Girls’ primary education and transition to secondary school in Nairobi: perceptions of community members at the onset of an education intervention

Abstract The objective of the paper is to examine the challenges of girls’ education in a poor urban context in Nairobi. This paper is motivated by the need to identify challenges affecting girls’ education in poor urban context in Nairobi. These challenges are identified from the lens of the community gatekeepers and the parents, who are key stakeholders in education at the local level. Data come from a qualitative component of the baseline study conducted in Korogocho and Viwandani, in Nairobi, collected in June and July 2013 using in-depth interviews and focus group discussions (FGDs). One of the most striking finding was role modelling to their daughters. This was evident from those households that had single parents. The study emphasizes that children should be better raised with the support of the whole community. This would enable parents to provide effective support to girls’ education and hence realize their full potential.


Introduction
Education enables people to make reasoned choices and provides them greater opportunities. It alleviates poverty and disease, and gives people a stronger voice in their respective societies. For nation states, education creates an enlightened workforce and citizens who are well adjusted -having the ability to compete as well as cooperate in a global environment, thereby enabling greater socioeconomic development (World Bank, 2004).
The global meeting on Thematic Consultation on Education in the post-2015 development agenda held in 18-19 March 2013 in Dakar, Senegal re-asserted that education is not only a human right but it is also the basis for the realization of all other rights. Moreover, significant development benefits accrue because education is a powerful empowerment tool. For instance, research evidence suggests that each year spent in school translates into a 10% increase in a country's potential income and a 1% increase in a country's Gross Domestic Product (GDP). Therefore, good-quality education, is key to achieving development priorities such as eradicating poverty and overcoming inequality across the globe (UNESCO 2013), particularly for girls. Providing girls with an education is a fundamental right. As has been documented since a decade ago, education enables girls to develop their human capacities, in addition to enabling them to make significant contributions to their nations in return leading to the ARTICLE HISTORY received 4 March 2016 accepted 29 april 2016 KEYWORDS challenges; primary; education; slums; parents; girls advancement of these nations economically (Abuya, 2010;Oxfam International, 2000;Klasen, 2002). Evidence has shown a catalytic effect that is an outcome of educating women and girls.

Challenges that girls continue to face
Significant progress has been made towards the goal of universal primary education, increasing enrolment rates and providing opportunities to many children who had previously been excluded from school (Ngware, Oketch, Ezeh & Mudege, 2009;Oketch, Mutisya, Ngware & Ezeh, 2010a). However, access to school for children has not been matched by quality of learning (Oketch & Somerset, 2007), even when governments emphasize access and transition. More children completing primary education has increased demand by students and families for secondary education worldwide (lewin, 2009), and particularly for girls and disadvantaged groups (Abuya, Onsomu and Moore, 2012). In Kenya, out of the pupils who sit the KCPE exam, only 48 per cent make it to secondary school, and fewer manage to graduate out of secondary education (UNESCO, 2012). However girls and women continue to be disadvantaged, particularly girls who live in marginalized areas. Marginalization occurs in several ways: those adolescent girls who are excluded from participating in mainstream education and programmes targeted at the youth. Such girls will tend to be illiterate; are susceptible to early marriage and early childbirth; belong to minority and ethnic groups; and live with or away from their parents in marginalized rural or urban slums (Sewall-Menon, Bruce, Austrian, et al. 2012). For instance, in the urban slums, APHRC longitudinal data show that in 2009/2010, children in poor households were less likely to make a transition into secondary school, with 52.1per cent of children making a transition in the poorest households compared to 57.8 in the middle poor and 60.8 per cent in the average households in the slums.
It is certain that girls continue to be critical in the demographics for social change and global development agenda, yet, they represent a large and underserved population in the developing countries (Warner et al., 2012). For example, girls in the urban informal settlements continue to be disadvantaged in terms of access and retention in secondary education. We focus on the informal settlements because APHRC education research data show that about 47% of children in the urban informal settlements attend non-state primary schools. For instance, in Nairobi 63% of children attend non-state primary schools (Ngware, Abuya, Admassu, Mutisya and Musyoka, 2013). Non-state schools are often staffed by teachers who are paid by parents. Moreover, research evidence shows that adolescents living in the slums are more likely to initiate sexual activity at an earlier age, thereby interfering with their schooling. For instance, adolescents who live in the slums in Nairobi initiate sexual activity about 3 years earlier than their peers in the non-slums (Kabiru, Beguy, Ndungwa, Zulu, & Jessor, 2012). Therefore, primary schooling and subsequent transition to secondary school for girls in the urban informal settlements remains a challenge that needs further investigation. This paper is motivated by the need to document these challenges that affect girls primary education in an urban informal setting, an example of a zone of exclusion. Since most of the literature is from the global north; these findings in the context of Nairobi will seek to shed light on the challenges that afflict girls from the perceptions of their parents and community leaders who are the gatekeepers in Korogocho and Viwandani. The paper seeks to answer the following questions: What are the challenges affecting girls' education in the slum of Nairobi, from the perspective of parents and community gatekeepers; and what are the potential solutions to these challenges?

