I confess that I have been avoiding doing this definition because I’m uneasy about whether I have the correct definition and because I am pretty sure it has become a very general term probably often too general.
Details #
The easy bit, on which people seem to agree, is that the term comes from Rensis Likert, an American social psychologist who introduced rating scales using items with response levels given sequential numbers as scores. Here’s one I made earlier (yes, that’s a joke for ageing Brits who watched “Blue Peter” in their childhoods).
That’s the first item of the English language CORE-OM. The catch for me is that I think that is actually a “frequency rating item” and not really a Likert item (I am not sure if I have ever called the CORE measures Likert scales, I know that for decades I have tried not to!) I think that to be a true Likert type item the anchors would be balanced and something like:
- Very rarely
- Fairly rarely
- Sometimes
- Fairly often
- Very often
I.e. terminology is used that is symmetrical and balanced. I think Likert’s original ratings were agreement, e.g.
- Strongly disagree
- Disagree
- Neither agree nor disagree
- Agree
- Strongly agree
I also think that Likert used “scale” for the aggregation of scores from a number of such items not for a single item. I confess I have never read his original work.
However, the term seems to be used very widely now for items and for multi-item scales where the items have a fixed set of responses that can be argued to be sensibly ordered so that scoring them using sequential integers: 0, 1, 2, 3, 4, 5 or 1, 2, 3, 4, 5 looks sensible. That distinguishes them from using binary choices or visual analogue scales. (Though binary choices can be seen as minimal rating scales, they are clearly not what Likert was using.)
The mapping of responses to integer numbers takes scoring of such measures into the whole arena of psychometrics and, sadly, the often rather polemnical wards between Classical Test Theory (CTT) and Item Response Theory (IRT).
Try also #
- Classical test theory
- Item response theory
- Visual analogue scales
Chapters #
Not covered in the OMbook.
Online resources #
Unlikely!
Dates #
First created 12.iii.25.
