by: Nico Tschöpe and Oke Brandt
Interviewing remains one of the most commonly used methods when selecting or hiring talent (Roth & Huffcutt, 2018). Over time it has developed from a resume-, or CV-, based conversation to take on a more structured and competency-based format. Interviews were nearly always in-person, perhaps sometimes over the telephone; that is, until digital developments led to online and real-time video interviewing. Such technology saved travel and time costs and were heralded as the new dawn of interviewing.
Besides real-time video interviews, asynchronous videos interviews (AVIs) emerged from the technological progress. These interviews require candidates to record their responses to predefined questions, which are then submitted and later rated by assessors within the hiring organization. It led to even greater resource savings, and the new ability to share candidates’ responses across a hiring team and the option to refer back to the recordings when needed.
With progressing digitalization, developing technologies, emerging global recruitment requirements and the ongoing war for talent, new selection methods have shaped talent acquisition. Talent decision makers have embraced these digital interview tools. However to some degree, research lags behind evaluating these new tools and technology for their feasibility and application.
My colleague, Nico Tschöpe and I, along with many other IO psychologists and researchers continue to conduct studies to explore the impact of AVI on talent decision making – and how best to employ advancements in AI to improve the value of asynchronous interviews. The latest development in the analysis of AVIs is the use of AI to make the process even more efficient. Each video no longer requires a human rating. This trend means the number of videos that need to be actively reviewed by a human can be reduced substantially, saving recruiters precious time and increasing the efficiency of the recruitment process.
My colleague, Nico Tschöpe and I are delighted to have successfully submitted a symposium addressing latest trends in AVIs with a special emphasis on the analysis of them with the help of Artificial Intelligence for the 2020 SIOP Annual Conference. We thank all internal colleagues and external research fellows for their contribution to this.
Predicting Personality Using Automated Linguistic Analysis
One of our studies focused specifically on whether we can predict personality from the words spoken by candidates in an asynchronous interview. We used the off-the-shelf IBM Watson service, Personality Insights (PI) (IBM, 2019) to identify Big Five personality factors and facets and gathered data from 128 job applicants. We examined the construct validity of the results by considering the results from PI and those from the personality framework – shapes, (Justenhoven, Lochner, & Preuß, 2016).
The study generated some evidence for a reliable measurement of PI’s variables although showed that PI could not prove construct validity in the context of AVIs. Literature suggests that automated language analysis can only be valid when training data and input data both stem from similar sources. As IBM Watsons’s PI was trained based on data from a social media platform, it explains why construct validity was low. When using AI technology in AVI it is suggested to watch closely which tool is a good fit based on the training data and the development of the algorithm. Following this rationale, it is crucial to generate training data from statements of actual interviews, which are reviewed by psychologists and expert.
How Artificial Intelligence Influences People’s Decision-making in Video Interviews.
Advances in AI have meant that AI can now support asynchronous video interviews and “score” the responses given by candidates to preset questions. These automatically generated scores can then be presented to the hiring manager who then can make a decision as to whether or not to progress a candidate. Such technology helps to make the interview objective and fair, eliminating human bias, but, on the other hand, the hiring managers must not be negatively influenced by the automated scores. This study focused on how hiring managers are influenced by the AI-score. Using data collected by 321 participants rating five video interview excerpts on five scales.
The results of the study suggest that those evaluating candidate performance in asynchronous video interviews are not more skeptical of scores suggested by AI compared to those of professional human raters.
Asynchronous video interviews, scored by AI hold huge promise to save recruiters and HR professionals time, mitigate human bias in response scoring and offer candidates more flexibility and convenience in the interviewing process. With continued research on the impact AVIs and AI have on hiring decisions, organizations can implement such tools confident that they will fulfill their potential.
IBM (2019). Personality Insights – Interpreting numeric results [Documentation]. Retrieved from: https://console.bluemix.net/docs/services/personality-insights/numeric.html Accessed: September 4, 2019
Junior, J. C. S. J., Güçlütürk, Y., Pérez, M., Güçlü, U., Andujar, C., Baró, X., ... & Escalera, S. (2019). First Impressions: A Survey on Vision-based Apparent Personality Trait Analysis. IEEE Transactions on Affective Computing.
Justenhoven, R. T., Lochner, K., & Preuß, A. (2016). Validierung des adaptiven Persönlichkeitsfragebogens shapes für die Personalauswahl. Journal of Business and Media Psychology, 7, 8-19.
Lievens, F., Schollaert, E., & Keen, G. (2015). The interplay of elicitation and evaluation of trait-expressive behavior: Evidence in assessment center exercises. Journal of Applied Psychology, 100(4), 1169.
Roth, P. L., & Huffcutt, A. I. (2013). A meta-analysis of interviews and cognitive ability: Back to the future? Personnel Psychology, 12, 157-169.
About the AuthorVisit Website More Content by Oke Brandt