Benefits of girls making a transition from primary to secondary school
Transition to secondary school provides the opportunities for young people to actively participate in the global economy, acquisition of civic skills, and social cohesion among communities (Schurmann, 2009). Particularly for girls, relevant secondary education brings well-demonstrated benefits and is correlated with related benefits for the whole society (Rihani, 2006;Hervish & Feldman-Jacobs, 2011;UNESCO, 2012). Research shows that when women have higher education levels, there is a lower fertility rate, family size becomes smaller, and health and economic status of women are stronger (Hervish & Feldman-Jacobs, 2011). For instance, each additional year that a mother spends in school reduces the odds of infant death between 5 and 10% (UNESCO, 2013).
In terms of economic growth, it is estimated that a one per cent increase in the proportions of women enrolled in secondary school, generates .3 per cent growth in the annual per-capita income (Dollar & Gatti, 1999), and boosts a girl's future earning by a margin of between 10-20 per cent (World Bank, 2004). In a recent study by Barro and lee (2013) education of girls was found to be a strong predictor of differential economic growth across countries. For example, there is a significant effect on output as a result of schooling, with a 5 to 12 per cent rate of return increase on economic growth for each year of schooling on the population. Other scholars argue that the quality of education is even more important in determining economic growth of countries, which goes along to enhance the demand for education (King & Winthrop, 2015;Sperling & Winthrop, 2015). Recent research validates the pioneer work of Shultz, (2002), who established that investments in women's education should be the focus of many countries who are seeking to advance human welfare and economic growth.
Girls' secondary education is also associated with increased social benefits to the society through increased political and civic participation, reduced instances of sexual harassment, and a lowered probability of young women being trafficked for labour and for sex. Secondary education offers girls greater ability to deal with HIV/AIDS by enhancing their levels of information not only to prevent the disease, but also to alter the way of thinking, and indirectly increase the chances of adopting self-protective behaviour (Rihani, 2006). With secondary education and beyond, girls get a renewed sense of responsibility -one that enables them to take charge of shaping their own future, without leaving it in the hands of their fathers or future husbands (Murphy & Carr, 2007).
Therefore, making a transition to, and obtaining secondary education remains a crucial goal in the agenda for development providing the opportunities for young people, particularly girls, to actively participate in the global economy, acquisition of civic skills, and social cohesion among communities (Schurmann, 2009).
Moreover, increased female education has also been associated with increased agricultural productivity in the context of SSA. For instance, in Kenya research has found that more years of female education is more likely to lead to improved agricultural output by 25 per cent (IFPRI, 2005). Research evidence further shows that educated women are more likely to adopt improved seeds and fertilizers, and the reason that they often do not is because of lower education levels (Saito, Hailu, Daphne, 1994).

Conceptual Framework
In conceptualizing the hindrances and educational resilience of adolescent girls, we adopted the risk and resilience framework. This framework seeks to explain how children and young adults in hardship and distress exhibit positive academic outcomes, emotional attributes, and show social competence (Nettles, Mucherah, & Jones, 2000). On one side of the framework are the risk factors that predispose pupils to vulnerabilities or unfavourable outcomes (Murray, 2003). The opposite side of the framework have resilient factors -protective factors that reduce the chances that an adverse outcome will occur, even when children, pupils and young adults are exposed to life stressors (Murray, 2003). Resilience implies that children can overcome severe hardship and have positive adult outcomes (Murray 2003). In the following section, I elaborate the meaning of risk and resilience, explain the relationship between risk and protective factors, and discuss the ecological dimension of risk and resilience.

Risk
'Risk' as a concept has been defined differently by different scholars. Risk is a composite term that comprises those stressors and vulnerabilities that are common to groups (Gordon & Song, 1994;Baldry & Farrington, 2005). In the case of this study, risk is viewed as the challenges that girls face in an attempt to navigate through primary schools and make a transition to secondary school. For the purpose of this study, I adapted Gore & Eckenrode's definition of risk (1994), which referred to risk as the presence of various life events and the effects of these life events on various groups.

