Introduction
Adolescence is a time of rapid change, not only in physical, cognitive and psychological competencies but also in social roles and expectations (Patton et al, 2012; Steinberg, 2015; Viner et al, 2015). Yet we still know relatively little about the patterning of these changes, and the types of support young people need in order to reach their full human capabilities (Patton et al, 2016). Given that many of the Sustainable Development Goals (SDGs) rest on investments in adolescents (Sheehan et al, 2017) – from eliminating harmful traditional practices (including child marriage and female genital mutilation) to ensuring quality education and training – investing in a more robust evidence base and improved measurement is critical.
The Gender and Adolescence: Global Evidence (GAGE) research programme addresses these evidence and measurement gaps. This chapter discusses the design and methodological choices made by the GAGE study – to date, the largest longitudinal study (covering nine years, 2015–24) focusing on adolescents (10–19 years) in the Global South. GAGE is following 18,000 adolescent girls and boys in three regions: East Africa, the Middle East and North Africa (MENA) and
This chapter highlights key features of GAGE’s longitudinal design: (1) working with two distinct age cohorts (younger adolescents aged 10–12 and older adolescents aged 15–17) to capture the dynamism of this life stage; (2) involving girls and boys to explore gender dynamics; (3) surveying primary female caregivers to understand intergenerational dynamics; (4) purposely sampling disadvantaged adolescents (those who married early, have a disability, or are internally displaced or refugees); and (5) working across diverse geographies (urban, rural, pastoralist, refugee camps) to understand the role of livelihood options, service access and infrastructural factors in shaping adolescent trajectories.
GAGE’s approach was informed by the need to address questions around adolescent experiences and perceptions, alongside questions on the impact of policies and programmes in mediating these experiences. The research design twins observational longitudinal research across six key capability domains closely linked to the SDGs (education and learning; bodily integrity and freedom from violence; health, nutrition and sexual and reproductive health (SRH); psychosocial well-being; voice and agency; and economic empowerment) with nested experimental and quasi-experimental evaluations of adolescent programming to explore the impact of different support packages.
The chapter concludes by reflecting on the strengths and challenges of longitudinal research, while also highlighting how research funders could do more to enhance the value, relevance and uptake of such research in academic, policy and programming circles, including in terms of achieving the SDGs.
Why adolescence needs a longitudinal lens
The confluence of the major changes that adolescence brings (see Box 6.1) is increasingly recognized by scientists, development actors and policy makers as a vital window in which to accelerate progress against the effects of poverty, inequity and discrimination, and to foster positive development trajectories (United States Agency for International Development [USAID], 2016; Sheehan et al, 2017; Lansford and Banati, 2018). However, the evidence base on the patterning of these changes and adolescents’ evolving capacities – and how they are shaped by gender, geography, political context, (dis)ability, sexuality, marital status and ethnicity, among other identities – is growing but still limited,
Multidimensional changes during adolescence
Biological changes: during adolescence the body achieves its maximum potential in terms of fitness, physical strength and reproductive capacity (De Sanctis et al, 2014). Growth during adolescence is faster than at any other time of life, with a 15–20% increase in height, and individuals gaining up to 50% of their adult body weight. With puberty (typically between 8 and 14 years), hormone changes lead to the development of secondary sex characteristics and, for girls, the onset of menstruation (Lerner and Steinberg, 2004).
Neurodevelopmental changes: early adolescence sees developments in the limbic system – the area of the brain responsible for pleasure-seeking and reward-processing, emotional responses and sleep regulation (Crone and Dahl, 2012). While risk-taking in adolescence may be perceived negatively, it can also lead to adaptive behaviours that promote acquisition of new skills and autonomy and long-term survival (Suleiman and Dahl, 2017). Later in adolescence, changes take place in the pre-frontal cortex – the area responsible for decision making, organization, impulse control and future planning (World Health Organization [WHO], 2018).
Socio-emotional changes: while parental support is critical to adolescent well-being, during adolescence young people typically become more independent from their families and interact more with their peers, valuing opportunities for peer socialization (Saxbe et al, 2015; UNICEF Innocenti, 2017).
Norm changes: in many low- and middle-income country (LMIC) contexts, gender norms and role expectations become more entrenched and personally salient during adolescence. This is manifested in increasing restrictions on adolescent girls’ mobility, and the framing of their future through the lens of marriage and motherhood only (Islam, 2012; Jones et al, 2015; Basu and Acharyam, 2016). By contrast, though adolescent boys tend to have more freedom and status within the family, gender norms around masculinity put them at greater risk of physical violence, substance use and even suicide (as men are discouraged from expressing emotions) (Pinheiro, 2006; UNICEF, 2017).
Legal changes: adolescence typically involves shifts in legal status. Internationally, 18 is recognized as the year of legal maturity (officially entering adulthood), but in many countries, 16-year-olds can legally work, marry and be conscripted (UNESCO, 2015).
A longitudinal lens is critical to capture the dynamism and diversity of adolescent transitions and to understand what works in improving adolescents’ outcomes and trajectories (Morrow and Crivello, 2015; Lansford and Banati, 2018). Given the particular ‘stickiness’ of gender norms, programmes such as GAGE can play an important role in exploring how gender norms shape adolescent development trajectories through observational research, while systematically disentangling the mediating roles that particular programme interventions can play in this process and at different stages of adolescence, through the impact evaluation component.
GAGE overview and conceptual framing
GAGE1 follows the lives of 18,000 adolescent girls and boys in six LMICs in Africa (Ethiopia and Rwanda), Asia (Bangladesh and Nepal) and the Middle East (Jordan and Lebanon) (Jones et al, 2018b). Three core research questions underlie the mixed-methods approach:
How do adolescents in diverse LMICs experience transitions from childhood to adulthood? How do these differ by age, gender, disability, geographic location?
What effects do adolescent-focused programmes have on adolescent capabilities in the short and longer term?
What programme design and implementation characteristics matter for effective delivery and scalability? (Jones et al, 2018b)
GAGE research builds on and complements seminal studies like Young Lives (Boyden et al, 2019). However, GAGE studies a larger and different set of countries (though it includes Ethiopia), focuses on marginalized populations (adolescents with disabilities, married adolescents, and refugee or internally displaced adolescents), and emphasizes understanding of what works to improve adolescents’ lives through embedded impact evaluations. This dimension of programme evaluation adds significant value to the research design, but also complicates longitudinal data collection in numerous ways (discussed later).
