The name is pretty accurate: it’s about whether items in self-report questionnaires are overt about what they are designed to measure or not. I have a feeling that the terms like “hidden” and “opaque” are also used.
Details #
The idea is that items that are overt about what it is they are aiming to measure are easier for respondents to deliberately “fake” their responses, perhaps “faking good” or “faking bad”. The idea is that making the intention behind the item or all the items in a measure may particularly prevent it suffering from “social desirability response bias”: that some of us are more concerned to present ourselves. I’m grabbing an example from the Roman Alexithymic Scale (not a measure I have used). Alexithymia, literally “not having words for feelings” is supposed to be a dimension on which people differ in their abilities to recognise and describe their feelings. In many cultures having strong abilities to describe one’s feelings is a positive attribute so might be liable to social desirability (SD) response bias. When raters looked at the itesm in the 27 item measure “I clearly recognize the emotions I feel”) was rated as overt and “My physical sensations confuse me” as more covert.
My sense is that there was some enthusiasm for using a mix of overt and covert items in questionnaires, or making up an entire questionnaire with only covert items, in perhaps the 1950s to 1980s but that the idea then largely dropped out of fashion. The idea has some overlaps with the idea of having “lie detect” items in measures and with the use of measures simply designed to give a measure of individuals’ social desirability bias (the 1960 Crowne Marlow Social Desirability Scale was a dominant example but there is also a Brief Social Desirability Scale).
I think we should be overt (if that’s not too recursive terminology!) about our feelings and I admit that I’ve largely avoided lie detector items, SD scales and covert items as I think that if we are interested in how clients perceive their change across interventions the more we can create an invitation that seems simply to encourage as much transparency as the clients can bear the better the information we will get. Having said that, I do think that the issues of clients easily imputing the issues we seek to address and having the ability, if they want, to adjust how they respond in the light of how they think we may use the information are not trivial issues and do need much more exploration.
Try also #
- Bias
- Social desirability
Chapters #
I think we were fairly clear throughout the OMbook that how measures are presented matters and may influence how clients respond but these issues weren’t overt (lovely word!) in the book!
Online resources #
None forseeable.
Dates #
First created 6.i.25.
