Not evidence-based

A lot of advice and practice in education is not evidence based.  This includes:
Authority:  “Professor Jones has shown….”;  “It’s in the National Strategy”
Anecdote:  “When I was at school we….”; “They do this at St Xavier’s and they get good results.”
Habit:  “We’ve always done it this way”;
Polls:  “86% of the pupils said they had learnt more…”

Poor evidence

Classroom experiments which have:

• a very short time scale
• few students involved
• no control group (to compare)
• lessons taught by the experimenters
• only one teacher involved

Sources of good evidence

• Classroom experiments: conducted with control groups.
• Cognitive science: studying how the brain learns.

References

To make the pages easier to read, EBTN does not give you the original references on the page.  You can find the original references by looking at the sources we quote.

See also:

Building evidence into education  by ‘Bad Science’ writer, Ben Goldacre.
Good intentions are no longer acceptable in education Wayne Harrison  Durham University
This blog post from Learning Spy on the problem with teacher-judgement.

Definition of EBT

Short version.
Rather than sifting through research papers themselves (with the problem of biased selection), teachers make use of professional research reviews.  There are several hundred of these, but, when we select only the methods which are the most effective, we find a handful of practical, but highly effective, ways to improve the learning of our students.

 Longer version

What does “evidence-based” mean in education?

Teachers and their students deserve the very best evidence. Rather than picking and choosing what research to base your ideas on, there needs to be a system for finding and sifting research. This is because we need an expert interpretation of the whole picture (or to get as close as we can to this). One piece of research can be contradicted by another, and enthusiasts for one interpretation of research will, consciously or unconsciously, choose research that harmonises with their own point of view. 

Systematic reviews of research 

To overcome these problems researchers have developed what they call ‘systematic reviews’ of research (often called ‘meta-studies’ or just ‘research reviews’). This is the best evidence we have.
There are over 900 of these just on the issue of educational attainment. They have clear criteria and methods for searching for papers, books etc, so that every view is harvested, and also clear criteria for allowing or disallowing such sources into the review. Usually the criteria require the research to be done with real teachers and students (rather than in laboratory simulations), an experimental design with a control group, or a good quasi-experimental design, an effect size, large enough numbers of students for valid conclusions to be drawn, etc.
 
Syntheses of over 800 Meta-Analyses Relating to Achievement have been published by John HattieVisible Learning (2009), and  Visible Learning for Teachers: Maximising impact on learning (2012)
Some systematic reviews are presented in teacher friendly language.  For example Classroom Instruction that Works , by Prof Robert Marzano. Other books present the findings of meta-studies in ways that teachers can understand and act upon. For example Evidence Based Teaching, Geoff Petty and other books recommended on this website.
 
When systematic reviews are not available, then the nearest approximation to it needs to found, but most factors affecting achievement have been reviewed.
 
Since the average effect size is 0.4, EBTN suggests that interventions with an effect size greater than 0.5 should be used first, as these will have the greatest effect on learning.