The significance of student voice: female students’ interpretations of failure in Tanzanian secondary education

In Tanzania, the national examinations are used as the primary tools for selection and transition from lower to upper secondary education. Female students are more likely to fail in the national exams and to drop out from education. This article examines the perspectives of female students concerning their advancement in secondary education. Two sets of qualitative data, responses to a research questionnaire from 100 female lower secondary school students and follow-up interviews with seven mature students enrolled in a non-formal school who had failed in the national examinations, were analysed to identify critical issues influencing the educational advancement of female students in an urban context of Dar es Salaam. Finally, the connections between students' perspectives and the national secondary education development plan are examined. The aim is to highlight how student voice can contribute to transform secondary education and achieve the equality and equity targets.


Introduction
When you have 177,021 out of 354,042 failing, as it was the case in 2010 national Form 4 exams, then you know the nation is in trouble, more so when you consider that the pass mark is 21 per cent! (The Citizen 26 November 2012, Tanzania) The recent poor results in the national lower secondary education examinations have generated heated debates in Tanzania. During the last years, there has been a dramatic decline in the results of final examinations in lower secondary education (Form 4) and the performance of girls has decreased proportionally more than that of boys. The pass rate in Form 4 examinations has steadily decreased from 86.15% in 2007 to 53.37% in 2011. In 2011, the pass rate for female students was 48.25% compared with 57.51% of male students (URT 2012). The national examinations are used as the primary tools for selection and transition. For students who aim to proceed to upper secondary education, (Forms 5 and 6) the Form 4 examinations form a bottleneck. 1 Transitions from primary to secondary and from lower to upper secondary education are critical stages for girls, in particular, and the dropout rates of girls in secondary education are higher than those of boys (Okkolin, Lehtomäki, and Bhalalusesa 2010;Posti-Ahokas and Palojoki 2014). A significant proportion of students in secondary schools are overaged due to repetition and drop out caused by financial constraints within families and the school system and the low quality of education provision 2 (e.g. Carr-Hill and Ndalichako 2005;URT 2011). In Tanzanian cities, the rapidly increasing private schools offer second chances for those who have dropped out of the formal system. This new phenomenon in Tanzania has stimulated public discussion on the role of national examinations and concerns over the quality of secondary education (e.g. Hartwig 2013). In the broader context of Sub-Saharan Africa, the situation in Tanzania looks serious, because the overall secondary education provision and enrolment rates are the lowest in the region (see, e.g. Al-Samarrai and Reilly 2008).
In Tanzania, the Ministry of Education and non-governmental organisations (NGOs) have analysed reasons for the dramatically deteriorating results with the focus on administration and management, school environment and the examination procedures, teacher-related factors and curriculum implementation (HakiElimu 2011a;URT and UNESCO 2012). What seems missing from these analyses are the perspectives of students, female students in particular, who are most affected by the situation. Recent research in two neighbouring countries, Malawi (Kamwendo 2010) and Uganda (Jones 2011) shows that insights gained from ethnographic studies on female students can unfold the construction of gender and gendered experiences in education and thereby contribute to the achievement of equity targets. Previous research further suggests that student voice is essential for the assessment of educational opportunities and critical aspects when aiming for meaningful educational change (Mitra and Gross 2009;Yonezawa, Jones, and Joselowsky 2009;Warrington and Kiragu 2012). Inspired by the evident potential of student voice work (McLeod 2011), and student engagement for prevention of dropout (Yonezawa, Jones, and Joselowsky 2009), this paper focuses on the female students' views on secondary education.
The high failure rate in lower secondary school national examinations triggered a heated public debate in Tanzania on how the 'nation is in trouble'. Less attention was given to the fact that most of those failed in exams were female students who, therefore, are at a greater risk of dropping out of secondary education. The specific adversities, including discrimination in and out of school, early marriage and pregnancy, that female students face at the critical transitions between levels of education need to be better understood to move towards gender-responsive strategies and practices (Chabaan and Cunningham 2010;Okkolin, Lehtomäki, and Bhalalusesa 2010). This study investigates the examination failure and needs for change in secondary education from the perspective of female students. Special attention is given to views of overaged students who are persistently working towards their personal goals in the margins of the education system (e.g. Warrington and Kiragu 2012). The connections between students' perspectives and the national secondary education development plan (SEDP) are examined to highlight how student voice can contribute to transform secondary education and achieve the equality and equity targets.
