Assess@Learning was the largest European policy experimentation to date on digital formative assessment, that is, on the question of how digital tools can help teachers to monitor their pupils’ learning progress continuously and to adapt their teaching methods accordingly. Nine partners from eight countries developed an online toolkit for this purpose, coordinated by European Schoolnet and funded by Erasmus+. Whether it works was tested in a randomised field trial in Estonia, Finland, Greece, Portugal and Spain, involving 208 schools, 846 teachers and around 2,500 pupils. We advised the consortium over four years on all methodological questions: on the theory of change, on sampling and the trade-off between randomised assignment and voluntary participation, on defining the primary outcomes, and on pre-registering the analysis plan. At the end, we reviewed the results. The toolkit measurably improved teachers’ subject knowledge and they could distinguish formative from summative assessment more confidently. Beyond that, effects were largely absent. This is precisely where the value of a clean design shows: because outcomes and analysis had been fixed in advance and publicly registered, the null finding could be named as such and not merely as a lack of significance. This laid the foundation for investigating the reasons in a targeted way and for developing the toolkit further.
Additional Information
Partners
