The voluntary resignation of employees has a direct financial impact on the organisation. Moreover, when the pandemic broke out, most organisations were seeking to cut staff costs, and voluntary employee resignation would cause great concern to the company. Therefore, the ability to predict employee turnover rates can not only help make smart hiring decisions but also help you save a lot of financial crises in an uncertain time.
Recognising that researchers and data scientists from AI recruitment startup PredictiveHire have built a language model that can analyze candidates’ open interview questions and infer the possibility of candidates changing jobs. The study, led by Madhura Jayaratne and Buddhi Jayatilleke, was conducted on the responses of 45,000 job applicants who used chatbots to conduct interviews and self-assess their likelihood of job-hopping.
The researchers evaluated five different text representation methods-term frequency-inverse document frequency (TF-IDF) abbreviation, LDS, GloVe Vectors for word representation, Doc2Vec file embedding, and language query and word count (LIWC). However, the GloVe embedding provides the best results, highlighting the positive correlation between the word sequence and the likelihood of an employee leaving.
The researchers further pointed out that there is also a positive correlation between employees’ job-hopping and their “open experience”. Because companies can provide the same forecasts for freshmen, companies that change careers can bring considerable financial benefits to the company.
In addition to the impact of new employees onboarding or outsourcing work to finances, increasing employee turnover rates may also reduce productivity and undermine employee morale. In fact, in this competitive landscape, the trend of leaving work in order to find a better job has received widespread attention. Therefore, it is critical for companies to assess the possibility of candidates for job-hopping before choosing.
Traditionally, this assessment is done by browsing the candidates’ resumes; however, manual intervention makes the process tiring and inaccurate. In addition, this method is only suitable for professionals with work experience and is useless for novices and amateurs. Therefore, the researchers decided to use the interview answers to analyze the personality traits of candidates and their chances of voluntarily leaving.
To keep up to date on all things “Hiring with Ai” subscribe to our blog! 😀
Get our insights newsletter to stay in the loop on how we are evolving PredictiveHire