Recent Trends in Human Resource Management

Monday, April 28, 2008

Selection Error

There are two types of selection error. In the "false positive error," a decision is made to hire an applicant based on predicted success, but failure results. In the "false negative error," an applicant who would have succeeded is rejected based on predictions of failure.
The False Positive Error
An organization that makes a false positive error incurs three types of costs. The first type of costs are those incurred while the person is employed. These can be the result of production or profit losses, damaged public relations or company reputation, accidents due to ineptitude or negligence, absenteeism, etc. The second type of costs are those associated with training, transfer, or terminating the employee. Costs of replacing the employee, the third type of cost, include costs of recruiting, selecting, and training a replacement. Generally, the more important the job, the greater the costs of the selection error.
The False Negative Error
In the case of false negative error, an applicant who would have succeeded is rejected because failure was predicted. Most false negative selection errors go unnoticed, except when the applicant is a member of a protected class and files a discrimination charge. Costs associated with this type of error are generally difficult to estimate. A situation in which the impact of both false positive and false negative selection errors can be detected and measured, however, is in professional sports such as football and basketball. Here, coaches and scouts analyze game films, physical statistics, scouting reports, and other data and decide whether they wish to draft a particular player. If they draft the player and his performance fails to meet expectations, a false positive selection error has occurred. Suppose, however, a team decides against drafting a player and another team chooses the individual. If the player subsequently turns out to be a star, the first team's rejection represents a false negative selection error.
Factors Affecting the Importance of Selection

Effective selection methods and procedures can result in fewer selection errors and their associated costs, and increased levels of performance and productivity due to hiring highly qualified applicants. The selection function takes on increased importance:
(1) when a job's base rate of success is low;
(2) when a job has greater importance to an organization; and
(3) when the selection ratio for a job is low.

Base Rate of Success.
Generally, the selection function is more important when a job's base rate of success is low. A low base rate of success indicates that relatively few employees reach an acceptable level of performance in a job. Improved selection procedures can raise base rates of success, thereby reducing costs associated with selection errors.
Job's Value to the Organization
The more important a job is to organizational effectiveness, the more important the selection function is. Selection errors are far more costly for important jobs than for jobs of lesser importance. One measure of a job's value to the organization is the standard deviation of job performance for a job (SDe). The SDe of job performance is a measure of the potential range, or variation, of the dollar value of job performance. For some jobs, differences in performance extremes (excellent to incompetent) have little effect in terms of dollar value to an organization. For example, variability in performance for a clerk/typist job is relatively insignificant compared with the effects of performance variability for the job of marketing manager.

Selection Ratios
A selection ratio is the proportion of applicants selected and placed to the number of job applicants for a job. The selection function increases in importance when the selection ratio is low enough so that meaningful differentiations can be made between job applicants. However, since there are costs associated with processing applicants, very low selection ratios may not be cost-effective for an organization.
Ref: Thomas.H.Stone

No comments: