The Reliability and Validity of the Clerical Test and Work Sample in Phonemin Company – Case Study Example
Case study 3 Reliability and Validity The reliability and validity of the clerical test and work sample are important considerations in deciding whether or not to keep them for use in recruiting. This is because these two elements will decide whether the two tests are Phonemin Company’s solution to the problem of ineffective assessment during recruitment. As such, it is imperative to assess how to interpret the reliability and validity of the results and the possible limitations.
The reliability results of the clerical test are supposed to be interpreted by getting the extent of the difference between the scores in the Time 1 and those of Time 2. The test would be inferred to be reliable if it produced consistent and stable results. Similarly, the reliability of the work sample should be interpreted by examining the difference between the corresponding scores in both Time 1 and Time 2. Since the two tests were re-administered a week after the first experiment, the correlation coefficient would work best in establishing the test-retest reliability of the two tests. At another level of interpretation, the results of the different raters should be established in order to account for the disparity that is occurs between different human observers. The internal consistency reliability of the two tests would be interpreted by assessing the difference between the corresponding scores produced by individual items of the two tests in the two times they were administered (Applications: Evaluation of Two New Assessment Methods for Selecting Telephone Customer Service representatives, Class notes, unidentified date).
The two tests are favorable enough for Phonemin to consider using them for keeps in selecting new job applicants. This owes to their fairly small difference between the scores they produced in both Time 1 and Time 2. For example, the difference between the mean scores of the two times in which the clerical test was administered is – 0.04 indicating that the clerical test is reliable. In addition, a difference of 0.16 between the mean score of Time 1 and Time 2 for the work sample shows that the test is fairly stable (Heneman, 2012).
The validity of the clerical test and work sample are to be interpreted by establishing how accurately they test measure applicants’ clerical speed, clerical accuracy, and interpersonal skills. In interpreting the results of the two tests, face validity would be the simplest for the raters and interviewers to establish because it would need them to consider whether the questions in the tests are addressing the said KSAOs. The construct validity of the two tests would be interpreted by establishing whether their individual items are measuring the intended variables and not others. Interviewers and raters who are experts in the field interviewed for would be best placed to assess construct validity. The criterion-related validity of the clerical test and work sample would be interpreted by examining how accurately these tests predict the future performance of prospective recruits (Applications: Evaluation of Two New Assessment Methods for Selecting Telephone Customer Service representatives, Class notes, unidentified date).
The high percentage agreement scores for raters in both tests are above seventy five and this shows the tests are favorable enough for Phonemin to consider using them “for keeps” in selecting new job applicants. The high scores for both applicants’ concern for customers and tactfulness in the two tests and for the two times in which they were administered show that the tests are quite on point (Blischke & Murthy, 2000).
A potential limitation with these tests is the length of tests not being long enough. Reliability increases with increase in the length of a test. For example, the four phone calls that were included in the work sample might not capture an applicant’s mastery of a given KSAO satisfactorily. The clerical test could be limited by the wording of the questions. The intensive use of terminology with which applicants are unfamiliar could be a potential set back with this test. Another disparity that should be kept in mind when interpreting the results of the tests is the disparity between the prevailing conditions at the time of administering the tests. For example, giving different instructions to applicants at the start of a test session is a disparity that can tamper with the reliability of the tests. The unpreparedness of the applicants is another limitation that can affect the reliability of the clerical test and the work sample. It is highly likely that the tests would produce inconsistent results if at Time 1, they were administered in the morning and in Time 2 they were administered sometime in a hot afternoon (Heneman, 2012).
Another limitation to be recognized when interpreting the results of these tests is the lack of good coverage of the criteria for the KSAOs under study. For example, the emphasis on the three chosen KSAO including clerical speed, clerical accuracy, and interpersonal skills might have sidelined others that are crucial to the consideration of an applicant’s tactfulness and concern for customers. Even so, there is a potential limitation in the tests if some of their items overlap (Blischke & Murthy, 2000). Finally, the design of the rating manual can be another limitation in interpreting the results of these results and so can the moderation of the analysis procedures of the results of individual applicant.
Blischke, W. & Murthy, D. (2000). Reliability: modeling, prediction and optimization. New York, NY.: Wiley.
Heneman, H. (2012). Staffing organizations. Middleton, WI: McGraw-Hill Irwin.
Unidentified author. Unidentified date. Applications: Evaluation of Two New Assessment Methods for Selecting Telephone Customer Service representatives, Class notes.