Gender equality in Tanzanian secondary education For the past decade, the development of secondary education in Tanzania has been guided by the national SEDPs 2004-2010 (SEDP I) and 2011-2015 (SEDP II) aiming at building a high quality, productive, and adaptive labour force for the economy to contribute to the realisation of the national development vision. The five key areas of the SEDP II include: (1) improvement of quality and relevance, (2) enhancement of access and equity, (3) improvement of the teaching force and teaching process, (4) improving management efficiency and good governance and (5) institutionalisation of cross-cutting issues, such as gender and HIV/AIDS awareness. SEDP defines strategic measures with the focus on the development of the education system and processes. The current SEDP recognises the lowered performance levels in examinations as an indicator of poor quality and, consequently, as a major issue to be addressed. One of the objectives of the SEDP is to improve the national examination assessment framework by reviewing the current status and the impact of examination practices on results, and by establishing an appropriate assessment and examination system, possibly with reference to changes in other countries in the region. Recently, Tanzanian media has paid attention to the further decline in the examination results since the beginning of the SEDP II period that has resulted in revision of the examination grading system and nullification of the examination results of 2012 (The Citizen Tanzania, 30 and 31 May 2013).
The current SEDP emphasises improving girls' participation and performance in the enhancement of access and equity in secondary education. While the gender parity has been reached in Tanzanian primary education, boys outnumber girls in secondary ducation and in the upper secondary level, the number of girls significantly declines (Al-Samarrai and Reilly 2008;Okkolin, Lehtomäki, and Bhalalusesa 2010;URT 2011). In 2011, the gross enrolment rate in lower secondary education was 45.9% for girls and 54.5% for boys, decreasing to 3.5% and 6.5% in upper secondary education. In terms of retention, the gross completion rate in 2011 for lower secondary education (Form 4) was 30.9% for females and 40.7% for males and, respectively, 3.0% and 5.4% for upper secondary education (Form 6). Furthermore, girls' performance in Form 4 examinations is significantly lower than that of boys throughout subjects (10-20 percentage points in subject pass rates), with the exception of Kiswahili language, where girls outperform boys. Consequently, the statistics suggest that girls remain remarkably disadvantaged with regard to access, advancement, and educational outcomes (URT 2011).
The poor quality is associated with the rapid expansion of access to secondary education, which in turn has had a negative impact on educational equality (Wedgwood 2007). The SEDP recognises the widening gender gap in participation and the decline in examination results, and explicitly targets retention and examination performance of girls. Promoting retention, improving performance in science subjects and enhancing transition of female students to upper secondary education are defined as critical issues. Retention target of 90% is to be reached through providing girlfriendly facilities (latrines and hostels), empowering girls' club activities, revising school rules and regulations and stakeholder sensitisation through media. Girls' examination performance is to be enhanced by providing remedial tuition, user-friendly learning materials and science camps (URT 2010). The experiences and views of girls and young women are not, however, suggested as a measure for improving educational advancement and learning outcomes.
In Tanzania, as in other low-income countries in Sub-Saharan Africa, girls' educational attainment is tied to a complex set of issues ranging from policy to socio-economic and sociocultural factors (e.g. Okkolin 2013). In Tanzanian families and communities, poverty and cultural constructions of gender have been identified as major causes of gender-based discrimination in access to education (Vavrus 2002(Vavrus , 2005. Furthermore, early pregnancy is a major risk for educational advancement of secondary school age girls (Stambach 2000;Helgesson 2006;URT 2011;Posti-Ahokas 2013). Within secondary education, analysis of the constructions of gender in the teaching and learning process (Thomas and Rugambwa 2011) and the gendered school environment (Sommer 2010) have shown how efforts to remove discrimination against females are challenged by the complex discriminative discourses and practices operating within school contexts. Egbo (2005) has argued that African girls are socialised to discriminatory gender role norms that are further reinforced when women themselves adopt and maintain a 'culture of silence' that excludes them from active participation in society. In this study, we privilege the voices of female students to better understand how they see the complex processes influencing their advancement in education.
Potential of student voice for educational change Grounded in critical research tradition and radical pedagogy (see Taylor and Robinson 2009), student voice is defined as processes that enable students to be consulted on their education (Czerniawski 2012) with the aim of increasing students' representation and participation in processes from which they have historically been excluded Robinson 2009). McLeod (2011, 181) identifies four common overlapping uses of voice in educational discourse: (1) voice-as-strategy (to achieve empowerment, transformation, and equality); (2) voice-as-participation (in learning and in democratic processes); (3) voice-as-right (to be heard and to have a say); and (4) voice-as-difference (to promote inclusion, respect diversity, and indicate equity). In this study, the female students' voice is understood as a significant contribution to the debate on secondary education, in McLeod's terms voice-as-right, in particular, to be heard and have a say concerning advancement and learning outcomes in secondary education in Tanzania.
