I have been following a discussion on LinkedIn about the Big5 model and tests created to measure it. The Big5 model is a description of personality that identifies 5 major personality factors. It has a long history and the curious should read the Wikipedia article about this model. Our own OP5 Test is measuring personality traits in the Big5 model and divides the 5 dimensions into 30 separate sub-scales, it is an intensive and very detailed test and used mostly by trained psychologists or specialized HR professionals.
The LinkedIn discussion went all over the place and it is in Dutch, but I thought I would share my contribution to that discussion here. The question under consideration was this (paraphrased):
‘Why would I use a likert scale (each question asks the respondent to rate a phrase from “very applicable” to “not applicable”) test to measure personality, wouldn’t a candidate just mark ‘high’ on every question to get the most socially desirable results?’
So, I waded into the thread with my own take on how these results should be interpreted:
A well constructed likert style test (with a rating for each item from low to high) test will have ‘mirror’ items. That is to say, that on a significant portion of the items, a high rating will be negative for the scale it is measuring. For example; “I like to be alone”, if rated high, would have a negative influence on the ‘Extroversion’ scale. This means that a candidate blindly marking the highest level on all items will not get the result they expect.
This is also a part of why norm groups are used, especially norm groups based on the situation of the test. If a candidate is answering with the goal of getting a job (selection), they tend to be more positive about themselves than in a strictly development (coaching) situation.
If you are looking for the ‘shape’ of a personality, you should use a forced-choice questionnaire or use an ipsative analysis of the results. But that ipsative analysis gives no indication of the ‘degree’ of a personality trait. If you are looking for the ‘degree’ of a personality, you should use a likert scale test (as most Big5 tests are). The reason most Big5 questionnaires are likert scale is because some people really do have more or less personality (as a whole) than others.
A huge part of helping with the reliability of a test result is in how you communicate with the candidate. Putting the candidate at ease, explaining exactly how the results will be used and why honesty is important, and verifying the results with an interview or cross checking them against other tests are all good ways to ensure accurate results. (cross checking against other instruments is helpful because it is difficult to lie consistently)
That is also one of the reasons our professional users employ several tests from our catalog when looking at a candidate as part of a selection assessment. The different tests look at a candidate from different perspectives and we know, from over 90 thousand candidates, how those different tests normally relate to each other. If your personality is conflicting with your values and work style and communication style, then that raises a big red flag that will be addressed in interviews and/or phone consultations.
—
If you have a topic or question you would like to see addressed here and on our mailing list, please drop me a line r.still AT onlinetalentmanager.com.