Resilience
There is no consensus on how to define the concept of resilience (Arrington & Wilson, 2000). Some of scholars have defined resilience as: 'the lack of development impairment, or adaptation, despite exposure to risk' (Taylor, 1994, cited in Arrington & Wilson, 2000, preserving positive adjustment in the face of adverse life events (Cohler, Scott, & Musick, 1995); being able to succeed despite the stress levels (Bartelt, 1994); the ability to adapt despite the manifestation of risk (Arrington & Wilson, 2000).
A key tenet of the construct resilience is the existence of risks and promotive factors, which improve positive outcomes or reduce the probability that negative outcomes will occur (Fergus & Zimmerman, 2004). However, Winfield (1994) posits that resilience is a process resulting from an individual's reaction to risk or vulnerabilities inherent in their environment. Therefore, the protective factors present in both the individual and the environment enhance the process of resilience. Implied in this definition is the fact that resilience is improved by an individual, the interaction with the environment(s), and the manifestation of protective processes that cushion the individual from the impact of risk.
In adapting this model to explain girls' primary schooling in the two slums of Nairobi, we concur with scholars who argue that a relationship exists between 'risk factors and protective factors' (Murray 2003, p. 18). For instance, girls from disadvantaged households -a risk factor-still have the capacity to 'exhibit positive outcomes' despite the adversity brought about by poverty in the urban informal settlements (Fergus & Zimmerman, 2004). Therefore, the interaction of the individual girls with their families, schools (and peers in the same school), and communities is bound to shape their aspirations and educational outcomes. In the context of sub-Sharan Africa and specifically Kenya, community factors that influence the participation of girls in school include; rural or urban location, level of urbanization, distances to urban centres, local economic activities (Ampiah & Adu-yeboah, 2009;Abuya, Onsomu and Moore, 2012). There is a well-established impact of household factors on girls' school participation, and that there might be a significant relationship between the communities where children, and particularly girls live and their educational outcomes (Gumus, 2014;Adams & Hannum 2005;Chudgar 2008;Connelly & Zheng 2003). It is for this reason that school participation and subsequent resilience of girls in school, in the context of the urban informal settlements has to take into account the contextual factors. In essence, the resilience of girls in school, is an outcome of the interaction of individuals, households, schools and community factors. The differences in the quality of families, the school (and peers in the same school), and the communities may either be protective or negate girls' primary schooling. When these factors are negative they become a challenge; when they are protective, they promote adjustment to life stressors and persistence in school. (See Figure 1).

Study sites
This study was nested in the Nairobi Urban Health Demographic System (NUHDSS), which has been operating since 2002. The NUHDSS is a longitudinal platform in two slums in Nairobi -Korogocho and Viwandani -managed by the African Population and Health Research Center (APHRC). The two slum areas are densely populated (63,318 and 52,583 inhabitants per square km, respectively), and are characterized by poor housing, lack of basic infrastructure, violence, insecurity, high unemployment rates, and poor health indicators (APHRC, 2002a;2002b). The NUHDSS involves a systematic recording of vital demographic events including births, deaths and migrations occurring among residents of all households in the surveillance area thrice a year since 2002. Individuals qualify to become Demographic Surveillance System (DSS) residents either through baseline enumeration, in-migration or birth (Emina et al., 2011) The NUDSS also provides a sampling framework and a platform for nesting other studies. It keeps a database of all dwelling units and their geographical position (GIS coordinates), households, and individuals in the two slums under study. Each of these data points is assigned an identifier that enables tracking in case of movement, updates, and linking with other information collected within the surveillance framework. The information is usually updated every 4 months (3 times a year) (APHRC, 2002a(APHRC, , 2002b.

Overall study sampling
This study had both a quantitative and qualitative component of the evaluation. Overall, the study had a sample of 1421 households, targeting a maximum of 2132 girls. The project targeted girls in the two urban settlements with a view to increase their after-school support through the use of peer mentors. In addition, parents of girls were targeted through community conversations in order to ensure that girls were given the necessary support by their parents and community leaders. At baseline, a total of 1134 households participated in the intervention study, with 1270 girls. The intervention that was to be implemented was three-pronged: after-school learning support and mentoring for girls; a subsidized primary to secondary school transition; and parent and community leader sensitization on girls' education. This was implemented in two treatment sites; treatment 1 (T1) and treatment 2 (T2). In treatment 1 girls would be exposed to all the three components of the intervention that was implemented, while in T 2 girls would be exposed to only after-school learning support and mentoring for girls; and a primary to secondary school transition subsidy. This paper is specific to the qualitative component that was focusing on identifying the challenges of girls with schooling from the perspectives of parents and community leaders, at baseline before the onset of the intervention.