As a political concept connected to justice and equality, voice is concerned with questions of power and knowledge, inclusion and exclusion and of being advantaged and disadvantaged (Thomson 2011). The popularity of student voice work within different educational systems has attracted significant academic interest and debate on its applicability in different contexts (for discussion, see Czerniawski 2012). Along with the idealised, emancipatory purposes of student voice work, practices to involve students are often reduced to solely serve the aims of organisational efficiency while losing the potential for democratic development and transformation (Fielding 2004;Gunter and Thomson 2007). Critchley (2003) argues that in student voice work for policy development, students are commonly used as a source of data rather than as policy-makers. Unterhalter (2012) has reminded of the importance of adequate contextual understanding for transformative use of girls' voice in African contexts, where young people's voices are easily silenced, respecting, as Kiragu et al. (2011) explain, the longstanding traditions of who talks, who listens and which topics are considered taboo. In Kenya, Jwan and Ong'ondo (2009) explored perceptions of students concerning participation in secondary schools, and found that the students felt their participation was peripheral and their contributions were not well articulated in school decisions. Practices that the teachers and principals considered to be in the interests of students were viewed differently by the students, even in direct contrast to the expectations of the teachers and principals (Jwan 2011). The involvement of students in decision-making was considered uncomfortable for teachers (Jwan 2011). Juffermans and Van Camp (2013) showed how their analysis using student voice on language policy in Gambian education has contributed to the critical evaluation of policy by unfolding the reasons for students' enthusiasm for learning in English despite the minimal learning outcomes. In her ethnography on girls' secondary education in Uganda, Jones (2011) concluded that the girls' views offered insights into their specific daily challenges and this knowledge is necessary for policy-making, if education planners are committed to achieving gender balance.
In the context of Tanzania, there is little documented evidence of student voice work within secondary schools. TUSEME ('Let Us Speak Out') clubs in secondary schools are the most well known and widespread initiatives aimed at empowering girls and tackling gender-related problems. Initially, an NGO initiative, TUSEME has now been scaled up by the Ministry of Education and Vocational Training and, where active, the clubs are forums for student participation in school improvement (FAWE 2004;URT 2010). Additionally, students have been involved as co-researchers to develop models for including students in curriculum development for HIV/AIDS prevention (Rutagumirwa and Kamuzora 2006;McLaughlin and Swartz 2011) and for science teaching and teacher training (Semali and Mehta 2012). Methodological explorations with participatory group activities (Sommer 2010), reflective writing and photography (Miles 2011), empathy-based stories (Posti-Ahokas 2013), and photo voice (Nguvu ya Sauti 2012) have been conducted to study the sociocultural context beyond school from the students' perspectives.
Previous analyses of student voice work call attention to issues of trust, inclusion, dialogue, power, and participation. Taylor and Robinson (2009) argue for the importance of recognising plural-and context-specific relations of power and participation. Similarly, Czerniawski (2012) emphasises the existence of different forms of professional trust in different national locations. To include the other, silent and, perhaps, controversial voices is a requirement for achieving the potential of student voice work that can increase social justice (Robinson and Taylor 2007). This is particularly relevant to the research involving student voice in African contexts where inequality and marginalisation in education are critical issues.
Previous research from Tanzania shows that research can be designed to give voice to students who are often left out of discussions around education (Sommer 2010;Billings 2011;McLaughlin and Swartz 2011;Posti-Ahokas 2013;Lehtomäki et al., 2014). Fielding (2009) sees the potential in research approaches involving student voice to encourage, enable, and liberateto remake the present and revision the future. Similarly, Cook-Sather (2006) argues for participatory research to counteract policy-making and practices that systematically exclude students' perspectives. Fielding (2001) outlines a four-fold typology of practices involving student voice: (a) students as data source, (b) students as active respondents, (c) students as co-researchers, and (d) students as researchers, indicating different levels of participation and authority of students where the last two are seen to have a stronger transformative potential. In this study, female students are seen as active respondents who reason and analyse their own educational advancement.

