Estimation of the Reliability of Job Analysis Data Using Generalizability Theory

The major aim of this research was to introduce Generalizability Theory and its application to estimate the reliability of job analysis data. Twenty employees of a state bank certified administrative specialist jobs were randomly selected from among qualified employees. After training, they were given job skills questionnaire to rate the importance of each of the skills for the job of administrative experts to determine. To estimate the reliability of job analysis data in the framework of Generalizability Theory, a five-step approach was introduced that can be used to estimate the reliability of job analysis data. The steps are: 1- Identify the object of measurement, 2- identify facets of measurement, 3- identify an appropriate measurement design, 4- estimate variance components and their interpretation and 5- identify the type of decision. It became apparent that the generalizability theory is useful to estimate the reliability of job analysis. It provides the opportunity for job analysis in such a manner that in different situations such as estimating the contribution of each systematic error sources, such as career history, age, and place of work assessors and achieving the desired reliability, it can perform the required activities.