Selection is an important management tool because it is an integral part of the management process. If a company intends to be successful and manage its resources without wastefulness, then personnel selection must be done in a competent, time-efficient manner.
Why is selection such an important management tool?.
As selection is an integral part of the management process, it is important to know the different types of selection available to recruiters. Strict rank-order selection, traditional banding and SED banding are three types of selection, which will be discussed in this essay. .
When creating a test for selection purposes, it is important to know which criteria are considered invaluable to the organisation. Then, a complementary scoring system must be affixiated.
Strict rank-order selection is based on the assumption that if a predictor-criterion relationship is linear, the average job performance of selected applicants in rank-order, will be higher than that of any other method. It treats any differences in scores as meaningful and involves narrow bands, starting with the highest score and moving downward based upon any differences in scores, until all vacancies are filled. This type of selection depends on validity coefficient so not accurate for every sample. It results in fewer people per vacancy than other methods that aggregate test scores.
Traditional banding and SED banding are 2 kinds of banding that can be used for selection purposes. An example of traditional banding are expectancy charts, which are presented as bar graphs showing the probability of above job performance for different bands of test scores. The result is a histogram, which is useful in conveying the meaning of validity to employers. A disadvantage linked with traditional banding is the loss of value as a result of treating scores with each band as equal. Statistical significance testing is based on the assumption that different scores should be viewed as equal unless they are found to be "statistically significantly different", using the standard error of the difference (SED) between scores.
