Measuring Wellbeing in Adolescents Living with HIV: Lessons from Zimbabwe

In the clinics and community support groups of Zimbabwe, adolescents growing up with HIV face a daily balancing act that goes far beyond swallowing antiretroviral pills. They navigate stigma, grief, poverty and the ordinary turbulence of adolescence, all while trying to stay adherent to a medication regimen that must be taken for life. Programs such as Zvandiri, a peer-led initiative that trains young people living with HIV to support their younger peers, have shown that strengthening mental health and wellbeing can improve HIV treatment outcomes. Yet a deceptively simple question has haunted such programs for years: how do you actually measure whether a teenager is flourishing? A new study published in PLOS Global Public Health by Webster Mavhu of the Centre for Sexual Health and HIV AIDS Research Zimbabwe and colleagues offers an unusually candid answer, and in doing so delivers a methodological wake-up call to the entire field of global mental health measurement.

The research team set out to build a wellbeing scale from the ground up, designed specifically for adolescents living with HIV in Eastern and Southern Africa. The motivation was straightforward. Most existing instruments either dwell on negative outcomes such as depression and anxiety, or they borrow from the positive psychology literature, where questionnaires were developed and validated in high-income countries among populations whose cultural assumptions about happiness, selfhood and success differ profoundly from those of rural and urban Zimbabwe. Concepts like self-esteem, life satisfaction and personal growth are not universal constants; they are shaped by family structures, religious belief, community expectations and the lived experience of a chronic, stigmatized infection. A questionnaire that assumes an individualistic, self-focused notion of wellbeing may simply fail to register what matters most to a fifteen-year-old in Harare.

Between 2022 and 2025, the researchers ran a multi-stage scale development process that would satisfy most textbook checklists. They began with a scoping review and a systematic review of the literature, mapping the crowded landscape of positive psychological constructs and the instruments used to measure them. They then conducted qualitative discussions with adolescents living with HIV, solicited input from an expert panel, prioritized constructs and candidate items, and ran cognitive interviews to check whether young people actually understood the questions as intended. Iterative item reduction followed, using exploratory factor analysis and Rasch analysis, two statistical techniques that test whether a set of questions hangs together as a coherent measurement instrument and whether the response options function as designed. The process whittled an initial item bank down to a 53-item candidate scale, then to a 38-item version, and finally to a 25-item instrument called the Zvandiri Character Strength scale, or ZCS 25.

The qualitative work produced findings that should reshape how anyone thinks about adolescent wellbeing in this setting. When young people talked about what made life good, they emphasized connectedness to others, hope for the future, self-acceptance, self-efficacy, a sense of purpose, and flourishing. Crucially, wellbeing emerged as relational and contextually situated rather than a private, internal state. Being well meant being embedded in supportive relationships, being able to contribute to one’s community, and having one’s identity as a young person living with HIV met with acceptance rather than shame. The discussions also revealed something that Western positive psychology instruments rarely accommodate: wellbeing and distress often co-occurred in the same person at the same time. A teenager could feel hopeful about her future while simultaneously grieving a parent or worrying about disclosure of her status.

Then came the quantitative testing, and this is where the study becomes genuinely instructive, because the results refused to cooperate with the researchers’ hopes. When the ZCS 25 was administered retrospectively before an intervention, immediately after, and at follow-up, the responses were overwhelmingly positive. So positive, in fact, that the data showed ceiling effects, with most participants clustering at the top of the response scales, leaving no room to detect improvement. Rasch analysis revealed disordered thresholds, meaning that adjacent response categories such as somewhat agree and strongly agree did not function as distinct levels of the underlying trait. The internal structure of the scale was weak, with items failing to group into the theoretically expected domains. Retrospective ratings fluctuated in puzzling ways across administrations.

The most illuminating explanation came not from the statistics but from the participants themselves. In feedback sessions, adolescents suggested that their glowing responses likely reflected affirmations, a kind of positive self-statement that carries social and psychological value in its own right, and that social desirability had shaped at least some of their answers. In a context where support groups celebrate resilience and where admitting struggle to a researcher or peer counselor can feel like a betrayal of the community’s ethos of hope, answering positively is not necessarily deception. It may be a meaningful social act. For measurement science, this is a profound complication: the very act of asking about wellbeing can invite a performance of wellbeing, and the boundary between genuine flourishing and aspirational self-presentation is not one that a Likert scale can easily draw.

The bottom line, stated with rare honesty in the published paper, is that the ZCS scale did not support a single interpretable total score. Adding up the 25 items and reporting a wellbeing number would have been statistically indefensible. Many research teams, under pressure to demonstrate impact to funders, would have been tempted to report the totals anyway. The authors instead treated the failure as the finding, arguing that the development process itself generated the most valuable lessons. This is an unusual and commendable stance in a field where null or messy results are routinely buried, and it transforms the study from a failed instrument into a methodological case study with implications far beyond Zimbabwe.

So what should the field do differently? The authors lay out a set of prescriptions that amount to a more humble and more rigorous approach to cross-cultural measurement. Future efforts should begin with a clearer construct definition, specifying exactly what aspect of wellbeing is being targeted before a single item is written. Response-process validation, checking how participants actually interpret and formulate their answers, deserves far more attention than it typically receives, because the cognitive interviews in this study foreshadowed many of the statistical problems that emerged later. Domain-level testing, evaluating each dimension of wellbeing separately rather than chasing a single grand total, would respect the multidimensional and sometimes contradictory nature of adolescent mental life. And researchers should make an explicit decision about the purpose of the measure: is it meant to capture lived wellbeing as adolescents experience it, to assess domain-specific positive outcomes, or to detect change attributable to an intervention? These are different goals that may require different instruments.

The wider stakes are considerable. Global funders increasingly demand mental health outcome data from HIV programs, and positive psychology instruments developed in the United States or Europe are being deployed across sub-Saharan Africa with limited validation. If those instruments produce ceiling effects and social desirability bias in the very populations they are meant to serve, programs may appear to have no effect, or worse, effects may be fabricated by measurement artifacts. The Zimbabwe study does not offer a ready-made scale, but it offers something arguably more valuable: a demonstration that rigorous, participatory measurement development can fail informatively, and that listening to adolescents about what wellbeing means to them is not a preliminary courtesy but the scientific foundation on which any valid instrument must rest. For the millions of young people living with HIV, getting this right is not an academic exercise; it determines which services get funded, which interventions scale, and whose flourishing is finally counted.

Subject of Research: Development and psychometric evaluation of a culturally grounded wellbeing scale for adolescents living with HIV in Zimbabwe

Article Title: Rethinking wellbeing measurement: Learning from a scale development study with adolescents living with HIV in Zimbabwe

Article References: Mavhu, W., Dambi, J. M., Simms, V., Martin, F., Lariat, J., Mbundure, R., Makoni, L., Kafata, L., Wogrin, C., Chitiyo, V., Mutsinze, A., Cowan, F. M., Willis, N., & Bernays, S. (2026). Rethinking wellbeing measurement: Learning from a scale development study with adolescents living with HIV in Zimbabwe. PLOS Global Public Health, 6(9), e0006845. https://doi.org/10.1371/journal.pgph.0006845

Image Credits: AI Generated

DOI: 10.1371/journal.pgph.0006845

Keywords: adolescent health, HIV, wellbeing measurement, Zimbabwe, psychometrics, scale development, global mental health, positive psychology, Rasch analysis, social desirability, Zvandiri, cross-cultural validation