Conceptual framework
The GAGE conceptual framework (Figure 6.1) is based on a capabilities approach, which explores the economic, human, political, emotional
The Global Early Adolescent Study conceptual framework (Source: GAGE Consortium, 2017, p 6)2
As Alkire and Deneulin (2018) argue, the capabilities approach is highly compatible with the SDG agenda. First, both the SDGs and the capabilities approach assess a country’s development based on how well their people are doing, rather than on income or consumption levels. ‘What ultimately matters are the types of lives that people are able to live – whether they can function well as human beings, and be and do things they value and have reason to value’ (Alkire and Deneulin, 2018, p 2). Both approaches also adopt a multidimensional perspective on poverty and inequality, recognizing multiple and overlapping forms of deprivation. Second, Alkire and Deneulin argue that the SDGs and the capabilities approach both recognize the interconnectedness of targets across dimensions, including between environmental, economic, social and political domains. Third, the capabilities approach, as advanced by Martha Nussbaum (2011), calls for a common set of capabilities to be strived towards for all individuals irrespective of gender, ability or citizenship – reflecting a shared emphasis on ensuring that no one is left behind. This is also reflected in the SDGs’ emphasis on disaggregated data, whether by ethnicity, disability, age, gender or geographical location. Finally, both approaches underscore the importance of agency and processes, with a strong emphasis on empowering people as agents of change in their own lives and communities, as well as on robust institutions and state action to address structural barriers to tackling poverty and inequality.
Globally, today’s adolescent population is the largest ever (estimated at 1 billion). With appropriate support, it is possible to offset childhood disadvantage, help adolescents transition successfully into adulthood, and contribute to national efforts to reap the demographic dividend that a youthful population offers. As such, investing in adolescent well-being will be a critical part of achieving the SDGs (UNICEF,
GAGE research methods
In order to understand the lived realities of adolescent girls and boys, GAGE’s mixed-methods research uses quantitative and qualitative data (described in more detail later). An overview is then given of the nested impact evaluations the study is conducting in Bangladesh, Ethiopia, Jordan, Nepal and Rwanda.
GAGE quantitative data
GAGE aims to collect at least three rounds of quantitative data in each country, ensuring that data can be analysed for questions of cross-sectional interest (such as the association between gender norms and health), panel data questions (such as what factors are associated with school dropout) and impact evaluation questions (such as what is the causal impact of gender transformative programming in Ethiopia). The focus is on younger adolescents (10–12 years), who are often overlooked by demographic and health surveys (DHS) and labour market surveys, but a smaller older cohort (15–17 years) is also included to be able to understand transitions into early adulthood.
Unlike many studies that focus solely on representative samples, GAGE oversamples certain vulnerable populations to provide more evidence on these understudied groups. The programme also focuses on fragile and conflict-affected contexts (where young people’s specific needs tend to be invisible), and on adolescents living in remote rural areas (those least likely to be reached by services). Finally, GAGE also oversamples adolescents with disabilities (for whom the knowledge base is exceedingly fractured) and never-married girls and boys (whose realities tend to be studied only retrospectively, through DHS questions
Child marriage findings from GAGE Ethiopia
To achieve SDG target 5.3 on eliminating harmful traditional practices (including child marriage) by 2030, strengthening the evidence base to inform programme interventions is critical. The Ethiopian Demographic and Health Survey (2016) shows that child marriage is still common, but GAGE baseline findings underscore stark regional variation in child marriage practices and age at marriage, which has important implications for programme design. In Amhara, for example, the data shows that age at marriage has been climbing sharply, and older girls are now negotiating with their parents about their marriage partner, although very young girls may still be forced to accept arranged marriages. By contrast, key informants and married adolescents in Oromia confirmed that age at marriage has been falling in some communities, mainly due to the fact that child marriage has become more adolescent driven (due to peer pressure, the involvement of brokers and limited educational opportunities beyond primary school). In Afar, where the national family law has still not been adopted, absuma marriage (mandatory marriage to a maternal cousin) is a deeply entrenched norm, and girls have little say in the choice of marriage partner or timing of marriage.
GAGE uses three core quantitative instruments, with both common and country-specific questions and modules. The Primary Female Caregiver module aimed to interview the main caregiver in each household (although secondary female caregivers or male caregivers were substituted as needed). Questions focus on the household structure, family background, assets, durables, dwelling and other household-level characteristics. It also has a specific focus on parenting of the younger cohort, as well as asking about attitudes and norms, mental health, financial inclusion, time allocation, exposure to violence, fertility, marriage and use of technology (with significantly more detail for parents of young adolescents). The core respondent module involves a face-to-face interview with the adolescent, exploring education, time allocation, paid work, health and nutrition, mental health, mobility, voice and agency, social inclusion, programme support, financial inclusion, economic empowerment, technology, marriage and relationships, SRH and violence. The questions are tailored to the age of the respondent.
Qualitative and participatory research methods
The quantitative survey is complemented by in-depth qualitative and participatory research with a subset of adolescents, their parents and siblings to better understand the experiences and perspectives of young people in diverse contexts. The study uses a qualitative toolkit designed to mirror the ‘3 Cs’ framework, and the specific tools used are now described in more depth (see also Jones et al, 2018a, 2019b).
Capabilities
A Few of My Favourite Things – a tool originally used in longitudinal qualitative research with UK adolescents (Thomson et al, 2011; Thomson and Hadfield, 2014) – was adapted to explore adolescent capabilities. Some talked about objects they would like to have (such as a bicycle, radio or mobile phone) rather than what they already have. Others chose objects that highlight the challenges facing disadvantaged young adolescents (for example, a piece of paper among visually impaired students so that they can take Braille notes during class, or a small table in the case of a newly married adolescent girl now responsible for keeping an independent household). Interviewers were able to probe the object’s significance in the adolescent’s life and how it relates to GAGE’s six capability domains. To explore psychosocial well-being and voice and agency, interviewers also used a Social Support Network exercise, asking who the respondent enjoys spending time with (and why), who they can confide in, and who they avoid.
GAGE has also set up participatory research groups that meet more regularly (at least once a quarter) for vulnerable adolescents (for example, those with disabilities, married or divorced adolescents, or internally displaced/refugee adolescents). These groups offer a space for these young people to explore and capture, using participatory photography, the experiences and perspectives of young people at risk of being left behind.
GAGE also included tools to explore the family, community and state-level contexts as well as gender and age dynamics that shape the development of adolescents’ capabilities. A sibling timeline explores how a sibling’s life compares with that of the focal adolescent, and why. A parent timeline explores key influences in the parents’ lives and how their experiences have shaped their approach to parenting, and their aspirations for their children. A similar exercise for grandparents focuses on intergenerational similarities and differences around social norms related to age, adolescent development, and gendered expectations, roles and responsibilities. Together, the individual interviews with family members across three generations provide a rich picture of intergenerational changes and their underlying drivers.
The toolkit includes a number of group-based exercises to explore community-level influences. A community timeline is constructed, with older adults who can remember changes over time. A social norms change tool is also used to explore key opportunities and challenges facing adolescents in each capability domain. These adult-focused discussions are complemented by community mapping with younger and older adolescent girls and boys, to identify the key places and spaces in the community where adolescents can and cannot go, as well as their favourite or secret places, dangerous places and places they aspire to access as adults.
To explore more sensitive issues (with individuals or groups), researchers can use body mapping (a visual tool) to facilitate discussions around emotional and psychosocial well-being, relationships, pubertal development, SRH and gender-based violence. It is also a helpful way to explore similarities and differences between adolescents who are able-bodied and those living with different types of disabilities. Vignette exercises – for instance, a girl about to be subject to forced marriage, or a child who is bullied due to having a disability – are read out to the group, which then discusses whether such a scenario could happen in their community (why it could or could not).
Change strategies
GAGE undertook a wide range of key informant interviews with government officials and civil society representatives at community, district and regional levels. Each interview began by exploring the main opportunities and challenges facing adolescents from the interviewee’s perspective – where possible, mapped against the six capability domains.
A set of experimental and quasi-experimental impact evaluations embedded within the panel data collection activities allows investigators to complement panel data analysis with causal questions on programme impact. Alongside quasi-experimental research in Jordan (Makani) and Nepal (Room to Read), GAGE has three ongoing randomized control trials (RCTs): the Act With Her and Her Spaces programmes in Ethiopia; the 12+ girls’ empowerment programme in Rwanda; and the Adolescent Girls Programme (AGP) in Bangladesh (see Table 6.1).
Overview of RCT programmes evaluated by GAGE
Act With Her (Ethiopia) |
Funder |
Bill & Melinda Gates Foundation (BMGF) |
Delivery partner |
Pathfinder & Care International and Pathfinder Ethiopia and Care Ethiopia |
|
Focal regions |
Act With Her: Amhara, Oromia, Afar Her Spaces: Afar, Amhara, Harari, Oromia and Tigray |
|
Value/project size |
$16,700,000 50,000 direct beneficiaries and 1.7 million indirect beneficiaries. |
|
Objectives |
Act With Her aims to enable adolescent girls (aged 11–16 at project start) to successfully navigate transitions to adulthood by addressing pervasive and harmful gender inequalities, empowering girls, enabling their capabilities and ensuring that girls are able to access essential services. |
|
Interventions |
Act With Her: Clubs (curriculum-based, led by ‘near peers’), interventions with boys, parental engagement, community and systems engagement – especially with ministries of Health and Women, Children and Youth. Her Spaces: A scaled-down 10-month girls’ club focused on life skills and safe spaces. |
|
GAGE evaluation approach |
Mixed-methods longitudinal RCT evaluation design – sample size of approximately 7,000 adolescents and caregivers. |
|
Investing in Adolescent Girls (Rwanda) |
Funder |
Department for International Development (DFID) Rwanda |
Delivery partner |
Non-governmental organization (NGO) partners (procurement process is still ongoing)+ Ministry of Gender and Family Promotion (MIGEPROF) |
|
Focal regions |
All regions – 26 out of 31 districts |
|
Value/project size |
£12.5 million. 200,000 adolescents (150,000 girls and 50,000 boys) |
|
Objectives |
To equip young adolescents with the knowledge and capabilities needed to achieve their full potential. |
|
Interventions |
Delivering out-of-school life skills lessons to 200,000 11-year-old girls and boys, covering basic health, nutrition and hygiene behaviours, sexual health and family planning, puberty, gender and violence, economic opportunities. |
|
GAGE evaluation approach |
Mixed-methods longitudinal RCT evaluation design – sample size of approximately 3,000 adolescents and caregivers. |
|
Adolescent Student Programme (Bangladesh) |
Funder |
World Bank |
Delivery partner |
Government of Bangladesh (Ministries of Education and Health) and partner NGOs. |
|
Focal regions |
Eventually nationwide but currently in Chittagong and Sylhet. |
|
Value/project size |
SEDP for Bangladesh $500 million; HSSP $500 million. |
|
Objectives |
The ASP Adolescent Girls Programme (AGP) aims to enhance cycle completion, gender equity in secondary school outcomes, improved voice and agency, and economic empowerment. |
|
Interventions |
AGP interventions will include: sexual harassment training and support in schools, outreach to teachers on adolescent physical health (including SRH and nutrition), mental health and bullying prevention. Awareness, guidance and community participation on adolescent health to delay age at marriage and timing of first birth. |
|
GAGE evaluation design |
Mixed-methods longitudinal RCT evaluation design – sample size of approximately 5,000 adolescents and caregivers. |
|
Makani/ My space (Jordan) |
Funders |
Multiple |
Delivery partner |
UNICEF, Ministry of Social Development and local NGOs |
|
Focal regions |
Camps, host communities, informal tented settlements across Jordan. |
|
Value/project size |
90,000 beneficiaries annually. |
|
Objectives |
Makani aims to support the development and well-being of refugee and vulnerable host community children and adolescents through an out-of-school one-stop-shop model. |
|
Interventions |
Makani provides safe spaces for children and adolescents aged 6-18 years, learning support, psychosocial support and child protection. |
|
GAGE evaluation design |
Mixed-methods longitudinal quasi-experimental design involving 4000 adolescents, with half in Makani and half not. |
|
Room to Read, Nepal |
Funder |
Room to Read |
Delivery partner |
Room to Read Nepal Country Office |
|
Focal regions |
Tanahun and Nukawot |
|
Value/project size |
4,800 girls total; 571 beneficiaries in the GAGE study. |
|
Objectives |
Exploring the impact of the Girls’ Education. Programme support for adolescent girls’ access to quality education and transitions. |
|
Interventions |
Curriculum-based life skills programme facilitated by social mobilizers focused on: self-confidence, problem solving, critical thinking. Focus also on reproductive health, nutrition, financial literacy and career choices. Programme includes mentoring, family and community support, and needs-based material support. |
|
GAGE evaluation design |
Quantitative quasi-experimental design with control and intervention groups. Sample size of 1200 adolescent girls. |
Emerging findings that have promise for longitudinal follow-up
Alongside impact evaluation questions of interest, GAGE’s mixed-methods dataset will allow for a large set of interesting longitudinal questions. Here, we discuss just a few, and highlight key baseline findings that motivate future longitudinal research. The conceptual framework highlights the interconnectedness of domains, which is clearly reflected in the baseline findings. In Ethiopia, for example, it is found that in Afar, where many very young adolescents are already out of school, they are more vulnerable to child marriage, gender-based violence and poor psychosocial well-being (Jones et al, 2019b). Understanding how this interconnectedness evolves over time will highlight how specific vulnerabilities or capabilities either persist or evolve.
The longitudinal data will also provide critical insight into how gender attitudes and norms change (or stay the same) over time, and how this change is associated with outcomes across the GAGE capability domains. This will build on baseline analysis showing strong associations between gender attitudes and norms and GAGE outcomes. For example, Baird and colleagues (2019a) find that restrictive gender attitudes and norms are associated with poorer physical and mental health outcomes for younger adolescents (aged 10–12) in Bangladesh and Ethiopia. Surprisingly, they find little difference in this association for boys compared with girls, and find stronger results in urban locations. Seeing whether similar findings manifest when we look at changes in attitudes and norms over time – and particularly how this
Given our focus on parenting, the longitudinal data will also illuminate the evolving relationship between parents and adolescents and how this drives adolescent outcomes. Baseline analysis on parent–adolescent communication about future marital and fertility expectations from urban Bangladesh and Ethiopia shows that increases in the primary female caregiver’s desired age at marriage or first birth for their adolescent child is associated with similar increases among adolescents themselves. The results also show that intergenerational dynamics matter, as the gendered attitudes of the primary female caregiver are associated with a lower desired age at first birth for the adolescent (Baird et al, 2019a). The longitudinal data will help unpack the continued role of parents, exploring whether adolescents’ attitudes and behaviours evolve to be closer to those of their parents or, as context and culture change, move further apart.
GAGE data (quantitative and qualitative) has a specific focus on refugees and internally displaced adolescents. In Jordan, the GAGE sample includes approximately 3,000 Syrian adolescent refugees in camp and host communities, and in Bangladesh approximately 1,000 Rohingya adolescent refugees in Cox’s Bazar (see Guglielmi and Jones, 2019; Jones et al, 2019a); there are also samples of indigenous populations in the host communities. Understanding the dynamics between refugee and host communities will highlight both the specific vulnerabilities of refugees, and how their presence affects host populations. Moreover, given uncertainty over the future of these refugees in the host countries, there is potential for rapid change, which the GAGE panel dataset is well placed to capture.
GAGE also focuses on the experiences of adolescents with disabilities (see Box 6.3) – a group for whom it is rare to have quantitative data, and especially panel data. As documented in Muz and colleagues (2019), to ensure a large enough sample of adolescents with disabilities within the overall sample, various strategies were used throughout the data collection process. These included adding a screening question on disability to the household census form prior to selecting the survey sample, identifying adolescents with disabilities during fieldwork by talking to knowledgeable community members, and including in the quantitative survey a set of questions (endorsed by the Washington Group) designed to identify disability.
Contributing to the evidence base on adolescents with disabilities in Ethiopia
SDG 4 calls for quality and inclusive education, including for persons with disabilities, yet approximately one third of children who are not in school (19 million) have a disability (Saebones et al, 2015; Male and Wodon, 2017). Other studies suggest a close relationship between poverty and disability, but evidence gaps obscure the specific mechanisms for this (Filmer, 2008; Mont, 2014). However, we still know relatively little about access to education for children with disabilities in low- and middle-income countries (LMICs), partly due to chronic underreporting (Lewis, 2009). In Ethiopia, the evidence base is especially fragmented due to an absence of routine gender-, age- and disability-disaggregated data collection systems. Indeed, while the World Health Organization (WHO) estimates that about 10% of the world’s population is affected by disability, Ethiopia’s most recent census puts the national total at just 1.09% (ACPF, 2011).
GAGE baseline data (collected in 2017 and 2018) highlights that adolescents with disabilities in Ethiopia face many interrelated challenges, not least in accessing quality and inclusive education. The quantitative findings show that adolescents with disabilities are less likely to be enrolled (64% versus 85%) than those without disabilities and have completed fewer years of education (3.5 versus 4.6). The type and severity of the disability also shape school enrolment: those with intellectual and hearing impairments are mostly out of school, and this gap increases by grade level and age (such that accessing upper primary and secondary education becomes a critical problem). Barriers to accessing inclusive education include inadequate infrastructure and education facilities, limited training and recruitment of suitably qualified teachers, lack of materials, and discriminatory attitudes from parents, teachers, peers and the wider community.
The short case studies from Bangladesh and MENA (see boxes 6.4 and 6.5) provide examples of emerging baseline findings that can
Threats facing adolescents in urban spaces: emerging findings from Bangladesh
Almost one fifth of Greater Dhaka’s 20 million people are adolescents (18.8%) (Index Mundi, 2018), yet there is little understanding of young people’s experiences in this rapidly changing context. Here, we briefly reflect on two examples from our baseline findings: adolescent drug use (in line with SDG target 3.5 on preventing and treating substance use) and internet connectivity (in line with target 9.3 on access to information technology), which will be key issues to track longitudinally.
Drug use: Substance abuse in Bangladesh has risen to prominence largely in the context of the government’s ‘war on drugs’, such that most studies on drug use in the country are couched in terms of crime rather than addiction (see Coyne and Hall, 2017; Gupta and Pokharel, 2018; Islam, 2019). Estimates of drug addiction levels nationwide range from 2.5 million (Shazzad et al, 2013) to 7 million (Haider, 2018), with youth disproportionately affected (comprising up to 80% of all Bangladeshi drug users). Almost half of all drug use is in the capital, Dhaka (Gaffar and Deeba, 2017). While recent evidence demonstrates a relationship between unemployment and drug use (Rahman et al, 2016), patterns of substance abuse among urban adolescents and the social dynamics that lead to increased risk remain poorly understood (Islam, 2019).
Although only about 10% of adolescent respondents in the GAGE baseline survey said they had smoked and only about 5% said they had ever had alcohol, the community focus group discussions, key informant interviews and interviews with boys revealed how common and accessible drugs have become, and how inextricably they are linked with violence and unemployment. Selling drugs is an easy source of income for adolescents who become users, often after they start out earning an initial income from acting as carriers. Interviews with adolescent girls also highlighted the fear and risks of sexual harassment in public spaces by young boys and men who are unemployed and sometimes using drugs. Not only is substance abuse a mental health issue, it is also related to violence and lack of security for girls, boys, women and men in urban settings.
Internet connectivity: Bangladesh has one of the world’s highest rates of internet coverage – 145 million sim cards for a population of 160 million – but there are significant disparities in access and use. As of 2013, national survey results indicated a pronounced gender disparity in internet usage among adolescents, with only 3% of adolescent girls surveyed having used the internet in the past year (Bangladesh Bureau of Statistics (BBS) and UNICEF, 2015), and significant differences between rural and urban access to information and communication technology (ICT) (BBS, 2012). GAGE 2018 baseline data highlight similar gender disparities: 56% of older adolescent boys compared to just 30% of older girls have their own phones, while 75% of the older boys have gone online (versus 35% of the older girls).
Complexities of psychosocial needs among adolescents in humanitarian contexts
SDG 3 on health and well-being for all has brought much-needed attention to this important issue, including among young people (target 3.4). Findings from GAGE, echoing studies in other parts of the Middle East and North Africa region (Okasha et al, 2012; Pocock, 2017), indicate that adolescents’ psychosocial needs are often acute, as they are disadvantaged by age- and gender-related hierarchies, political turbulence, poverty, and lack of resources and skills that would help them cope with these multiple stressors (Samuels et al, 2017; Jones et al, 2019c). The literature indicates that many adolescents in the region have developed anxiety-related disorders, depression, loss of hope and susceptibility to risky behaviours, including suicidal ideation and substance abuse (Abu Hamad et al, 2015; Saymah et al, 2015; WHO and Palestinian National Institute of Public Health, 2017). The GAGE baseline survey in Jordan found that 32% of respondents meet the threshold for ‘caseness’ on the General Health Questionnaire (GHQ) (Jones et al, 2019a), while 46.2% of adolescents in Gaza reported psychosocial ill-being (Hamayel and Ghandour, 2014).
GAGE findings suggest that older girls are 11% more likely than older boys to exhibit emotional distress, mostly attributed to their greater social isolation, anxiety about being forced to marry young, subsequent experiences with child marriage, and limited access to psychosocial services (Jones et al, 2019c). In terms of context, poorer adolescents living in highly conservative informal tented settlements were especially likely to exhibit distress (40% versus 33% in host communities and 29% in camps) (Jones et al, 2019c).
While most studies on psychosocial well-being in the MENA region are medically/epidemiology oriented (focusing on prevalence of illnesses), GAGE uniquely focuses on the gendered experiences and perceptions of adolescents about their psychosocial well-being. For example, conforming to expectations of ‘family honour’ is a source of considerable stress for older girls, while physical violence was a major concern among boys.
GAGE also fills an important gap by exploring the extent to which age- and gender-sensitive psychosocial support is available to adolescents, and their experiences accessing such services. Adolescent uptake of formal services is limited by lack of knowledge about service availability as well as stigma attached to mental illness (Samuels et al, 2017). Boys reported that social norms around masculinity and self-sufficiency restrict their participation in psychosocial activities, and that they tend to resort to substance use and/or violence as a coping strategy (Samuels et al, 2017). By contrast, social norms around marriageability tend to preclude adolescent girls from accessing psychosocial services.
Overall, at the system level, our findings highlight that such services are fragmented, poorly governed and monitored, reactive in nature, and are neither gender- nor age-sensitive, while access to specialist services is limited (Samuels et al, 2017). It is therefore essential to launch cross-sectoral interventions, including but not restricted to strengthening the links between psychosocial support and health services, education systems and social protection programmes (Jones et al, 2019c).
Learning so far
GAGE has generated some important learning so far, with implications for longitudinal work more generally. Here we discuss three lessons: (1) the promise and perils of embedded RCTs; (2) a truly mixed-methods approach; and (3) ensuring that the sample includes the most vulnerable.
While these RCTs are an essential part of GAGE, they have also introduced complications that other researchers doing similar work could learn from. First and foremost, as GAGE funding did not include any funding for implementation, GAGE researchers had to find implementing partners willing to work with a research team on an evaluation design without any additional financial support. Also, the timing of implementation had to line up with the research (and vice versa) – with some, but limited, flexibility given GAGE’s reporting commitments to funders. The original plan (to find existing programming to work with) was ultimately thwarted, largely due to these two constraints. The exception is 12+ in Rwanda – also funded by DFID – but even in this case, GAGE delayed data collection for more than two years to accommodate programme delays (baseline is now scheduled for 2020).
For the other two RCTs, GAGE was ultimately able to operate within these constraints by working closely with partners to bring in additional money for implementation. This is exemplified in the Ethiopia case, where GAGE worked closely with Pathfinder International to secure implementation funds from the Bill & Melinda Gates Foundation for programming that aligned with the GAGE conceptual framework. In
Incorporating programming and developing RCTs has had some drawbacks though. As mentioned, aligning GAGE data collection with programming led to long delays in data collection (Rwanda) or less than ideal timing between data collection and programming (in Ethiopia, for example, baseline data collection took place in late 2017 but programming only started in mid-2019). It also complicates some of the panel aspects of the data collection, in terms of comparisons within and across countries. In Ethiopia, for example, Act With Her focuses on 10–13-year-olds in rural areas, so in order to have sufficient sample size for the RCT, GAGE data collection focused solely on young adolescents in these areas. However, GAGE is also interested in adolescent transitions in urban locations in Ethiopia, where data was collected on both younger and older adolescents. This data structure implies that we cannot distinguish whether certain findings among 15–17-year-old urban adolescents are due to age or the urban environment.
There are similar challenges with comparisons across countries. Due to delays in programming, as well as differences in target age groups and/or duration, data collection sometimes focuses on different age groups (for instance, 11-year-olds in Rwanda versus 14–15-year-olds in Bangladesh and 11–13-year-olds in Ethiopia); nor are these groups aligned in terms of timings (for example, adolescents in Ethiopia were 10–12 years in 2017, while those in Rwanda will be 11 in 2020). While we believe the added value of the impact evaluation component outweighs the drawbacks, maintaining a perfect multi-country panel data structure while at the same time nesting impact evaluations is probably impossible, except perhaps if a funder funded both research and implementation (although the reality of fieldwork would still make this a challenge).
Mixed methods
While other longitudinal research claims some level of a mixed-methods approach, almost inevitably one approach takes priority over the other. GAGE puts equal emphasis on quantitative and qualitative methods, which is reflected in the research team composition. This ensures that quantitative data is appropriately contextualized and that qualitative data is forced to confront issues around lack of representativeness. Yet
Vulnerable adolescents
As already mentioned, GAGE’s central focus is on the most vulnerable but typically under-researched young people – adolescents with disabilities, adolescents who have experienced child marriage, and refugee and internally displaced adolescents. Again, though, there are important trade-offs. First, adolescents with disabilities and those who experience child marriage are often not reached by traditional quantitative methods, so we undertook complementary purposeful sampling. This not only adds costs, but raises the question of whether conventional ‘census’-style listing misses key populations, or whether GAGE is actually over-representing these populations in our research (we believe the former to be the case). Second, both of these groups are usually left out of programming or, if they are included, programmes are not tailored to their needs, so including them in the sample is at odds with exploring programme impact. Third, given taboos around child marriage, quantitative and qualitative data may record an individual’s marriage status differently, leading to challenges in triangulating data. Finally, tracking respondents over time is already a challenge with adolescents as they are a very mobile population; when you add in refugees and internally displaced populations, these challenges become very considerable, particularly as some may return to their home countries.
Implications of longitudinal research design for policy and programming impacts
What lessons can we draw from GAGE’s experience of longitudinal research design and implementation, and how does this approach strengthen the evidence base on gender and adolescence and inform the SDGs?
First, given the relative rarity of longitudinal research funding but the huge potential of such research, public goods research arguably has an obligation to seek to balance demands for methodologically robust social science with practice- and policy-relevant research evidence.
Second, research and programming teams within the same funding entities need to design this work jointly from the outset if investments into longitudinal research are to have maximum use for policy and programming. While productive partnerships can be forged across programmes with different funding streams, GAGE’s experience to date suggests that these are often complex, time-intensive and imperfect; many of the obstacles faced could (arguably) have been avoided through joint design and mutual contractual mandates.
Third, longitudinal research evaluations of programming typically rely on programme stability, but in reality, programming is often more dynamic; indeed, there is growing evidence that an adaptive management approach is critical for context-responsive programme implementation. In this regard, GAGE is exploring how to harness its hybrid research design to allow for programming adaptations in response to emerging findings while maintaining a panel study and continuing to learn about the medium- and longer-term legacy effects of the original design on adolescents’ lives.
Fourth, given the rapid cognitive and psychosocial shifts that adolescence brings, research instruments over successive rounds need to be adapted to be age-responsive while maintaining sufficient continuity so as to assess impact. This requires considerable investments in developing research instruments and piloting.
Fifth, in order to contribute to the breadth of the SDG agenda as it pertains to adolescent lives, considerable resources are needed to develop research instruments that cut across thematic areas. This is in turn compounded by efforts in line with the ‘leave no one behind’ agenda to ensure an inclusive sample that encompasses married adolescents and adolescent parents, those in remote rural areas, adolescents with disabilities and those who are internally displaced or refugees.
Finally, there is a need to educate stakeholders about the increasing value of longitudinal research insights over time. Because the start-up costs (resources and time) are considerable, managing expectations in the early years and in monitoring and evaluation of such programmes is critical. When the focus is on finding out ‘what works?’ it is important to understand that expectations around baseline and first follow-up should be appropriately modest. Ultimately, the rationale
Acknowledgments
The authors wish to thank Eric Neumeister and Kate Pincock for research assistance for this chapter.
Notes
The GAGE Consortium, managed by the Overseas Development Institute (ODI), includes 35 partner organizations from around the world known for their expertise in research, policy and programming in the fields of adolescence, gender and social inclusion. GAGE is funded by UK aid from the UK government. For more details see www.gage.odi.org.
References
Abu Hamad, B., Jones, N., Al Bayoumi, N. and Samuels, F. (2015) Mental Health and Psychosocial Support Service Provision for Adolescent Girls in Post-Conflict Settings: The Case of the Gaza Strip. London: Overseas Development Institute.
African Child Policy Forum (2011) The African report on child wellbeing 2011. Available at https://resourcecentre.savethechildren.net/library/african-report-child-wellbeing-2011-budgeting-children
Alkire, S. and Deneulin, S. (2018) The real wealth of nations. SDGs: delivering change. Blog, 19 March. Available at www.sustainablegoals.org.uk/real-wealth-nations
Baird, S., Bhutta, Z., Hamad, B., Hicks, J., Muz, J. and Jones, N. (2019a) Do restrictive gender attitudes and norms influence physical and mental health during very young adolescence? Evidence from Bangladesh and Ethiopia. SSM – Population Health 9: 100480.
