Psychometric properties career of commitment instrument using classical test theory and graded response model

Abstract

A good instrument is one that is a measuring tool with good psychometric properties. This exploratory study aims to describe the psychometric properties of the career commitment instrument using the classical test theory approach and the graded response model. The data was obtained from a career commitment questionnaire with 12 statement items and 250 randomly selected respondents. The summated ratings (Likert) method was used for scaling, with five response options. Data were analyzed using R Studio's traditional Graded Response Model Theory Test technique. According to the findings, the career commitment instrument's quality has an estimated reliability of 0.77 (reliable) and a standard error measurement of 3.3. This instrument has a good Index of Endorsements and a Discrimination Index, with a classic and modern approach. Furthermore, the graded response model analysis revealed that 10 items fit and 2 did not. If given to respondents with low ability levels (θ=-2) to high ability levels (θ=2), this instrument provides complete information of 58.93 with a standard error of 1.0. As a result, an instrument is created that is ideal for companies to use in assessing their employees' career commitment.