Methods
The research was conducted as part of a broader study in three public and in two private schools 3 in Dar es Salaam, the largest city in Tanzania (Posti-Ahokas and Palojoki 2014). Participants were selected from different types of schools to include the perspectives of girls representing a variety of socio-economic backgrounds. Data on female students' assessment of their own opportunities to proceed to upper secondary education were collected using a semi-structured research questionnaire on personal plans and future aspirations, distributed to and responded by 100 female students enrolled in Form 3, in December 2008. The students were asked about factors that they felt influenced their opportunities to realise their plans to advance in education. The focus of the analysis was to identify factors influencing advancement and retention in education, as seen by students. To engage the voices of female students who had dropped out and faced the risk of marginalisation, follow-up interviews were conducted a year later with seven young women aged between 23 and 28 years who were enrolled in a non-formal school, 4 which mainly served mature students who had failed in the national examinations and, consequently, dropped out of the formal system. This group represents, as Warrington and Kiragu (2012) define, those who against odds continue their education and, therefore, provide essential information about the education system's functionings. The students interviewed talked about their extended transition experience, the reasons for the generally low examination results, and their perspectives on how the performance of students and the educational system in general could be improved. In the results section, participants of the questionnaire study (N = 100) are referred to as 'Form 3 students' and the seven interviewees of the follow-up study as 'mature students'.
The data collection was conducted in Kiswahili, the official language in Tanzania, the medium of instruction in primary education and the lingua franca in Dar es Salaam. In secondary education, the official medium of instruction is English, although, in practice, Kiswahili is often used in teaching (see Billings 2011 for a discussion on language use and acquisition in Tanzanian secondary schools). The qualitative content analysis of SEDP II (URT 2010) focused on the critical issues identified during preparation of SEDP II, and the plan's objectives and strategies regarding the enhancement of access and equity. Finally, connections between the female students' perspectives and the policy articulations were examined.

Advancement in education
In the responses to the questionnaire (N = 100), 131 mentions of the factors influencing advancement were coded and categorised as positive (contributing) or negative (constraining). The mentions were further grouped into self-related, school-related, and sociocultural factors. The distribution between the factors and categories is presented in Table 1.
Considering their situation prior to the final examinations in Form 4, the students emphasised the role of sociocultural factors, particularly family, in determining their transition to Form 5 (55% of all mentions). The role of economical and psychological support provided by the (extended) family was crucial: Because my parents very much like to see me acquire adequate education and they do provide me with every service that I need, even though they themselves attained a simple formal education. In contrast, almost half of the respondents doubted if they could continue at all, mainly due to financial constraints in their families: These expectations will never materialise because the guardians I am living with are not (financially) capable … Until right now, I am studying under problemsschool contributions are a bother. I therefore think that I may completely not be able even to continue due to the condition of my guardiansthey are elderly and unable. However, where support from the family was available, students took the responsibility for succeeding themselves (39%): I will study with the utmost diligence, and since my whole heart is in it, I will simply succeed.
Our analysis of the students' perspectives to advancement in education suggests that school-related factors play only a marginal role in students' considerations. Only 6% of the reasons mentioned for attainment were directly associated with school or government support. Eight school-related issues, three of which were negative effects of the schooling system, were mentioned in the responses. Students relied on support by the Government in providing opportunities for quality education I know the Government of Tanzania will assist me. Peer support and being close to teachers were considered important for personal success. When considering their own opportunities to proceed to upper secondary education, three respondents saw high enrolment and competition as threats to their own advancement: there are many of us studying in Form Four and there is very stiff competition in education nowadays, so, in fact I am not certain whether I will attain my dream.
These findings are in accordance with our previous findings (Posti-Ahokas 2013) on the importance of sociocultural factors (support from parents and extended family, peer support, and religion), and the very limited role of school-related factors in students' considerations regarding advancement in education. While in school, persistence, effort, and effective learning strategies are seen to greatly determine success. Similarly, previous studies in Kenya (Kakenya 2011) and Tanzania (Okkolin 2013) have shown that educated women consider their success as a combination of personal qualities, familial support, and school-related factors. In this study, the limited role of schoolrelated considerations in the thinking of the Form 3 students places pressure on the female students themselves as the major cause for success or the lack of it. Internalising and personalising the lack of success caused by structural constraints can result in creation of a false sense of failure (Ansell 2004;Crivello 2011). Therefore, participation in evaluation and improvement of the learning environment at school (e.g. Sommer 2010) could both increase female students' awareness on the structural constraints to  (131) advancement in education and also contribute to dialogue required for improvement of equality in education.

