Research Brief:
Teller

Setting:
The retail banking division within one of the nation's largest banks sought to reduce turnover and improve profitability through improved selection practices within a large metropolitan market. Specifically, the bank hoped to identify the personality characteristics required of high performing incumbents in the teller role for the purposes of developing a valid selection model. The bank provided an incumbent sample of 75 tellers for whom performance and tenure data was matched with scores on the Craft Personality Questionnaire (CPQ). The resulting sample provided the basis for the development and local validation of a formal teller selection model.

Model:
Using the data provided by the bank, a model was developed using both quantitative and qualitative performance criteria. Specifically, the model identified ideal ranges on each of the CPQ's "basic eight" scales and calculated an overall score based on an individual's results in comparison with the ideal. This score was reported as a low, moderate or high rating based on an overall assessment of compatibility to the teller role.

Findings:
The teller selection model provided a statistically significant correlation (r=.622, p<.001) to performance based on cross-selling activity between 2002 and the first quarter of 2004. High scoring incumbents achieved a level of sales performance that was 98% higher than that of their lower scoring peers. Additionally, the model provided a significant correlation to the incumbents' most recent performance reviews as provided by their immediate supervisor (r=.332, p<.05). In short, the selection model accurately predicted performance across the entire scope and context of the teller role.

To properly asses the applicability of these findings outside of the bank referenced above, a cross-validation study was conducted. To ensure that local market and cultural conditions did not exert undue influence on the teller selection model, the cross validation sample included a regional bank within a rural Midwestern US market. In this cross validation study, the teller model produced statistically significant correlation to recent performance reviews based on two ratings provided by the tellers' immediate supervisors: selling services (r=.232, p<.05) and dealing with customers (r=.237, p<.05).

Conclusion:
Across culturally and economically diverse samples, the teller model accurately predicted performance consistent with the critical success factors found within the teller role. This study provided evidence of strong criterion validity supporting the appropriateness and strength of the teller selection model as an employment selection assessment.