How to stop research interventions that don't seem to work: 10 key steps

In early December, there was a strange alignment between the real-world and Twitter when the issue of evidence-based practice in education was discussed in three conversations:

• The Coalition for Evidence-Based Education talked about the notion of strategic abandonment (thank you @DrCarolineCreaby)
• Later in the evening, #DebatED discussed "whether an interest in education research is more about identifying what doesn't work as suggesting what will"
• Then on Thursday, the topic for #UKEdResChat was, "How do we define 'what works' in educational research? Should we also be focusing on what doesn't work?"

With that in mind, it seems sensible to examine a process for how to disengage from strategies and interventions which appear not to be working. In her 2011 paper, Failing By Design, Rita Gunther McGrath identified a disciplined process for getting out of projects. Her steps include:

1. Decide in advance on periodic checkpoints for determining whether to continue or not
2. Evaluate the project's upside against the current estimated costs of continuing. If it no longer appears that the project will deliver the returns anticipated at the outset, it may be time to stop
3. Compare the project with other projects that need resources. If this one looks less attractive than they do, it may be time to stop
4. Assess whether the project teams may be falling prey to escalation pressures
5. Involve an objective, informed outsider in the decisions about whether to continue, instead of leaving it up to the project team members
6. If the decision is made to stop, spell out the reasons clearly
7. Think through how capabilities and assets developed during the course of the projects might be recouped
8. Identify all who will be affected by the project's termination. Draw up a plan to address disappointments or any damage they might suffer
9. Use a symbolic event – a wake, a play, a memorial – to give people closure
10. Make sure that that the people involved get a new, equally interesting opportunity.
    
    

Given what we know about educational research and interventions, it is impossible to avoid things that do not work. As such the choice is simple: continue with practices and interventions that do not work or release the resources for use in some area where they might. However, in doing so, it is important to maximise what can be learnt from failure – and which may lead to success next time.

This piece is an edited version of a blog that was originally published here.