Using Binary Regression to Predict Employee Turnover Intention

The present study attempted to mathematically predict the employee turnover intentions in one of the Qatar-based companies following a quantitative methodology conceptually built on the binary logistic regression model. A questionnaire specifically developed to measure a five-factor impact: organizational culture, subjective social status, personal orientation, expectations, and career growth was administered to a random sample of 165 employees. Data were assessed in SPSS, and employee responses were dichotomized (agree/disagree) from a 5-point Likert scale in order to optimize predictive accuracy of the model. Findings indicated that each of the related constructs — organizational culture, social status, and personal orientation — had a high and significant negative correlation with turnover intentions, indicating that employees perceiving each of these perform worse in their search for the next job. In contrast, career growth was positively related to turnover intentions, as employees quit their respective workplaces in search of a promising future career. The model also displayed a fairly good predictive power (61.2% overall accuracy rate), validating the fact that binary regression is a powerful statistical tool that assists organizations in identifying employees at risk of quitting, enabling proactive and targeted measures to prevent their attrition.