Qualitative sampling
The qualitative component randomly sampled 120 parents and guardians from the populated list of recruited households which was generated to mobilize participants for the focus group discussion (FGDs) at baseline. Out of the 120 parents sampled, 71 participated in the FGDs. We conducted 6 FGDs in the two sites, 3 in each (T1, T2, and C), respectively. Both men and women participated. Specifically, of the 38 FGD participants in Korogocho, 15 were men and 23 women. In Viwandani, 22 were men and 11 women (see Table 1). Furthermore, 12 in-depth interviews (IDIs) were conducted; six in each site, and all the key informants were men. The IDI protocol investigated community gatekeepers' understanding of their role and that of the community in the education of girls and the challenges that affect girls' education. The discussions were recorded to ensure that all of the data were captured.

Data source and collection
Data collection approval was through the Institutional Review Board at Kenya Medical Research Institute (KEMRI). The main method for data collection was of focus group discussions and IDIs. FGDs were conducted by a trained moderator and an assistant moderator between 12 June and 15 July 2013. The FGD and IDIs were conducted in specific sites already identified within the respective communities. The FGDs lasted a minimum of one to one and a half hours, while the IDIs lasted on average 45 min. All interviews were tape recorded for later transcription and analysis. The discussions began with a short conversation to establish rapport, and to have a formal introduction by all the participants, as well as the moderator and assistant moderator. The moderator led the discussion based on a series of questions in the protocol FGD (See Table 2), and the In-depth interview guide (See Table 3) which were designed to investigate parental understanding of their role and that of the community and community leaders' views towards the education of girls and to understand the challenges that affect girls' education in the two urban informal settlements where an education intervention is being implemented.

Analytical strategy
This study used NVivo software to facilitate storage and working with data. We generated codes from the research question: what are the challenges affecting girls' education in the slum of Nairobi from the perspective of parents and community gatekeepers? A review of the transcripts enabled us to identify Focus group discussion guide (1) to investigate parental understanding of their role and that of the community towards the education of their daughters (2) to understand the challenges that affect girls' education in the two urban informal settlements where an education intervention is being implemented (1) Give an account of your daughter (s) school attendance to date (2) What in your view has encouraged you to keep your daughter in school? (a) What issues have you been found to be a challenge if any? (6) What is your expectation of your daughter being that she is involved in this program?
the relevant codes that would emerge from the chunks of data, looking either for phrases that occurred frequently as affecting the education of girls in the slum context -either Korogocho and Viwandaniwhile having an eye for unique occurrences within the data (Miles & Huberman, 1994;Maxwell, 2005). The first reading of transcripts was to familiarize with the responses and to gain insights and clues as to what was contained in the data and thereby enrich the inductive codes (Maxwell, 2005). In reading the transcripts, we were looking out for the issues and concepts that characterized risk, challenges and vulnerabilities that impair the ability of girls to continue with their primary school, and subsequently joining secondary schools. In the subsequent readings, we were looking for the ideas, phrases, concepts and words that were most pronounced in the data; for example, peer pressure, puberty, parental role modelling, domestic chores, and poverty (poverty as thematic category is described in a separate paper). These words and phrases formed the root of the themes that emerged from the data. Subsequently, the initial codes allowed us to tentatively group the data, make descriptions, and extract quotes from the data chunks to support the emerging categories, based on the patterns and interpretations given to a code or sets of codes. This is what Patton (2002) describes as developing multidimensional categories to give the researcher a basic frame of analysis. The data chunks were grouped in matrix format (Miles and Huberman, 1994) in order to highlight the differences between the two sites, and between the treatment 1, treatment 2 and the control. The main thematic areas highlighted in this section of the challenges affecting girls' education include: inadequate parental monitoring; poor role modelling, prevalence of domestic chores, and onset of pubertal issues.

Results
Among the key risks to schooling challenges identified in the informal settlement prior to the onset of the education intervention included: inadequate parental monitoring; peer pressure and puberty; poor parental role models; and domestic chores. The key protective factors in the opinion of the parents and community members included: Harnessing community social capital; and periodic meetings of parents to discuss challenges affecting girls. In this section, we start with a discussion around the risk factors and follow with a discussion of the protective factors as highlighted by the parents and community members.

