Have you ever preferred a candidate because of their similarity to you on various criteria? Did you grow up in the same village? Did they come from the same foreign country as you? Perhaps they also liked running or had four kids just like you? If you preferred and selected this candidate, then you were influenced by a bias called the similarity effect. This effect has shown that the more similar two individuals are, the greater the attraction between them. Wondering whether recruiters might be attracted to candidates with similar personality traits, we recently conducted a study to investigate this question, and you might be surprised by the results.
The similarity effect has been studied extensively over the last decades. It can be observed based on various elements. Byrne, in 1961 , was one of the first to explain that people with similar opinions have a mutual attraction. The similarity effect can also appear in different situations. While it can influence decision-making in the world of recruitment, it can also occur during a sports match or a TV Show, for example.
If, as described above, general similarity and opinion influence the degree of attraction, other widely studied criteria can influence this similarity effect. These are the socio-demographic criteria. We can mention race, gender, age, sexual orientation, etc.
To illustrate this point, a study in sports showed that NBA referees favor players of the same skin color. Players earned up to 4% fewer fouls and scored up to 2½% more points on matches in which their race was similar to that of the refereeing crew. A similar observation has been made in baseball, where a shot is more likely to be considered a catch when the referee shares the player's skin color.
Last but not least, besides the socio-demographic criteria, individuals appreciate other people with the same skills as they do. They, therefore, judge them more positively because of the similarity effect.
Considering that the use of personality testing is on the rise in organizations, we conducted a study to examine if recruiters would be more inclined to hire candidates whose personality profile is similar to theirs. The sample comprised 135 recruiters, with 63% of women whose average age was 36 years old. In the first part of the study, recruiters had to evaluate the hireability of 14 fictitious female candidates for an internship position in their company (as shown in the example below).
All candidates differed regarding attractiveness, extraversion, agreeableness, and conscientiousness (openness to experience and emotional stability were held constant across candidates). In the second part of the experiment, recruiters completed a short personality questionnaire measuring their personality traits.
Two results stood out in this study. First, candidates' conscientiousness was the strongest predictor of recruiters' hireability ratings. In other words, recruiters emphasize conscientiousness more than other information available (attractiveness, agreeableness, and extraversion) to decide if a candidate is hireable.
Second, no personality-based similarity effect was observed. In other words, conscientious recruiters did not put more emphasis on candidates' conscientiousness than their less conscientious counterparts, and extraverted recruiters did not put more emphasis on candidates' extraversion than more introverted recruiters.
How can we explain this unexpected result?
The start with, there is a large body of evidence on the relationship between personality traits and the ability to perform in several work environments. Giving preference to a candidate because they like the same sport as you is one thing if they meet the job's other criteria. On the other hand, favoring an introverted candidate because you are introverted when the job in question requires a lot of social interaction is quite another.
It is reasonable to assume that recruiters are looking for the most competent and suitable person for the specific job in question, despite the fact that they might like another person who is more similar in other respects.
Furthermore, the methodological approach of the research also influences the results obtained. The question of the perception of the trait does not come into play here. Personality results were given to the participating recruiters, not perceived by them. Each candidate profile presented to the recruiter detailed her five personality traits standardized and visually (as shown above).
Yes, without a doubt, the similarity effect does exist. People tend to have favorable impressions about others who share similarities. But this assertion raises another question. Is it common for hiring decision-makers to suffer from the similarity effect? Concerning this question, the results are far more nuanced because most of these professionals have been adequately trained in the use of hiring methods and the existence of perception biases.
Large-scale studies conducted in the field (analyzing what happens in actual and not simulated interviews) found no support for the similarity effect (or very little). For instance, a team of U.S. researchers analyzed over 20,000 structured interviews conducted in a large organization to fill managerial-level positions. No similarity effect between applicants and interviewers was observed, whether for gender or race. These results are encouraging because they show that professionals are immune to the similarity effect with proper training and valid selection methods.
Our results have demonstrated something of considerable importance. When recruiters receive information from personality tests, they do not suffer from any personality-based similarity effect. Instead, they focus on hiring candidates who best fit the position of interest. Most hiring decisions imply at least one interview between candidates and interviewers, and very often, interviewers form impressions about candidates' personality traits through these interactions and not from personality testing.
We have yet to determine if these impressions are accurate and to what extent they are, considering that most interviews are short and that candidates engage in impression management. Therefore, a personality-based similarity effect may occur during selection interviews, mainly when interviews are unstructured.
Further research is therefore needed to examine if a personality-based similarity effect influences hireability ratings in selection interviews. To conclude, our results advocate for using personality testing in hiring decisions.
First, even if psychometric personality tests are not without limitations, they have a proven track record in predicting a range of outcomes such as employee engagement, performance, citizenship behaviors, or employee turnover, to name a few. Second, they can prevent recruiters from making biased decisions by considering solely interview impressions. As such, personality testing can be used in addition to interviews to make more accurate hiring decisions.