Baird, S., Hicks, J., Jones, N., Muz, J. and GAGE Consortium (2019b) Ethiopian baseline survey 2017/2018. Core respondent module. London: Gender and Adolescence: Global Evidence.
Bangladesh Bureau of Statistics (BBS) (2012) Population and housing census 2011. Socio-economic and demographic report. National series, vol 4. Dhaka: Bangladesh Bureau of Statistics.
BBS and UNICEF (2015) Bangladesh: multiple indicator cluster survey 2012–2013. Progotir Pathey, final report. Dhaka: Bangladesh Bureau of Statistics and UNICEF.
Basu, S. and Acharyam, R. (2016) Gendered socialization of very young adolescents: perceptions and experiences of adolescents and their parents from a disadvantaged urban community of Delhi, India. Paper presented at conference on Adolescence, Youth and Gender: Building Knowledge for Change, 8–9 September, Oxford.
Boyden, J., Dawes, A., Dornan, P. and Tredoux, C. (2019) Tracing the Consequences of Child Poverty: Evidence from the Young Lives Study in Ethiopia, India, Peru and Vietnam. Bristol: Policy Press.
Coyne, C.J. and Hall, A. (2017) Four decades and counting: the continued failure of the war on drugs. CATO Institute, Policy Analysis No. 811.
Crone, E.A. and Dahl, R.E. (2012) Understanding adolescence as a period of social-affective engagement and goal flexibility. Nature Reviews Neuroscience 13(9): 636–50.
De Sanctis, V., Soliman, A.T., Fiscina, B., Elsedfy, H., Elalaily, R., Yassin, M. et al (2014) Endocrine check-up in adolescents and indications for referral: a guide for health care providers. Indian Journal of Endocrinology and Metabolism 18 (Suppl 1): S26–38.
Filmer, D. (2008) Disability, poverty, and schooling in developing countries: results from 14 household surveys. World Bank Economic Review 22(1): 141–63.
GAGE Consortium (2019) Gender and Adolescence: Why Understanding Adolescent Capabilities, Change Strategies and Contexts Matters. London: Gender and Adolescence: Global Evidence.
Gaffar, M.B. and Deeba, F. (2017) Mental health conditions among adolescents of substance dependent parents. South East Asia Journal of Public Health 7(1): 48–50.
Guglielmi, S. and Jones, N. (2019) The Invisibility of Adolescents within the SDGs: Assessing Gaps in Gender and Age Disaggregation to Leave No Adolescent Behind. London: Gender and Adolescence: Global Evidence.
Gupta, S. and Pokharel, S. (2018) Bangladesh defends war on drugs as body count mounts. CNN, 8 June. Available at https://edition.cnn.com/2018/06/08/asia/bangladesh-drug-war-intl/index.html
Haider, L. (2018) Bangladesh has 7 million drug addicts, ‘over half of them are addicted to yaba’. Bdnews24.com, 21 May. Available at https://bdnews24.com/bangladesh/2018/05/21/bangladesh-has-7-million-drug-addicts-over-half-of-them-are-addicted-to-yaba
Hamayel, L. and Ghandour, R. (2014) Assessment of the Psychosocial Health Status of Adolescents 12–18 Years Old in the Occupied Palestinian Territory. Ramallah: Birzeit University, Institute of Community and Public Health.
Index Mundi (2018) Bangladesh demographics profile 2018. Available at www.indexmundi.com/bangladesh/demographics_profile.html
Islam, N. (2019) War on drugs: the case of Bangladesh. Harvard Public Health Review 25. Available at http://harvardpublichealthreview.org/bangladesh
Islam, T. (2012) Causes and consequences of eve-teasing in urban Bangladesh: an empirical study. SUST Studies 15(1): 10–20.
Jones, N., Tefera, B., Presler-Marshall, E., Gupta, T., Emirie, G., Gebre, B. and Berhanu, K. (2015) ‘Now I Can Propose Ideas That Can Solve Any Problem’: The Role of Community Awareness Interventions in Tackling Child Marriage in Ethiopia. London: Overseas Development Institute.
Jones, N., Camfield, L., Coast, E., Samuels, F., Abu Hamad, B., Yadete, W., et al (2018a) GAGE Baseline Qualitative Research Tools. London: Gender and Adolescence: Global Evidence.
Jones, N., Baird, S. and Lunin, L. (2018b) GAGE Research Design, Sample and Methodology. London: Gender and Adolescence: Global Evidence.
Jones, N., Gebeyehu, Y. and Hamory-Hicks, J. (2019a) Exploring the role of evolving gender norms in shaping adolescents’ experiences of violence in pastoralist Afar, Ethiopia: victim, perpetrator, or what else? Sociological Studies of Children and Youth 25: 125–47.
Jones, N., Presler-Marshall, E., Małachowska, A., Jones, E., Sajdi, J., Banioweda, K. et al (2019b) Qualitative Research Toolkit: GAGE’s Approach to Researching with Adolescents. London: Gender and Adolescence: Global Evidence.
Jones, N., Baird, S., Presler-Marshall, E., Małachowska, A., Kilburn, K., Abu Hamad, B. et al (2019c) Adolescent well-being in Jordan: exploring gendered capabilities, contexts and change strategies – a synthesis report on GAGE Jordan baseline findings. London: Gender and Adolescence: Global Evidence.
Kabeer, N. (2003) Gender equality, poverty eradication and the millennium development goals: promoting women’s capabilities and participation. Gender and development discussion paper series (13). New York: United Nations.
Lansford, J. and Banati, P. (2018) Handbook of Adolescent Development Research and Its Impact on Global Policy. Oxford: Oxford University Press.
Lerner, R. and Steinberg, L. (2004) Handbook of Adolescent Psychology (2nd edn). Hoboken, NJ: John Wiley.
Lewis, I. (2009) Education for disabled people in Ethiopia and Rwanda. Background paper prepared for the Education For All Global Monitoring Report 2010.
Male, C. and Wodon, Q. (2017) Disability Gaps in Educational Attainment and Literacy. The Price of Exclusion: Disability and Education Series. Washington, DC: World Bank and Global Partnership for Education.