Inadequate parental monitoring
Parental and community gatekeeper perceptions of parental efforts to ensure girls complete the primary cycle of schooling is interpreted as effective monitoring, which includes but is not limited to the Table 3. the in-depth interview guide questions.

In-depth interview protocol
(1) to investigate the community gatekeepers understanding of their role and that of the community towards the education of girls (2) to understand the challenges that affect girls' education in the two urban informal settlements where an education intervention is being implemented (1) What is your role as a village elder in this community?
(2) What is your general understanding of children's education in this community? amount of attention parents give to their daughters, the tracking mechanisms they use to keep their daughters in check, and their ability to structure contexts for their daughters. Inadequate parental monitoring as a risk to girls was explicitly expressed in Korogocho than in Viwandani. Community gatekeepers were most explicit about challenges emanating from the parents, while parents were more explicit about the challenges emanating from the girls, especially puberty and peer pressure, which impair adolescent girls' ability to seek and internalize the advice of both parents and teachers. The narratives showed that girls in Korogocho experience deep-rooted challenges in attempting to complete primary school. The community gatekeepers in Korogocho explained that by the nature of their daily economic activities, parents are away from their homes for extended periods of time. This absence impairs their ability to effectively monitor their children's school attendance and the work covered in school. An elder explained: Most parents leave early in the morning so they are unable to monitor whether the child went to school and arrived on time or didn't go at the right time… Does the child attend school regularly or not? … There are parents who leave at 5 A.M. in the morning and come back at 8 P.M. or 9 P.M. This parent is unable to check the child's school books, so monitoring is very poor, especially for the girl child… (IDI, Village Elder Koch) At times, parents were forced to rent separate houses for their adolescent children, girls included, and making monitoring what they are doing during the day and, worse, at night impossible.

A Korogocho elder summed it up: 'For instance, where I live is in a one-room house, and my child, who is in class six or seven -I will rent a house for [him/her] some distance away from me. So they go there, and I cannot know what transpires between where I live and where they are…' (IDI Village Elder, Koch)
Poor parental role models ….many of them start being involved with men at an early age. What contributes to this is, as I have said, when the mother comes to live in the slum with her children. The house measures ten by ten. The child is aged seventeen, and the mother comes home with her male friend. The children sleep on the floor while the mother is on the bed with her friend. Don't you think this exposes the child to such behavior, yet she is only aged sixteen or seventeen? (IDI, Village Elder, Viwa) This statement by a village elder in Viwandani sums up the significance of a parent's actions in either building or destroying her daughter. His question exposes the link between the structural deficiencies in the slum, the family process, maternal behaviour, and the possible outcome for an adolescent girl who is supposed to be attending school. The community gatekeepers in the two urban sites agreed that the parents of the girls in the study were not good role models. It appeared mothers' behaviour was singled out as exposing girls to early sex and multiple sexual partners.
Moreover, parents felt that some of them engage in sex when the children are watching due to the constraints of their living arrangements. Their consensus was that the children will try to emulate what the parents are doing. A father of a girl in the Viwandani control site said: Again, here in Nairobi, the houses we live in are very small. There are men who aren't good…you start having sex earlier, and the children are in the same room and haven't even fallen asleep. The children will also want to try and imitate the parents (Male Respondent, Male FGD, Viwa, Control).