Voices of the 'failures' in Form 4 examinations
At the time of the individual follow-up interviews, the students, families, and schools throughout Tanzania were digesting the shockingly low examination results that also received significant media attention. We asked students who had dropped out of the formal educational system what, according to them, had caused the low examination results, and what, from their perspective, could be done to improve the performance of students and the educational system in general. In addition to the individual interviews with seven young women, we conducted complementary interviews with members of the school's student parliament, 5 the school principal, and the teachers to contextualise the individual experiences of the interviewed students. The interviewees were aged between 23 and 28, which indicate irregularities in their educational paths. On their first attempt in taking the examinations in different schools, all seven women had failed to secure the credits required to enrol in upper secondary education. Three of the women had already taken the Form 4 examination twice (in 2008 and 2009). In the academic year of 2008-2009, all seven had attended a oneyear programme of the non-formal school comprising essential contents of the Form 3 and 4 syllabuses, preparing them to sit for the national Form 4 examinations as private candidates in special examination centres. They had all taken the Form 4 examinations in November 2009 and had recently received the results when we conducted the interviews in March 2010. Two of them had passed the second time and proceeded to Form 5, three had re-registered for the preparatory programme, and two were considering whether and how to proceed.
Shock and disappointment regarding the examination results were shared by the five students who had not passed the examinations with adequate marks to proceed to Form 5. All of them thought they had been well prepared and were expecting good results. Nuria (age 28) describes the shared disappointment: Therefore, most of us think that results have not been good this year, and most of us did not so expect. We studied diligently, but now we do not know what the problem is. We do not understand, even though in our hearts and our understandings we deemed that we had performed well. Two of the interviewees had immediately re-enrolled in the same preparatory class and were planning to retake the examinations later the same year. Two were trying to find the financial resources to enrol in the preparatory class as soon as possible. Angelina (age 24) was now working and found her situation fairly hopeless: Actually, for now it pains me, that is, I having not ... I have done nothing. I still see that I have done nothing because my goals were attached to reaching Form Four, and then find a course to pursue. Therefore, for now I see that it is difficult, that is, I have not yet got any light.
The disappointment was also shared by the school principal, the teachers, and the members of the student parliament whom we interviewed. Obtaining a full record of the results was difficult because students were taking the examinations as private candidates in different examination centres. After the results came out, the student parliament analysed the situation together with the school administration. This resulted in the initiation of an additional, teacher-led revision session each morning.
Analysing the reasons for one's own low examination results was difficult for the students. However, they articulated some possible reasons that had led to the overall poor performance: Researcher: What do you think has caused the results to be so poor?
Kemi (age 23, passed the exam on second attempt): As for me, what has made the results to be undesirable is probably on our side as students. This is because we fail to read topics in the books written by the Ministry (of Education) or the Institute (of Education) for us to read. As a result we go and take books that people have already written, like those of Nyambari Nyagwini. Even last year, I did likewise. So you take and study the points; consequently, you find the points shallow, then you do not understand everything. When you go to write, you do not elaborate things adequately. It is just like those are the things misleading us students, reading things … you do not get into class for meaningful reading, you begin struggling for other things.
As described by Kemi above, high failure rates have increased the demand for concise 'question and answer' materials that are marketed as a quick fix solution to students' learning needs. Both the students and the teachers recognised the risks related to the popularity of such materials. However, it was also recognised that the school was not able to provide adequate reading materials to all students. Decisions are made in favour of the commercially produced materials that are often cheaper than the government approved materials. The school principal confirmed the increase in using alternative, exam-oriented books and expressed his concern over access to proper learning materials. Also, student parliament members saw the availability of materials as a key issue in improving students' performance.
The students expressed their concerns about the unequal position of students in government schools compared to those enrolled in private, non-formal schools. Abolition of examination fees from students in government schools was seen to increase inequalities between students enrolled in different types of schools. Although some of the private schools provide for the more privileged, many schools, including the one where we conducted our research, are enrolling students from the poorest segments of the society. Therefore, it was suggested by the student parliament that the government should exempt the poorest students from the examination fees. Furthermore, the status of private candidates in the National Examinations seems generally vulnerable.
