Research and Evaluation Projects

As the map above shows, nearly half of the studies were done in the US, while the Middle East and Western Europe also contributed a large number of experiments and quasi-experiments in this area. An explicit MC/SRL focus and reporting the study in English were required for inclusion in the review, which limits how geographic trends are interpreted. Still, it appears there is a lack of evidence on the effectiveness of MC/SRL programs in mathematics in some regions of the world, indicating a need for further research that takes context into account. 

This graph shows the types of activities most commonly used in the synthesized studies to train "meta" skills in mathematics. During-task reflection and control activities were the most common, and these were usually practiced while solving realistic problems. Students learned metacognitive questions for prompting their next problem-solving steps, for example. Prompts were also embedded into digital problem-solving activities in some studies. Metacognition and self-regulated learning were also trained through pre- and post-task activities, such as planning strategies and reflecting on learning or performance. Verbalizing thinking through discussion or writing, as well as modelling a mathematical situation, were also highly used program activities. 

Sixty studies between 2005 and 2019 that trained "meta" skills and had mathematics achievement outcomes were statistically combined in a meta-analysis. For each study, the table above shows the effect size given in standard deviation units, the standard error, and the number of participants. The studies can be ordered by any of these measures by clicking on the column labels. This table shows  no consistent relationship between effect size and standard error or sample size, but the most "effective" studies often featured smaller samples. As with other educational interventions, there may be challenges in "scaling-up" effective metacognitive training programs.