Peer Pressure
According to Rose and Rudolf (2006), while boys experience peer stress in the form of explicit verbal or physical victimization, girls are exposed to friendship stress, not taking into account conflict with their best friend, as self-reported by girls. Boys generally experience less stress than girls, and this difference intensifies in adulthood. Peer interactions among girls can have intense ramifications if the stress increases to the extent that they are not able to cope with both their social networks and the pressures of school. Community gatekeepers' narratives underscore this effect in the sense that girls whose parents are not keen on school attendance may influence other girls in the neighbourhood to stop attending school altogether. A Korogocho elder said: …And also peer pressure. you find that your child is going to school, but the neighbor's doesn't attend school… the neighbor's child encourages your own not to go to school, telling her to go somewhere else. So peer pressure contributes a lot… (IDI, Village Elder, Koch) Moreover, peers in the neighbourhood lead other children into criminal activities, such as thuggery. Girls struck close relations with boys who were already part of the criminal gangs in the community. With time, they became the leaders, luring unsuspecting friends to the criminal gangs. A Korogocho elder lamented: …yes, since they associate with the boys…who do the planning? At times they even bring boys from town to come and steal from people here in the community. The girl brings two boys and calls her colleagues claiming that I have customers. [All laugh] The customer comes and is mugged here, and the girl is beaten… a few slaps to make it look real, and a deal is sealed… (IDI, Village Elder, Koch) In Viwandani, the gatekeepers related the effect of peers on early sexual debut, which led to prostitution. An elder observed, ' …you know that when girls reach a certain stage, they get a boyfriend and tell their fellow girls who might not already know… the dangers of early sex. They might influence each other…. ' In addition, those girls who have already engaged in sex lure and teach other girls in the community not only to enter into sexual relations but to perfect them as an everyday tool of trade. They rent houses in the community and lure girls to come and entertain men. In the views of the community gatekeepers, the girls who are most easily lured are orphans and those from single-parent households.
Concurrent with the concern expressed by community gatekeepers, parents also felt that their daughters' friends and peers may jeopardize their daughters' attempt to complete primary school. Peers can influence girls negatively, and the outcome can be early pregnancy, which means the end of school for the girl. The mother in Korogocho said: In my view, girls face a lot of challenges when learning in this Korogocho schools. Mine really wanted to learn, but she got friends who disturbed and misled her, teaching her bad habits until she got pregnant. I wish you could help me get her another school so that she may continue with her studies… (A Female Respondent, Female FGD, Koch, Treatment 2) The parents in Viwandani had a similar opinion: that peers can jeopardize their daughters' chances of completing school, particularly when the friends and peers are not attending school. The mother of a treatment 1 girl in Viwandani noted: Girls in this community are influenced…by the girls who aren't attending school. If they become friends…she wants to go out with her friends…. Even these girls who go to the streets, it is not that the parents don't talk to them, but it is their peers' influence. (Female Respondent, Female FGD, Viwa, Treatment 1)

Onset of puberty
Parental narratives showed that when some adolescent girls undergo biological changes, their attitudes towards their parents changed. They became out rightly rude and felt that their parents were in direct conflict with whatever they wished to spend their time on. A father in Korogocho said: The main challenge is when the girls start developing breasts. They start becoming rude to the parents; they see that they are grown because the boys start following them. Also, the issue of health is a challenge because when they get to 13 years they can now become pregnant… (Male Respondent, Male FGD, Koch, Treatment 1) For some parents, the onset of puberty begins at a specific age. Suddenly, they are not able to keep pace with what their daughter does, and eventually the girl succumbs to peer pressure, which may lead to early pregnancy. A mother puts it this way: The child goes to school well, but it reaches a certain age where things change…As a parent, you can't be with the girl all the time. The girl herself wants to learn, but it reaches a point where they are cheated by boys and are impregnated. As a parent, you lose faith and no longer want to educate that girl… (A Female FGD, Female Respondent, Koch, treatment 2) Parents in Viwandani agreed that at the onset of puberty, people's perception of girls' changes; many think she is a grown-up, presumably because of the physical changes. A mother of a treatment 1 girl in Viwandani said, ' …if the girl is of age, others look at her as a grown-up woman, but you as a parent know she is still a child. Men look at her as a woman, so a mother has to check when the girl left school and arrived home after how long' .

Domestic chores
When girls try to combine school with household chores and looking after younger siblings, they spend more time than boys in activities that are not related to school (Coclough, Rose, & Tembon, 2000), and their school attendance is affected. Community gatekeepers pointed out this challenge, while parents were very silent about it. The chief in Korogocho explained the girls' involvement in domestic work and what he sees as its effect on their education: …I had said that girls in Korogocho are vulnerable and what I mean is when they go home from school in the evening, they have to wash clothes, clean the house, they cook, do homework, fetch water, fetch firewood for those who use it, they go to purchase paraffin, they go to Korogocho open market to buy vegetables [emphasis added]. So you can imagine a girl leaving school at five and doing all that, including homework, before sleeping…. (IDI, Chief, Koch) An elder in Viwandani concurred. He believes that domestic chores do not leave the girls the time they need to attend to their schooling activities. He explained: yes, the girls are normally given a lot of work when they come from school… kitchen work, house chores, washing of clothes, looking after the young ones and ensuring they have eaten. …And this is a very big challenge that highly affects the girl's performance in school. (IDI, Elder, Viwa) Protective factors that will make girls resilient Some of the protective factors that were put forward by parents and community members broadly encompassed encouraging parental and community support for girls' education. Some were already in the pipeline, and the parents were involved in them. Their outstanding suggestions were harnessing community social capital and holding periodic meetings for parents to discuss the challenges that girls experience.

