What is a meta-analysis?

Combining the results from several studies can be helpful when it comes to investigating the effectiveness of different teaching and learning approaches

In a meta-analysis, researchers combine the results of multiple studies to calculate an overall effect. This involves more than just pooling data from studies together - the process uses established statistical methods to account for differences in sample size, variability and findings.

The EEF view: What are the implications for education research?

For schools and teachers, meta-analysis can be a useful way of finding out the average effectiveness of a particular teaching and learning approach. A good meta-analysis will contain a large amount of information and compare effects across a number of different studies, drawing out patterns and trends.

Single studies looking at the same research question can often produce contradictory results. Differences in context, how an approach was implemented or how it was measured can make findings difficult to compare.

The good thing about meta-analyses is that they translate results from different studies to explore relationships between findings. By comparing findings from many different studies, we can establish how “generalisable” they might be. Meta-analyses can also examine inconsistency across different studies and present hypotheses about why impacts might be different.

If you’re using meta-analyses to support your decision making, it’s important to look beyond the average - in academic studies, this is often described as the heterogeneity (or diversity) of results.

A good example of this is the research on teaching assistants. While teaching assistants can provide a considerable positive impact on learner outcomes, the high average impact hides a large variation in different approaches. Targeted deployment - where teaching assistants are trained to deliver an intervention to small groups or individuals - has a higher impact, whereas deployment of teaching assistants in everyday classroom environments has not been shown to have a positive impact on outcomes.

Where can I see meta-analyses in action?

There are many examples of meta-analyses in education research.

One recent example is from 2020: Patrick Leman, who was dean and professor of psychology at the Institute of Psychiatry, Psychology and Neuroscience at King’s College London at the time, and Dr Harriet Tenanbaum, reader in developmental psychology at the University of Surrey, published a meta-analysis of peer learning. They found that it can be beneficial for students - and that when compared with direct one-to-one teaching, the outcomes were the same.

In another example, from 2018, a team of researchers from Simon Fraser University in Canada, headed by educational psychology doctoral student Kiran Bisra, conducted a published meta-analysis that looked at 64 research reports, involving almost 6,000 students, and found better learning outcomes in those who received prompts to self-explain while studying or solving problems.

Further reading:

Recent

What is a meta-analysis?

Read more