Low stakes quizzing
Description of the innovation
The evaluation sought to establish whether weekly low-stakes quizzing of Year 8 geography and history pupils across five half-terms would lead to an improvement not only in the breadth of knowledge but also its application during end-of-year assessments in both subjects. The evaluation also sought to establish whether this form of quizzing is sustainable in everyday classroom practice.
Summary of the evaluation
Heathfield Community College is a rural, comprehensive secondary school in the south of England. The Year 8 pupils taking part in the study were a mixed prior-attaining cohort and were taught in eight mixed prior-attainment classes throughout the academic year in geography and history by subject specialists. Four classes (approximately 50% of the cohort) were randomly selected to receive the weekly, 10-minute low-stakes quizzing intervention which consisted of 10 closed questions. The remaining four classes were the control classes who only took part in retrieval practice on an ad hoc basis throughout the evaluation period. All pupils sat a baseline assessment in each subject at the start of the evaluation and these mean baseline scores were compared with the mean scores achieved in the end of year assessments in each subject.
Summary of findings
The intervention in geography showed that it did have an impact with an effect size of +0.3 for all pupils. It had a greater effect on girls (+0.4) and on pupils who are not regarded as pupil premium (+0.5). The greatest effect (+0.8) was on pupils with special educational needs (SEND), but it has to be considered that the sample size was small (n=9), especially compared with the whole cohort (n=218). Our key conclusion is that low-stakes retrieval practice in the form of short-answer recall tests do boost meaningful learning in geography. My view is that this is because they enable pupils to use what they know to better understand and write about the subject.
In contrast, the intervention in history showed that it had a much smaller impact with an effect size of +0.1 for all pupils. It had a slightly greater impact for boys (+0.2) in comparison to girls (+0.1). Similar to geography, the greatest effect was on pupils with SEND (+0.2), but the sample size was also small (n= 12) in comparison to the whole cohort (n= 215). Consequently, the conclusion drawn here is that low-stakes retrieval practice in the form of short recall tests in history leads to a smaller increase in meaningful learning and may be of less help with pupils applying knowledge to write about the subject than in geography.
One limitation of the study is that the conclusions are predicated on the final assessment truly being a test of meaningful learning and this could be contested. There is also the need to be very clear about what we mean by “works” – works for what purpose? (Biesta 2015). In the case of geography, I would say that our study shows that regular quizzing of previously studied material led to pupils being able to write better answers to geographical questions.
Although the overall sample sizes for the intervention group and the control group were fairly large compared to other practitioner research studies some of the subgroups were much smaller, such as the numbers of SEND and pupil premium pupils. There were differences in the numbers of pupils who eventually sat the history and geography assessments due to pupil absence. This is particularly seen where the number of SEND pupils who successfully completed all parts of the study was different in geography (n= 9) to history (n= 12). Similarly the final number who sat the history end-of-year assessment was lower (n= 215) than those who sat the geography end-of-year assessment (n= 218), as not all pupils in the cohort completed the end-of-year assessment.