Vocational interests – the activities, processes and environments you prefer at work – are, compared to ability and personality, the neglected child of occupational psychology. This is partly thanks to a 1984 meta-analysis, which reported a weak correlation with job performance of just .1. However, recent focus on the idea of person-job fit has drawn attention back to this domain, and a new meta-analysis appears to further rehabilitate interests by showing a rather stronger relationship to performance.
Lead author Christopher Nye and his team gathered 60 studies by searching the literature for terms such as vocational interests, job performance, and turnover, and by perusing the bibliographies of texts such as interest inventory technical manuals. Half the studies followed the 1984 meta-analysis, 42 involved employment (the remainder looked at academic achievement), and these related interests to various measures of performance, such as job outcomes or organisational citizenship behaviours. Interests were measured in various ways, but common to many studies was John Holland's six-interest taxonomy, comprising work that is realistic (e.g. technical), investigative (research), artistic, social, enterprising, and conventional. Interest scores can be treated in a fairly absolute way - your standardised interest score is above average, so you should be somewhat suited to this role - but the team suspected that stronger relationships would be found in by taking a different approach. An individual's interests can be portrayed in terms of a personal ranking or interest profile, where the absolute scores matter less than the relative priorities; a strong investigative interest matters less if other areas matter even more to you. Matching individual interest profiles to job profiles produces 'congruence scores' that, by more closely reflecting fit or misfit, could be better predictors of outcomes.
Overall, the regression-based meta-analysis revealed the baseline relationship between interests and job performance to be .20, already twice as strong as the original 1984 analysis. Moreover, the congruence indices had a much stronger relationship, on average .36. Highest correlations were found around organisational citizenship behaviours, which makes sense: if you enjoy what you do you are more likely to go over and above what the job asks of you. Across the measures, the team found that "interested employees are likely to perform better, help others in the organization, and stay with the company longer." These correlations are substantial and suggest that interests are of greater value than previously believed.
The analysis also made it clear that choosing the right measure is critical. In Holland's taxonomy certain domains are more closely related than others: social interests can be partly compatible with artistic or enterprising job features, but opposed by realistic features. Following that example, studies that correlated a social interest measure with job performance found stronger relationships when the jobs were dominated by social features, weaker ones for artistic, and weakest for jobs that were essentially realistic. This re-emphasises that for interest to be valuable, it must be considered in terms of fit to a particular role, rather than as a more-or-less proxy of motivation. “Because past research has indicated that interests are not strong predictors of performance, vocational interests have seemingly been ignored in selection contexts”, concludes Nye's team, inviting a new wave of research to fill in the gaps.