The steps begin with collection of all relevant information about the applicants and conclude with induction of the right candidate into the organization. The selection process in an organization is determined by the organization's strategy, goals and objectives, the tasks and responsibilities of the job and the organization, and the attributes required in the prospective candidates. The selection criteria for evaluation of candidates should be compliant to the organizational objectives. The selection methods should be designed in such a way that they evaluate candidates based on these criteria. To enable an accurate prediction of the candidates' success in the prospective job, the selection methods should meet the standards of reliability and validity.
A selection method can be considered reliable if it produces consistent results across different situations and times (Selection 2003, p.88). When a test taken by the same individual at different times produces results that are remarkably different, that test cannot be considered reliable. There are a few tests using which the reliability of a selection method can be measured.
The first test is the repeat or test-retest approach (Selection 2003, p.88). In this approach, a group of candidates is given a test which is repeated after an interval of 2-3 weeks. The similarity in the results obtained and an analysis of the scores determine the reliability of the test. ...
If the scoring pattern is similar for both, then the test is reliable. The split-halves procedure is a method where the same test is divided into two parts and given to the candidates. Again, the degree of similarity in the results determines the degree of reliability of the selection method.
The degree to which the success in the selection method reflects the candidate's success in the job is the measure of its validity. There are three general methods which can be used to measure or ensure the validity of a selection method, namely, the criterion validity, content validity and construct validity (Selection 2003, p.89).
Criterion validity refers to the correlation between the test scores and the job-performance scores. If the correlation is substantial, then the selection method can be considered valid. Content validity refers to the degree to which the content of a selection procedure represents the important aspects of job performance. Construct validity evaluates how well a selection method measures those characteristics in the candidates that are essential for his/her successful performance in the job. Depending on the jobs and the organizational factors, these methods can be used in different combinations to evaluate the validity of the selection methods.
Validating a test
The validation process comprises of five steps including analysing the job, choosing the tests, administering the tests, relating the test scores and the criteria, and cross-validating and revalidating (Dessler 2003, p.167).
Analysing the job involves identifying job descriptions and specifications. For this, human traits and other skills necessary for a job - for example, whether a person should possess people skills or