an impact on parental participation.The findings are also in line with the suggestion by Grolnick et al. (1997) that parental involvement is no longer conceived as a single phenomenon, but that a multidimensional perspective is needed that incorporates emotional and personal aspects.This suggests that parental participation should be approached ecologically so that various aspects, as opposed to unitary aspects, can be explored.

Conclusion
We began by introducing the concept of parental participation in the activities of their children's schools.We have discussed how the concept has been understood internationally and locally and presented empirical data on parental participation in schools.One suggestion that emerges from the findings is that there is a need to increase parental participation in order to improve on children's education.
To increase participation of parents, more direct impact programmes and interventions aimed at enhancing relationship between parents, pupils and teachers, other than their governance, are needed.For example, to increase participation of parents in the No Child Left Behind Program (NCLB), an American programme aiming at offering low-performing schools opportunity to switch schools, several key recommendations were made including engaging schools to provide information to parents and making parents more aware of the need for improvement in education (Vernez & Li, 2009).The same recommendations can also apply as a strategy to increase parental participation in their children's education and school activities in South Africa.Making parents more aware of the need to improve education for their children not only increases their participation in their children's schools, but can also promote and improve the image of school and standard of education in general.

Table 1
Descriptive statistics of the respondents' personal characteristics

Table 2
Hosmer and Lemeshow test (Goodness of fit test)

Table 3
Model summary

Table 4
Results of the binary logistic regression model