A new study is offering credit unions suggestions on reducing employee turnover and improving performance.
The study, sponsored by the Filene Research Institute and entitled "Predicting Employee Turnover and Performance: Pre-Employment Tests and Questions that Work," was based on analysis of four credit unions. Authored by Murray R. Barrick, Tippie School of Management, University of Iowa, it suggests that when applicants are ranked on a set of criteria for predicting voluntary, avoidable turnover and performance, high ranking applicants have a substantially greater probability of performing well and staying with the organization.
Applicants who rank in the top 30% on the criteria for low turnover have a 95% probability of being with the credit union nine months later, whereas applicants who rank in the bottom 20% have only a 25% probability of staying, the study states. "In other words, the top 30% of applicants are almost four times as likely to remain with the credit union as applicants who rank in the bottom 20%," the author noted.
Applicants ranking in the top 30% on the performance predictor have a 70% probability of average performance or better and a 64% probability of superior performance or better in evaluations 3-6 months after hire. Those who rank in the bottom 20% have only a 39% probability of at least average performance and a 17% probability of superior performance.
These results are based on actual applicants to credit unions and actual outcomes.
Applicants who score well on the criteria that predict low turnover also score well on the criteria that predict high performance. Therefore, the study suggests, using these criteria for selection enhances performance and lowers voluntary avoidable turnover.
Barrick uses 10 criteria to predict turnover and performance. Three-"biodata" based on the applicants' background and experience -are readily collected on employment applications. The remaining seven are from questionnaires completed by applicants based on tests available commercially at low cost.
The study also addresses legal issues in selecting of new employees. Legally, new employee tests must be validated as effective predictors of relevant, job-related outcomes. There must also be a valid theoretical basis for why such a relationship between the predictor and the outcome exists.
"Appropriate testing and a few key items on an application can screen applicants to substantially reduce voluntary, avoidable turnover and significantly enhance performance on the job," says Bob Hoel, Filene executive director. "This report provides detail on how credit unions can apply Professor Barrick's results to improve employee performance and turnover rates."