Mont, D. (2014) Childhood disability and poverty. Working Paper 25. London: Leonard Cheshire Disability and Inclusive Development Centre, University College London.
Morrow, V. and Crivello, G. (2015) What is the value of qualitative longitudinal research with children and young people for international development? International Journal of Social Research Methodology 18(3): 267–80.
Muz, J., Baird, S., Hicks, J., Rashid, S., Sultan, M., Yadete, W., Jones, N. (Eds) (2019) A Mixed Methods Approach to Including Adolescents with Disabilities in Survey Research. Comparative Research Programme on Poverty (CROP). Available at https://expert.gwu.edu/individual/doc-3e9ba5a32d4dc2f0e4e8a7dc157af5bb
Nussbaum, M. (2011) Creating Capabilities: The Human Development Approach. Cambridge, MA: Harvard University Press.
Okasha, A., Karam, E. and Okasha, T. (2012) Mental health services in the Arab world. World Psychiatry 11(1): 52–4.
Patton, G.C., Coffey, C., Cappa, C., Currie, D., Riley, L., Gore, F. et al (2012) Health of the world’s adolescents: a synthesis of internationally comparable data. The Lancet 379(9826): 1665–75.
Patton, G., Sawyer, S.M., Santelli, J.S., Ross, D.A., Afifi, R., Allen N.B. et al (2016) Our future: a Lancet commission on adolescent health and wellbeing. The Lancet 387(10036): 2423–78.
Pinheiro, P.S. (2006) World Report on Violence against Children. Geneva: United Nations.
Pocock, L. (2017) Mental health issues in the Middle East: an overview. Middle East Journal of Psychiatry and Alzheimers 8(1): 10–15.
Rahman, F.N., Ahmad, M. and Ali, M. (2016) Socio-economic status of drug addicted young people in Dhaka City. Journal of Armed Forces Medical College 12(2): 15–20.
Sæbønes, A.M., Bieler, R.B., Baboo, N., Banham, L., Singal, N., Howgego, C. et al (2015) Towards a disability inclusive education. Background paper for the Oslo Summit on Education for Development, 6–7 July.
Samuels, F., Jones, N. and Abu Hamad, B. (2017) Psychosocial support for adolescent girls in post-conflict settings: beyond a health systems approach. Health Policy and Planning 32(5): 40–51.
Saymah, D., Tait, L. and Michail, M. (2015) An overview of the mental health system in Gaza: an assessment using the World Health Organization’s Assessment Instrument for Mental Health Systems (WHO-AIMS). International Journal of Mental Health Systems 9(41): 1–8.
Saxbe, D., Del Piero, L., Immordino-Yang, M.H., Kaplan, J. and Margolin, G. (2015) Neural correlates of adolescents’ viewing of parents’ and peers’ emotions: associations with risk-taking behavior and risky peer affiliations. Social Neuroscience 10: 592–604.
Sen, A. (2004) Elements of a theory of human rights. Philosophy & Public Affairs 32(4): 315–56.
Shazzad, M.N., Abdal, S.J., Majumder, M.S.M., Ali, S.M.M. and Ahmed, S. (2013) Drug addiction in Bangladesh and its effect. Medicine Today 25(2): 84–9.
Sheehan, P., Sweeny, K., Rasmussen, B., Wils, A., Friedman, H.S., Mahon, J. et al (2017) Building the foundations for sustainable development: a case for global investment in the capabilities of adolescents. The Lancet 390(10104): 1792–806.
Steinberg, L. (2015) Age of Opportunity: Lessons from the New Science of Adolescence. New York: Eamon Dolan/Mariner Books.
Suleiman, A.B. and Dahl, R.E. (2017) Leveraging neuroscience to inform adolescent health: the need for an innovative transdisciplinary developmental science of adolescence. Journal of Adolescent Health 60(3): 240–48.
The Palestinian National Institute of Public Health (2017) Estimating the Extent of Illicit Drug Use in Palestine. Ramallah: The Palestinian National Institute of Public Health/UNODC/KOICA (Korea International Cooperation Agency).
Thomson, R. and Hadfield, L. (2014) Day-in-a-life microethnographies and favourite things interviews. In V. Johnson, R. Hart and J. Colwell (Eds) Steps to Engaging Young Children in Research: Volume 2: The Researcher Toolkit (pp 126–30). Brighton: University of Brighton. Available at https://bernardvanleer.org/publications-reports/steps-engaging-young-children-research-volume-2-researcher-toolkit/steps-for-engaging-young-children-in-research-volume-2-the-researcher-toolkit7591
Thomson, R., Kehily, M., Hadfield, L. and Sharpe, S. (2011) Making Modern Mothers. Bristol: Policy Press.
United States Agency for International Development (USAID) (2016) Acting on the Call. Ending Preventable Child and Maternal Deaths: A Focus on Equity. Washington, DC: USAID. Available at www.usaid.gov/sites/default/files/Final-AOTC-file-v2.pdf
UN Statistics Division (2016) The Sustainable Development Goals Report 2016. New York: United Nations. Available at https://unstats.un.org/sdgs/report/2016
UNESCO (2015) EFA Global Monitoring Report 2015. Education for All 2000–2015: Achievements and Challenges. Paris: UNESCO.
UNICEF (2016) Harnessing the Power of Data for Girls: Taking Stock and Looking Ahead to 2030. New York: UNICEF. Available at https://data.unicef.org/resources/harnessing-the-power-of-data-for-girls
UNICEF (2017) A Familiar Face: Violence in the Lives of Children and Adolescents. New York: UNICEF.
UNICEF Innocenti (2017) The adolescent brain: a second window of opportunity. A compendium. Florence: UNICEF Office of Research, Innocenti.
Viner, R., Ross, D., Hardy, R., Kuh, D., Power, C., Johnson, A. et al (2015) Life course epidemiology: recognising the importance of adolescence. Journal of Epidemiology & Community Health 69(8): 719–20.
WHO (2018) Guideline: Implementing Effective Actions for Improving Adolescent Nutrition. Geneva: World Health Organization. Available at www.who.int/nutrition/publications/guidelines/effective-actions-improving-adolescent/en
