Assessing behaviours

Assessing behaviour is when we measure behaviours as a particular time point. We can then compare these assessments over time to evaluate whether change is happening.

Developing and using a behavioural checklist

A behavioural checklist is useful to see if your partnership work is changing important practices. It is usually easiest if someone who understands the clinical situation is completing the checklist (unless the IBOs are very simple)  so that they can make a judgement about the appropriate standard. The checklist, of course, must be completed ethically, with full knowledge of those being observed. Sometimes these can be completed as part of a supportive visit, where people are offered support after the checklist with any situations they found difficult.

Create a behavioural checklist with each of the IBOs, and three columns a) if the behaviour was expected; b) if it was observed and c) if it met the appropriate standard.

Discuss with the team who will be conducting the observations and completing the checklist. Make a plan about when and where the observations will take place. Sometimes, this will be observing people who will be
attending training, sometimes it will be observing a whole ward, outpatient clinic or other physical area.

Observations should take place before and after interventions to fully understand any impact that interventions are having on health professional practice.

Example observation checklist

Example of additional information you might include on checklist

Self-report behaviours

Sometimes, it is not possible to observe people in their own environment. In these cases, it is possible to ask people to self-report their behaviours. This is not perfect because sometimes people will forget and sometimes people will over or underestimate their behaviours. It can give some indication in the absence of being able use a checklist., for example, before and after training and perhaps in a three-month follow-up. You will need a way of matching up participants e.g. by a code.

You can ask:

How many times in the last week could you have done [insert behaviour]?

How many times in the last week did you [insert behaviour]?

Or

How many people did you see last week who [insert the situation in which the behaviour should / should not be done]?

With how many of these people did you do [insert behaviour]?

For example, if you were interested in doing an antibiotic review 48 hours after surgery, you would say:

How many people did you see last week who were 48 hours after surgery?

With how many of those people did you do an antibiotic review?

You can vary the time frame to fit, bearing in mind people are more likely to remember things that have happened fairly recently.

Behavioural expectation

Sometimes, you want to get an idea of how an intervention might change someone’s intention or expectation of doing. This is usually because you are assessing them before they have had a chance to go back to work after your intervention. This might be at the start and end of a training session, for example. Expectation has been shown to be a good predictor of behaviour, although it is obviously not as strong evidence of change as observed or self-reported behaviour.

We usually ask a behavioural expectation question in the following form, with the answer being between 0 and 10:

For every 10 [insert situation/ context] in how many would you expect to do [insert behaviour].

For example:

For every 10 people you see with a upper respiratory tract infection in how many would you expect to prescribe an antibiotic.

For information about how to explore behavioural influences, see

Qualitative evaluation methods?

Sometimes it is useful to conduct interviews or focus groups following training or other interventions to understand in-depth participants’ experiences and how they have put learning into practice. This also gives insights into what they haven’t managed to put into practice and why, helping you refine your training and other support you might be able to include going forward.

Routinely collected data

It helps if routinely collected data can be used to understand behaviour. For example, the number of prescriptions for a particular antibiotic would help us to understand the behaviours of prescribers. The benefit of these types of data is that they are collected routinely and therefore can be less time consuming to collate.