Like the way we did our examinations, our centre had so much disorder. And that is what made them decide to nullify it this year; it is no longer there, that is, examinations will no longer be conducted there, because they breached the procedures of the (National) Examination Council. Many students were taken away and some transferred to churches and whatever, intensifying the disorder. So I would like the government to take heed on such issues. (Suzy, 28) When I see bad results, it is a big issue. Because we can blame Ministry of Education, in other side you can blame students. Because some students, for example I, I struggle with myself, I need to succeed but sometimes results can come negative … I think there are mistakes but because I'm not familiar with government issues that's why I can't give you full information. But I need to advice my fellow students -I used to struggle hard in order to escape these obstacles from the Education Ministry. (Eve,25) Similarly, the school principal considered the private candidates as victims of the system and said that the students' rights only seem to apply to students within the formal system. As a solution, the student parliament members called for more dialogue between the public and private education providers and the government to increase coherence in practices. As an example, more common tests for both school types were suggested to increase equal opportunities. Students were worried that they would be seen as second-class workers by the employers. Despite all problems, students saw the non-formal, fast-track provision as a necessity: There are so many Tanzanians who are poor … that's why all of us are not able to take all the years; Form, 1, 2, 3, and 4that's why we take a short cut. So I would like to say that the government should not stop this [programme]. (Interview with student parliament member) Connections between the students' views and the SEDP When assessing their own opportunities to proceed further in education, students emphasised the role of economical and psychological support provided by the (extended) family to greatly determine their transition. The students' perspectives of factors influencing advancement in education contrast with the objectives of the SEDP regarding access. The plan seems to place insufficient value on the role of families in supporting the secondary education of girls. The SEDP suggests using media, particularly newspaper articles, to sensitise stakeholders to strategies to promote the retention of girls. Our data from urban areas do not point to major problems in familial attitudes towards girls attaining higher levels of education, but rather reflect great commitment of families to support their daughters. However, previous research findings from Tanzania suggest that advocacy work is still needed to promote education of girls and women, particularly in rural communities (Stambach 2000;Vavrus 2005).
Due to the emphasis given by the students on the critical and controversial, both enabling and constraining, role of their families and community in their advancement in education, a further analysis of the SEDP documentation was done to identify policy articulations regarding school-community connections. Based on the analysis, the schools are seen to operate in isolation to a great extent. The community is mainly seen as a provider of income for the schools via various types of students' fees in both public and private schools. According to the latest Education Sector Analysis (URT and UNESCO 2012), household direct and indirect investments are equivalent to 21% of public recurrent education expenditure. This estimation does not include the significant contributions from the community to the construction of schools in terms of cash and labour. Furthermore, the SEDP designates new, significant responsibilities to communities for the construction of houses for teachers and hostels for students and for the provision of meals for students. Other than their contribution to fees and construction, the families and communities are not given any role in implementing the strategies of the SEDP. The only reference to networking with other governmental agencies and NGOs is found in strategies to establish care and support programmes for students and personnel affected by HIV/AIDS. Form 3 students and mature students viewed economic problems in their families, high enrolment numbers, and competition as threats to their own advancement in education. The disproportional expansion of lower secondary education has resulted in problems to accommodate the Form 4 leavers in further education. In 2010, the pass rate for Form 4 examinations was only 50.4% and only 4.3% gained the scores required for transition to Form 5 (URT 2011). The SEDP recognises the challenges resulting from the expansion and the increasing number of students from poor households. It includes a plan to expand provision of upper secondary education and a public scholarship scheme to assist youths from poor families. The qualitative evidence presented here regarding the scale of economic problems, even among students residing in the largest city of Tanzania, raises doubts about the ability of a small-scale scholarship scheme to make a difference. Even if all 80,000 eligible students availed of the scheme, scholarships would be available only to 5% of students enrolled in public secondary schools (calculated from enrolment figures for 2011 [URT 2011] and SEDP projections). The SEDP acknowledges the roles of private and nonformal schools in expanding access and the non-government schools are included in national enrolment and performance statistics. The non-profit making non-government schools are also eligible for the capitation grants at the rate of 50% of what is paid to government schools.
The Form 3 students emphasised the role of personal effort and learning strategies in advancement. The SEDP outlines the promotion of independent learning, as well as analytical and critical thinking skills of students, as objectives leading to improved performance in examinations. Evidence from studies conducted in government schools (Makombe et al. 2010;HakiElimu 2011a) suggests that active learning is an objective yet to be reached in practice. Both the mature students and the headmaster of the nonformal school criticised the increased use of alternative materials containing old exam questions and model answers. The limited access to textbooks and to other quality learning materials is also recognised in the SEDP. Resources for the provision of materials are drawn from school fees paid by the students and their families, in addition to capitation grants paid by the government. However, a monitoring report by a Tanzanian NGO (HakiElimu 2011b) suggested that capitation grants do not reach the schools on time and that the central government has not been able to budget for the capitation grants throughout the SEDP period.
