How to Guide - Dealing with Social Desirability in Assessments

A common question from clients and recruiters is “what do I do if my candidate(s) show evidence of social desirability?” This article helps you to consider this situation from the angle of the assessment itself and the interview process.

What is Social Desirability?

Most high-quality personality and self-rating assessments these days include social desirability scales. These measure a candidate’s approach to the assessment and his/her response style, and provide an indication of the validity of results.

Social desirability is measured on the basis of audit questions. These questions are separate from the construct-related questions in the personality assessment. The social desirability scores are not specifically linked to one or more personality constructs.

At Lumenii we measure two types of social desirability, both of which are rated on a scale of one to three:

1. No indication (results are valid)

2. Some indication (results may not be valid)

3. Definite indication (results are not valid)

Impression Management of your Future Staff

Evidence of impression management implies that the candidate consciously shared a more positive image of him/herself, typically aligned to their expectations of the role.

A higher score on impression management is more common in people from collectivist cultures. In addition, higher scores on impression management are far more common in selection scenarios than in developmental scenarios - this is due to the high-stakes environment of recruitment and selection, as people often feel that creating a more positive image of themselves will help them to “get the job”.

Individuals applying for sales, commercial and management-level positions typically score higher on impression management. According to Jan Meijning, senior psychologist, this is logical from a sales perspective as the candidate is “selling themselves” when completing the assessment. Since selling themselves for such functions is quite logical behaviour, this impression management score is self-explanatory and not necessarily problematic. “My way of dealing with it is to discuss the score with the candidate. In general, he/she understands the point. I ask the candidate if they can help identify where they might have presented themselves more enthusiastically than normal.”

Self-Deception

A candidate will show evidence of self-deception if they unknowingly have an unrealistic image of themselves. A very high score on Self-deception is rare (1% of candidates), while some indication of self-deception occurs in 4% of candidates.

Social Desirability and the Assessment

If a candidate shows evidence of self-deception, there is no point asking them to re-complete the assessment. Because their bias is unconscious, they will not recognise the need to be more transparent, and they will typically argue that the assessment itself is not an honest reflection.

When dealing with a higher score on Impression Management, it is acceptable to ask the candidate to re-complete the assessment (depending on advice from the assessment provider). At Lumenii we follow a set procedure for impression management scores of 2 or 3:

• Upon receiving the relevant report, the administrator reviews the social desirability scores to determine the validity of results. An impression management score of 2 means that the candidate CAN redo the assessment (optional), but a score of 3 means that the candidate SHOULD redo the assessment (strongly advised).

• The trained test administrator or an Lumenii psychologist then has a face-to-face discussion (ideal) or telephonic discussion with the candidate. During this discussion, it is explained that he/she may have represented themselves in the way that they would like to be, rather than how they currently are. It is important to be very conscious of the way this is explained, so that the candidate does not get defensive. The candidate is asked if he/she is willing to redo the assessment.

• If the candidate agrees to complete the assessment again, the administrator resets the questionnaire on the platform and resends a link to the candidate.

• He/she completes the assessment again as normal, and the results are once more reviewed by the administrator.

• If a candidate continues to show evidence of impression management after completing the assessment again, it is typically not recommended that they complete a third or fourth time. The results can either be thrown out or probed further in an interview with the candidate, depending on the degree of impression management and the company/assessment policy.

Social Desirability and the Interview

Making use of competency-based interview questions is the best way to validate scores found in the competency assessment. This is even more true in the case of impression management, as it allows the interviewer to explore the results and compare them to the candidate’s interview responses.

A few tips on Impression management:

• Ask critical questions and push the candidate to provide details when answering, and to avoid answering in short sentences.

• Combine the results of the assessment with the interview by asking for concrete behavioural examples of certain competencies. For example, “when last did you have to make a decision with limited information? How did you go about making the decision?”

• Probe the same competency in a few different ways. There is a greater risk of being fooled by impression management in shorter interviews.

• Evaluate the candidate's responses to each question separately and do not just look at the overall picture. When dealing with the big picture, you are susceptible to the effects of impression management.

Self-Deception and the Interview

When faced with a high self-deception score in the interview, do not get into a discussion with the candidate about the detailed assessment results. Someone with an unconscious unrealistic self-image will always insist that the results are not accurate. Instead, ask the candidate about his/her strengths and weaknesses, and ask for examples of these.