The ongoing global pandemic has affected all the sectors equally in terms of both profit and loss. Many organizations are running into loss during this hard time. Therefore, organizations are now tending to cut down the employee cost in other words layoff, in order to balance the financial crisis or decrease the financial impact.
Voluntary employee turnover has a financial impact on organizations. It occurs when an employee leaves their position. For example, an employee who resigns, or simply leave their positions for other reasons is voluntary employee turnover.
Apart from layoff, voluntary employee turnover can create a big headache for many organizations during this pandemic. Therefore, foreseeing the probability of employee turnover would help in making hiring decisions as well as in taking relevant financial decisions during the unsure time.
Researchers and scientists from PredictiveHire, an AI recruiting startup has constructed a language model that can analyze from the open-ended interview questions answered by candidates to understand the probability of a candidate’s job-hopping. The research was carried out by Madura Jayaratne, Buddhi Jayatilleke. A total of 45000 responses were collected from candidates, who used a chatbot to attend an interview and self-rated their probability of Job-hopping. Researches evaluated five methods of text representations. They are term frequency-inverse document frequency (TF-IDF), LDS, GloVe Vectors for word representations, Doc2Vec document embeddings, and Linguistic Inquiry and Word Count (LIWC). Among the five methods, GloVe embedding showcased the good results. It showcased a positive correlation between words and the probability of Job Hopping. It also showed another positive correlation between job-hopping and openness to experience. These results would help companies in the financial area.
Along with the financial impact due to employee turnover, it also impacts decreased productivity as well as decreased employee morale. Even the trend for searching for a new job better than the existing is increasing. Therefore, companies have taken job-hopping as a critical factor to understand a candidate.
Earlier days the job-hopping was assessed from a candidate’s resume. It was a manual interpretation that makes work so difficult. Therefore, companies decided to make the work easier. Companies decided to understand the personality traits as well as job-hopping through interview questions. To find out the correlation between answers of the candidate and the probability of job-hopping, researchers built a regression model. This model uses the 5-point scale to rate the answers to understand the probability of job-hopping. Thus, it is understood that open-ended interview questions could predict the candidates’ turnover rate by using the model.
