Deciding Whether a Program Improved Health and Wellbeing
A district of 240,000 people introduced a school-based program combining daily physical activity, free fruit, and mental health literacy lessons for students aged 8 to 14. After three years, the evaluation reports: mean student body mass index fell slightly; self-reported happiness rose from 61% to 70%; reported bullying incidents rose from 140 to 210 per year (attributed by staff to greater willingness to report); Year 9 dropout fell from 9% to 6%; and youth clinic visits for anxiety rose 40%. The program cost $2.1 million. A councillor argues the program failed because anxiety visits rose and bullying reports increased.
1. Assess the councillor's argument using the dimensions of health and wellbeing.
Read the worked answer
The councillor treats rising anxiety visits and bullying reports as evidence of harm, but both are ambiguous indicators. The physical dimension shows improvement (lower mean BMI, more activity, better nutrition). The mental and emotional dimension shows mixed signals: higher happiness suggests better emotional wellbeing, while more anxiety clinic visits may reflect greater help-seeking and reduced stigma rather than more illness, especially given the mental health literacy component. The social dimension likely improved: increased reporting of bullying is a recognised sign of improved trust and reduced normalisation of violence, and lower dropout suggests stronger connection to school. Spiritual and cultural dimensions are not directly measured, so no conclusion can be drawn. The correct judgement is that the evidence is consistent with overall improvement in health and wellbeing, with better detection of problems. The councillor's inference confuses increased measurement with increased prevalence, and ignores the favourable indicators.
2. Identify what additional data would be needed before concluding the program caused these changes, and explain why.
Read the worked answer
Causation requires a comparison. Needed data include: (1) a control or comparison group of similar schools without the program, to see whether BMI, happiness, bullying reports, dropout, and anxiety visits changed similarly there; (2) baseline and trend data for several years before the program, because some indicators (for example, anxiety visits) may have been rising already; (3) dose and participation data, to test whether students with higher exposure showed larger changes; (4) measurement consistency, confirming that bullying reporting and clinic referral practices did not change; and (5) cost-effectiveness information, comparing the $2.1 million with alternative uses. Without these, observed changes could reflect secular trends, changed reporting culture, maturation of students, or concurrent policies. The strongest inference available now is association, not causation.