By David Creelman

HR loves behavioural rating scales. Yet it is odd that some of the original work on competencies by Signe and Lyle Spencer used behavioural scales for things we do not normally think of as behaviours at all. Those competencies focused on traits such as analytical thinking, conceptual thinking and achievement motivation—things that are more cognitive than behavioural. What is going on?

As HR professionals, behavioural rating scales are something we should understand in-depth and that means knowing a bit about their history.

A historic debate in psychology
Part of the reverence HR has for behavioural scales is a holdover from the views of behaviourist psychology, a movement that held sway for much of the 20th century. Behaviourists were scornful of psychologists who relied on introspection and vague concepts about what was going on in the brain. Behaviourists felt you could only do scientific research on things you could observe, i.e. behaviours. Some of that scorn for vagueness still shows up as pride in behavioural rating scales.

There was value in the behaviourist perspective but it went too far. Behaviourists tended to deny the value of anything cognitive, and ultimately the narrowness of that perspective gave way to a more balanced view. As HR professionals we need to make sure that we too take a balanced view.

A broad view of what counts as behaviour
The competencies of Signe and Lyle Spencer are often better understood as mental habits rather than behaviours. People with a high degree of analytical thinking tend to break a problem into parts; it is how they habitually approach the world. For the Spencers, the important point was that you could assess these competencies in a reliable way within a structured interview. Asking people “what were you thinking at the time?” is a good method for finding out if people have the habit of analytical thinking. The fact that we talk about this sort of interview method in terms of “observable behaviours” and “behavioural rating scales” is something of a holdover from the days when behaviourism was seen as the scientific way to do psychology. The takeaway is that we should have a broad view of what counts as a behaviour. It doesn’t need to be “observable” in that you can look out the window and see it. There simply has to be some reasonably reliable way to assess the factor.

Strengths and limits of behavioural rating scales
If you want your store managers to assess employees on “reliability” it is very helpful if you can illustrate what that means with specific examples; examples that help distinguish “unreliable” from “somewhat reliable”. In other words we should give the store managers what we normally call a behavioural rating scale.

However, humans are always messier than even the best designed scales imply. I remember many cases where highly trained competency consultants would look at the same interview transcript and one would say it showed one competency (e.g. initiative) while another argued it showed something else (e.g. interpersonal understanding). This was neither a problem with the behavioural rating scales for the different competencies nor with the consultants. Humans, and the situations we find ourselves in, are multifaceted and ambiguous. Behavioural scales cannot “fix” that, they simply help make it more manageable.

The takeaway is that helpful as behavioural scales are they still rely on a great deal of judgement and the results are indicative not certain.

Healthy humility
HR specialists and HR consultants often talk about scales based on observable behaviours as if they were finely calibrated laser range finders. The truth is that sometimes what they measure is not something we would normally think of as a behaviour, and almost always there is a good deal of ambiguity and imprecision in their use. The scales are useful. We just need a little more humility in how we talk about them and expect less precision in the results.

David Creelman is CEO of Creelman Research, providing research and writing on human-capital management. Currently he is co-leading a club on evidence-based management with Dr. Denise Rousseau. Mr. Creelman can be reached at dcreelman@creelmanresearch.com.