Harnessing community social capital
Many parents believed that children are raised better with the support of the whole community than by individual parents. Stronger community partnership would enable parents to support each other and particularly single parents who have to fend for their children alone. The parents agreed that if responsibility for children's upbringing were collective, then the absence of a parent would not be an excuse for a daughter to misbehave. In such a case, other parents in the neighbourhood would take charge. A father of a girl explained: you know, in the past, the child belonged to the community and not one person. So if I see a child making a mistake, I should be allowed to punish that child and then take them to the parent… (Male Respondent, Male FGD, Koch, Treatment 1) The men in Viwandani who represented the control zone had similar sentiments. They felt that raising a child should be a concerted effort, and parents should be open to each other and not suspicious of those willing to correct children in the neighbourhood when they make mistakes. One explained: …I want to agree with this man. Today, responsibility for raising your children is yours alone, not like in the past, where you see a girl misbehaving, and you report to her parents… The community should work together to raise the children. (Male Respondent, Male FGD, Viwa, Control)

Periodic meetings of parents to discuss challenges affecting girls
In addition to the community coming together to keep girls focused on school, parents also believed they should hold regular meetings to discuss the challenges affecting their daughters. They felt that such discussions would allow them, particularly mothers, to brainstorm about solutions to the common problems their daughters faced. A Father said: …If we can come up with a group, and meet like once in a month, then we discuss the challenges that we are fac-ing…It will help…I think we need to sit as members of this community and discuss our children… (Male Respondent, Male FGD, Koch, Treatment 1) These sentiments were shared by parents representing treatment 2 in Korogocho. They felt that they should also interact more with teachers, who are the custodians of their daughters for a longer period of time in any given school day. A mother explained: We should sit…talk to mothers with girl children. We should discuss the issues and see how to solve these issues and look for one answer…We should talk to our teachers since they stay with the children the whole day and know them well… (Female Respondent, Female FGD, Treatment 2)

