The Pros And Cons Of The Bell Curve

Performance management processes that follow a Gaussian distribution (bell curve) and primarily focus on past performance rather than the future promise have come under critical focus in several research papers in recent years. A comprehensive literature review of the same reveals some interesting insights. In the earlier half of twentieth century, Ferguson (1947) suggested that “ratings for a large and representative group of assistant managers should be distributed in accordance with the percentages predicted for a normal distribution.” This concept that was proposed by Ferguson carried on over the years and subsequently also led several management scholars and academicians to assume that normal distribution in the performance of a job is

To nurture manpower in the organization that can work on both exploitative and exploratory endeavors, the appraisal process needs to be more development-oriented with long-term orientation (Chan, Shaffer, and Snape, 2004; Collins and Clark, 2003). Normally, appraisals have been found to be result based with little weight given to behavioral based criteria. The system must be adaptable to shift between the twin criteria of evaluations when assessing employees working on exploratory or exploitative work activities. Many times, force fitting results in distributions that are not actually normal but are forced to be so. Several scholars have in fact gone against the prevalent view and argued that employee performance may not always follow the normal distribution it is expected to (Bernardin & Beatty, 1984; Murphy & Cleveland, 1995; Saal et al., 1980; Schmidt & Johnson, 1973). It was proposed by Jacobs (1974) that in roles such as Sales there is always a minority group of employees who possess superior selling abilities as compared to the general group, and that they tend to dominate revenue generation. In such cases where it is seen that performance output does not follow a normal distribution, there is reason to suppose that non-normal power law distributions may then be applicable (West & Deering, 1995). Many non-normal distributions such as Pareto’s Law (1897) have fatter tails as compared to a normal curve. Paretian probability distributions therefore allow a greater number of extreme values to be present in the curve. Studies performed by researchers such as Aguinis & O'Boyle (2014) have in fact shown that contrary to popular belief, performance distributions usually tend to follow a Paretian distribution. Force fitting a Paretian distribution curve into a normal distribution curve results in a flawed bunching of employees resulting in a decrease in motivation, increased tendency to quit