I have been blogging about (primarily) educational research for over a decade now. Often, this involves new studies that I summarise, annotate, and try to place within what we already know. But what if the way you communicate research actually matters hardly at all? Through René Kneyber, I discovered a new study by Samuel Sims and colleagues that poses this rather uncomfortable question.

They had nearly 3,000 teachers read a short summary of a hypothetical study of a teaching approach. They varied nine elements that one might expect to make the message more persuasive. For instance, some teachers read a story from a colleague, while others received concrete advice on how to apply the approach. Sometimes the effect size or sample size was mentioned. Sometimes there was an endorsement from an expert, an explanation of why the approach works, or a reference to a credible university. Because each of these elements could be present or absent, 512 different versions of the same research summary were created. As a research design, that sounds quite convincing.

But what did the results show? In fact, almost nothing made a difference. For eight of the nine elements, the researchers found no statistically significant effect on teachers’ intention to use the approach. The only significant effect was very small and, moreover, negative. Even various additional analyses did little to change that picture.

So, time to close my blogs and find something else to do early in the morning? Hold your horses. Not quite.

What the researchers tested here is quite specific. They gave teachers a brief summary of a single study and immediately asked whether they intended to use the approach described in it. So they weren’t even measuring actual behaviour.

And in fact, we should probably be happy about that. Personally, I certainly hope that no one reads just one blog post about one study and thinks, “Starting tomorrow, I’m completely overhauling my teaching.” In my opinion, that would often be a particularly bad idea.

What I try to do with my blog is something else. I place new studies alongside what we already know. Sometimes a study confirms previous findings, sometimes it contradicts them, and I may also add an important nuance. Very occasionally, a study is particularly interesting precisely because you shouldn’t draw any major conclusions from it. Like this study, by the way.

Interestingly enough, the authors themselves end up moving in a similar direction. They suggest that we might have a greater need for knowledge brokers, such as those at the E.E.F. or the Flemish Leerpunt, which I head. They refer to people and organisations that bring together different studies, synthesise them, and translate them into a more coherent picture, rather than expecting a single well-written research summary to change practice.

So no, I am not stopping blogging yet.