Discussion
The objective of this paper was to establish the risk factors to girls' education in the slums of Nairobi from the perspective of parents and community gatekeepers; and what the parents and community members thought were the protective factors to some of the risks? The study found that among the key risks that girls encountered in the slums included: inadequate parental monitoring, poor parental role modelling, and onset of puberty. Inadequate parental monitoring was as a result of the nature of their daily economic activities, which kept them away from their homes for extended periods of time. In essence, parents were not able to pay enough attention to their daughters, and maintain the tight tracking mechanisms necessary to keep their daughters on track for school. This finding is similar to the findings of scholars (Dishion & McMahon, 1998;Borawski, levers-landis, lovergreen, & Trapl, 2003;Hill & Tyson, 2009) who have found that parental monitoring and tracking are essential for adolescent behaviour and for quality parental-children relationships. This finding underscores the importance of this study at the onset of an education intervention, since it reinforces that girls still face similar challenges towards the close of the MDG goals.
Moreover, this study found that inadequate parental monitoring made parents unable to structure contexts for their daughters -and their ability to effectively monitor their children's school attendance and the work covered in school. This finding suggests that parents in the informal urban settlements whose schedules are too tight, and being away from home for long periods of time, inadvertently miss out on opportunities to get involved in their daughters' schooling. This finding is similar to lawson's (2003) who posits that lower class parents who do not have the privileges of being in jobs with flexible schedules miss out on the opportunities to get involved with their children's education. However, this finding raises a significant question, which is that some of these parents would still have higher expectations for their daughters' academic achievement. In which case, they would still be involved parents, albeit not being around to assist with homework (Wilder, 2014). However, an extant body of research has shown that there is a positive effect of parental involvement at school (Epstein & Sanders, 2002;Hill, 2001;Hill & Craft, 2003;lareau, 1987), with children achievement. This means that even if the parents were not at home and they found time to go their daughters' respective schools, they still can contribute to their daughters learning process. Parental narratives also showed that poverty inhibited their ability to fulfil their obligations to their daughters. Girls' needs for food, lighting, uniforms, school fees, and oil for their skin were often unmet ), making it difficult for them to attend school and jeopardizing their education. The perception among the people in Korogocho that they are poor seems to have contributed to their ambivalent attitudes towards their daughters' education. The internalization of this belief discouraged them from engaging in activities, economic or otherwise, that would change their own and their daughters' circumstances.
This study suggests that parents and community gatekeepers believed that within the urban slum context, girls' education cannot be the prerogative of parents alone. Parents and community gatekeepers saw a link between themselves and the wider community to ensure that girls attend school. Their narrative of a community united to educate girls, was identified as a key protective factor -meaning that there was need for a certain level of community social capital for better education of girls. This implies that schools cannot do their job without the support of the other stakeholders, key among them community members. Rather, there is need for a multipronged approach bringing the three key stakeholders around the girls to work together in order to enable girls to attend and complete school. Key stakeholders included; parents in the households, schools attended by girls, and the communities around the schools. Both parents and community stakeholders thought that a communication model -"community-communication-knowledge" -would be useful. It would encourage a knowledge flow; the whole community would be aware of girls' peers, and, at any given time, a community member would be able to offer guidance to a girl to help her reap the benefits of education. This would be very beneficial even in the absence of the parents for long hours during the day.
The study findings show that parents were not good and effective role models. Consequently, this corrupted the girls in terms of them engaging in risky sexual behaviour, brewing and drinking alcohol. This was made worse in single-parent households. Their assumption was that single parents are unable to monitor the children effectively while scrambling for resources to sustain the household. This finding reinforces the classical study (Bandura, 1977b) which underscores the importance of the social learning theory. This theory emphasizes the role of modelling in acquiring and maintaining certain behaviour (Bandura, 1977b). Thus, parental modelling of sexual behaviour is important for adolescents. This study raises a lingering notion that it is only in single-parent households that are headed by mothers, where the monitoring is not adhered to. This notion calls for an investigation of the differences that exist between the parent households headed by men and women. This study was limited in the sense that it covered two urban informal settlements in Nairobi. We may not be aware of what challenges, and community perceptions in other urban informal settlements in Nairobi, or other parts of Kenya.
This study had one significant policy implication in relation to the education of girls in a poor urban slum context. The study found that girls face a myriad of risks in an attempt to navigate through the primary school years. In order for girls to be able to complete their primary cycle of education, schools cannot do it alone. That there is need for an even stronger school community partnership for success of girls' in school. Moreover, communities' social capital should be harnessed to support girls who come from single-parent households. The study suggests parental involvement with daughters both in their respective schools and at home by tracking, and monitoring their whereabouts, an important ingredient for school attendance. Further research would be able to tease out the differences in community members' perception and support to girls education in selected rural areas of Kenya. In conclusion, the quality of education is important even for girls who come from the urban slum context. Risks may exist but if the parents and the schools work together, these challenges can be minimized. Most importantly, when goals are shared, and relationships become trustworthy, with mutual respect, and complementarity (Christenson, Rounds, and Franklin, 1992;Minke & Anderson, 2005), family school, and community collaboration becomes more beneficial to the pupils, beyond parental involvement. Therefore, gaining support from families to support the education of their daughters go hand in hand with providing the necessary support to families and learning from these families on the best ways in which they can be supported to offer the necessary support to their children.

Funding
This work was supported by the Anonymous [4435].

B. A. Abuya is a research scientist at the African Population and Health
Research Center (APHRC), in Nairobi, Kenya. Her research interests include comparative education, education policy issues of access, equity, and quality; the linkages between education and health outcomes for adolescents; gender issues in education; the impact of families on the education, resources allocation, and health outcomes of their school going children; and intimate partner violence. W. M. Ngware is a Senior Research Scientist and Head of the Education Program at APHRC. He designs research projects, conducts policy relevant research in the areas of access and quality of education with a special focus on learning outcomes. He joined APHRC in 2007 prior to which he was a policy analyst at the Kenya Institute for Public Policy Research and Analysis and also served as a Senior lecturer and Chairman of the Department of Education Administration and Planning at Egerton University, Kenya. Ngware is driven by success and passion to influence change that can lead to improvements in social outcomes particularly among marginalized populations in sub-Saharan Africa.
M. Mutisya works with the African Population and Health Research Center, in the Education Research Program. He is currently a CARTA fellow pursuing his PhD in Public Health at the University of the Witwatersrand South Africa. He has over 7 years extensive experience in data management, and analysis of both qualitative and quantitative data, with special interest in mixed methods. He has special interest in understanding the linkages between education and health. Maurice has authored and co-authored a number of articles on Education and Health in international peer-reviewed journals.
M. Nyariro is a research officer in the Education Research Programme at APHRC. Her research interests include quality of life of women and girls in the society particularly those living in the urban informal settlements. She is now a PhD student at McGill University, Canada.