Over the past few days, I have read several pieces about how schools influence pupils’ well-being (often negatively, according to the articles). But how big is that influence actually? It seems a simple question, but the answer is much less so. Previous research has already shown that schools often inherit much of their pupils’ well-being from outside school. After all, when pupils enter the school, they do not leave their home situation, personal experiences and possible mental health problems at the gates. Of course, schools can make a difference. But I do think we should be careful when we simply attribute differences in well-being between schools to the schools themselves.

But there are other questions we could ask. Can a positive school environment contribute to pupils’ mental well-being? Or could it also be the other way round? Could pupils’ mental well-being partly determine how they experience their school environment? A new study by Esparza and colleagues, published in Journal of Adolescent Health, examined this interaction as far as possible.

The researchers used data from the American ABCD Study. This is a large longitudinal study that follows children and young people over several years. For this analysis, they used data from 8,454 young people. In total, there were four annual measurement points between approximately ages 10 and 14.

In their research, the authors studied three factors: family conflict, the school environment as perceived by students, and psychological problems. With regard to the latter, they differentiated between internalising problems, such as anxiety and depression, and externalising problems, such as aggression. They assessed the school environment mainly using the relationships with teachers. They also looked at pupils’ sense of safety at school.

The authors utilised what is known as a random-intercept cross-lagged panel model. Behind this fancy-sounding term is an important distinction. It is one thing to show that adolescents with more mental health problems are, on average, also less positive about their school. It is quite another to examine whether an adolescent who experiences more mental health problems than usual at a particular moment also rates their school environment less positively a year later. And, of course, you can also examine the reverse direction.

So, what did they find? Young people’s mental health problems more often predicted later changes in their family and school environments than the other way round! For example, young people with more externalising problems were more likely to experience more conflict within their family a year later. Internalising problems also predicted more family conflict at certain ages. The reverse direction, in which family conflict predicted later mental health problems, was found less consistently.

For the school environment, the picture was actually even more striking. After correcting for multiple statistical tests, the researchers found no robust associations in which a more positive school environment predicted fewer mental health problems a year later. Conversely, there were several associations. Young people who experienced more internalising or externalising problems were more likely to rate their school environment less positively later on.

There were also some differences between boys and girls. For example, among girls aged 11 to 12, internalising problems predicted a less positive school experience a year later. This was not the case for boys.

There is a catch, however. The effects we are talking about here were rather small (most standardised coefficients between 0.04 and 0.09). Several of these associations even disappeared when the researchers ran additional analyses to check how individual pupils developed over time. Still, young people with more externalising problems in their early adolescence tended to experience more family conflict later on.

We certainly cannot conclude from this study that schools have no influence on their pupils’ mental well-being. Despite the strong research, this study is less suited to answering that question. And besides, other research does show effects of the school environment. What the study does show, however, is that the interaction between pupils and their environment is more complex than you sometimes read. Pupils are influenced by their family and school. In turn, they also influence those environments, or at least how they experience them.

And that brings us back to the original question. When we measure pupils’ well-being, what exactly are we measuring? The influence of the school? The circumstances pupils bring with them from outside school? Or a combination of both?

I suspect mainly the latter. And that is why we should be careful about holding schools responsible for their pupils’ well-being. Not because schools cannot do anything about it, but because they often have far less under their control than we think.