The increasing failure rates have had severe impacts on the transitions of large number of students. The reliance on national examination results as the sole indicators of achievement results in prolonged transitions and generates an externally imposed sense of failure (Ansell 2004;Abdi and Cleghor 2005). The SEDP includes a plan to improve the national examination assessment framework. The further declining examination results and the heated public debates call for an urgent action. The interviewed mature students had experienced disruption in the examinations and had suffered from the nullification of their results. Taking examinations as a private candidate places students from non-formal schools in a more vulnerable position and leaves their schools with neither responsibility nor possibilities for a proper follow-up. The students enrolled in the private, non-formal institution regarded their status to be lower than that of the students in the public schools. These mature students called for more dialogue between public and private institutions and the government to increase the coherence and equality in secondary education. They valued the student parliament as a dynamic body intermediating issues between students and school administration. The current SEDP does not acknowledge student councils as part of the institutional arrangement and no data are available on their prevalence.
To summarise, the key findings regarding advancement in education, the transitions through Form 4 examinations and female students' suggestions for system improvement are presented in Table 2. Increasing enrolment and economic problems are a concern shared both by the female students and the SEDP. The deteriorating quality of secondary education is reflected in the decreasing examination results, public debates, and the lives of female students. According to the students, their (extended) families usually do their best to educate their daughters. Our findings suggest that more could be done to further benefit from the strong societal demand for female secondary education (see Billings 2011 for a discussion) and the valuable insights of female students who are the major targets of equity efforts of the SEDP (Table 2). Towards transformative student voice work in secondary education development From the perspective of these findings, critical issues are what is the potential of student voice work in opening up new opportunities for female students' engagement in education and improving their educational transitions in Tanzania, and how can this research contribute to educational planning and policy implementation. In this study, education policy is seen as a multi-level process of negotiation, contestation, and a struggle between competing groups (Ozga 2000). In common with previous studies on girls' voice in Tanzania (Billings 2011;Sommer 2010), the girls and young women in this study had several relevant recommendations on how to improve schools and the education system. Their perspectives highlight the importance of understanding advancement and transitions in secondary education as a combination of personal effort, family support, and the enabling structures of the education system and thereby complement the structural analysis on the construction of gender inequalities in education in Tanzania (e.g. Vavrus 2005;Chaaban and Cunningam 2010). The experiences of the mature students in non-formal school illustrate the impacts of deteriorating quality and poor performance on individuals and families. Examining the connections between the female students' perspectives and the current SEDP measures shows that though the development plan does to some extent recognise the critical issues raised by students regarding advancement and transitions, it ignores the issues beyond the immediate school structures. Vavrus (2002) has argued that rather than traditional attitudes, economic policies undermine opportunities of Tanzanian children to stay in school. Our study supports Vavrus's statement by suggesting that the most important issue for both the female students and the realisation of gender equality in secondary education is finances, significantly influencing quality and participation in secondary education. Therefore, further understanding of the complex connections between female students, their families, and communities and the school system is needed to effectively work towards equity and equality targets (see also Okkolin, Lehtomäki, and Bhalalusesa 2010;Jones 2011;Okkolin 2013).
The perspectives of mature students who are struggling to make the transition to upper secondary education provide valuable insights to the realities of female students within the examination-oriented system. The female students in the non-formal school who had previously failed in national exams had actively sought to re-enter school. However, they reported that as private candidates they did not receive equal treatment in the national exams. To include these silent and, perhaps, controversial voices is a requirement for achieving the potential of student voice work that can increase social justice (Robinson and Taylor 2007;Warrington and Kiragu 2012). This issue is particularly relevant to this research involving the voices of female students in Sub-Saharan African contexts where inequality and marginalisation in education are critical issues, particularly in post-primary education. Therefore, improving the examination system in a gender sensitive manner would greatly benefit from information gathered through the qualitative analysis of the experiences of female students in different school contexts and, as Vavrus (2005) suggested, in local settings beyond education.
When making recommendations, the participants referred to 'the government' instead of any levels in between, such as secondary school teachers, school leadership, and committees. This may mean that the students perceived the educational system as centrally controlled and viewed the teachers and leadership as powerless to change this system, e.g. how the examination process was designed and conducted. An alternative interpretation is that their teachers and school leadership pointed to 'the government' when trying to explain the evident failures of the secondary schooling due to the poor quality and massive expansion. This refers to what Taylor and Robinson (2009) highlighted as the importance of recognising plural-and context-specific relations of power and participation. Similarly, Czerniawski (2012) emphasised the existence of different forms of professional trust in different national locations. From the perspective of those female students who participated in this study, 'the government' needs information, but the students do not have mechanisms to speak back to policy. The findings provide a clear call for educational leadership and research to search for ways to include students' voices in dialogue at different levels of the educational system. They also highlight the need to redefine responsibilities for schooling and schools and to enforce collaboration between the government, different types of schools, NGOs, and community-level initiatives (see Phillips 2011 for a discussion on the tensions of democratising policy-making in Tanzania).
