Black Dog Institute study of 1,500-plus NSW students links school belonging to later wellbeing

In this study, connectedness is something students said they felt. They rated statements such as "I feel like I belong at school" and "I make friends easily at school" ^[5], and their mental health was measured with standard psychological questionnaires ^[7]. Both measures came from the same teenagers.

The two-year gap is the strongest part of the design. Belonging was measured in eighth grade and wellbeing in tenth ^[8], so the order in time is fixed. The researchers say the link held after accounting for gender and a history of mental health issues ^[9]. That adjustment removes one obvious confounder. A student who already felt well in eighth grade might also have found friends easier to make, though. The published summary reports neither an effect size nor whether eighth-grade wellbeing was held constant.

The denominator matters most for the null results. More than 1,500 students across 35 schools ^[6] works out to about 43 students per school on average ^[1]. Connectedness varies from one student to the next, so it was tested across roughly 1,500 data points. Sector, gender mix and student-teacher ratios vary only between schools, so they were tested across 35. A study built this way is far better placed to detect a student-level association than a school-level one. The finding of no link for ratios, sector and single-sex status ^[10] is weak evidence that those features do not matter for wellbeing. The authors note that the same factors have been linked to other outcomes, such as academic achievement ^[15].

"Our findings suggest enhancing connection early in high school is likely to improve well-being later in high school," the Black Dog Institute researchers wrote ^[11]. That is a causal claim. The study observed students. It did not raise anyone's sense of belonging and then wait to see what happened. The authors acknowledge that more research is needed to find the most effective strategies ^[12]. Their candidates include connection committees with teachers, staff and students as members, channels for students to raise concerns with school leaders, mentoring across grade levels, and lunchtime activities run by teachers ^[13]. They also cite a recent meta-analysis that found programs built on mentoring and on fostering problem-solving skills with others were more effective ^[14].

I think the result is enough to put belonging on a school's wellbeing plan, with one condition. The factors the team measured included school income, attendance and staff views of school culture ^[4]. Connectedness was the only one that tracked with wellbeing two years on ^[8]. Earlier work has also tied it to higher achievement and less bullying ^[3], so a program aimed at belonging would be judged on more than one outcome. The condition is evidence. A school running mentoring or lunchtime clubs would need to show wellbeing scores moving against a comparison group before it can call belonging a lever. The hours involved are large: by age 15, an Australian student may have spent as many as 11,000 hours at school ^[1].