The mature female students and members of the student parliament called for including the voices of students in dialogic improvement of schools and the education system as a whole. According to Critchley (2003), for students to actively take part in development and implementation of education policy, their participation has to be formalised at all levels. In contrast, Green (2010) argues that the heavily institutionalised participatory planning mechanisms used in Tanzanian development activities have made participation to work against its intended benefits. Furthermore, social structures and cultural norms may limit students to voice their concerns (Unterhalter 2012), whether formalised or not. Therefore, introducing participatory approaches within the development of education should be done with caution to retain the transformative potential of student participation at all levels. We argue that qualitative research can greatly complement the formalised student voice by including the silent and marginalised voices, allowing for a rich contextual understanding.
Recognising the threats regarding misuse of student voice (Gunter and Thomson 2007) and lessons learnt from previous work done in African countries (Jwan and Ong'ondo 2009;McLaughlin and Swartz 2011;Unterhalter 2012;Posti-Ahokas 2013), we emphasise the importance of cultural and contextual understanding for successful implementation of student voice work. Involving students as co-researchers and creating opportunities for student-led research could enhance collaborative development in schools and contribute to the education of teachers. Previous experiences of involving Tanzanian students and teacher students in curriculum development (Rutagumirwa and Kamuzora 2006;McLaughlin and Swartz 2011;Semali and Mehta 2012) provide with encouraging experiences of student participation beyond serving as a source of data. Examining the student voice and the connections between their voice and the SEDP but failing to include the voice in the development dialogue with policy-makers is a limitation of this study. Inclusion in the dialogue is, however, the long-term plan of the broader research project that this study is a part of.
Based on her analysis of the potential of student voice in equity initiatives in higher education, McLeod (2011, 187) argues that rather than inciting students' voices, the challenge is to convert those voices into meaningful and practical recognition and actions through 'a politics of listening'. She further suggests that re-framing student voice in a politics of listening would result in recognition, engaged dialogue, and relevant action, thereby reviving the voice as a strategy for equity and inclusion. By highlighting female students' views on failure and advancement in secondary education, this paper contributes to the research dialogue, though further efforts are necessary to include the student voice in educational development in Tanzania. The government is committed to achieve the equity and inclusion targets in secondary education and here the politics of listening could provide a new strategy to enhance the desired transformation in education. The discriminative structures, policies, and practices within education place females in and out of school at a higher risk of marginalisation (Chabaan and Cunningham 2010;Warrington and Kiragu 2012). This study accords with the view that the perspectives of students can provide valuable, complementary perspectives to the complex problems related to equity and equality of education (Okkolin, Lehtomäki, and Bhalalusesa 2010;Jones 2011;Kiragu et al. 2011). Also the research community and NGOs can contribute with their findings and knowledge concerning student voice work. As recommended by the students, the 'government should listen'. Our argument is that listening to female students, both to those who succeed in national exams and those who fail and face the risk of dropping out from the formal educational system, opens a window to relevant recommendations from motivated young females, eager to learn and find ways towards meaningful learning. Notes 1. The Tanzanian education system consists of seven years of primary education (Standards I-VII, age group 7-13), followed by four years of lower secondary education (O-level, Forms I-IV, age group 14-17) and two years of upper secondary education (A level, Forms V-VI, age group 18-19). Primary education is public and free. Secondary education in both public and private schools is fee-based (URT and UNESCO 2012). Within the categories of public and private schools, there is a variety in resources and performance (Hartwig 2013;Posti-Ahokas and Palojoki 2014). Therefore, the division between public and private is not of relevance for this research. 2. In 2011, only 68.7% of O-level students and 39.9% of A-level students were enrolled at the right age (URT 2011). 3. The schools were given the labels of: (1) 'new community school', (2) 'established community school', (3) 'high performance government school', (4) 'established private school', and (5) 'NGO education centre'. Schools 1-3 are classified as public and schools 4-5 as private. All schools were following the same national curriculum. 4. School 5 'NGO education centre' is run by a non-governmental organisation and is classified as a private school. N.B. The status of private schools varies from high-performance and well-resourced ones to low-performance and poorly resourced institutions (see Hartwig 2013; Posti-Ahokas and Palojoki 2014). 5. The student parliament consists of three members elected through a general vote. The purpose of the student parliament's work is to link students with school administration, to help administration to solve problems, to assist in achieving results and to organise extracurricular activities, including academic clubs and study tours.