A Brief Review on The Datafication of Talent
“The Datafiction of Talent: How Technology is Advancing The Science of Human Potential at Work”
Tomas Chamarro-Premuzic, Reece Akhtar, Dave Winsborough and Ryne A Sherman (2017)
What are the researchers interested in?
“[Talent Management] is focused on identifying, developing, and engaging employees to increase organizational effectiveness.”
Unsurprisingly, if an organization is unable to accurately predict and understand human behaviour, it will not be able to manage its workforce to its full potential. Thus, researchers turn to psychological assessments as a core foundation of effective talent management practices. Advancements are allowing talent management practices to be less based on intuition and increasingly based on research and data.
A search for peer-reviewed research reveals that there is a lack of scholarly scientific research on Talent Analytics. Subsequently, these conclusions shone a light on several issues facing academics and HR professionals. Chamarrow-Premuzic (2017) and colleagues believe the issue is three-fold: professionals require substantial training so that they are able to use and interpret psychological assessments correctly; academics will need to validate the emerging technologies and practices, and without compelling scientific evidence organizations must resist the “hype” surrounding big data and new talent assessment tools.
What methods did they use?
The researchers [with their own organizational and research experiences] aim to raise awareness regarding the utility, effectiveness, and current limitations of three promising innovations regarding the datafication of talent. In doing so, they hope this will increase the use of, and benefit from a data-driven and evidence-backed approach to identify and develop talent (Chamarrow-Premuzic et al., 2017).
Machine-learning Algorithms
Digital Interviews
Gamification
What did they end up finding?
Machine-learning Algorithms
A candidate's digital records (e.g. social media footprint) can be translated into a psychological profile; for example, their personality, cognitive ability and values. In other words, data-mining has shown compelling academic evidence for the validity of job performance, leadership potential, and abnormal work behaviours. However, the most important consideration is its ethical and legal constraints. Firms who are intending to use digital tools will have to allow candidates to have full control over their data, and decide whether or not to share their psychological profiles with recruiters.
Digital Interviews
Hiring managers can post their questions on a platform to ensure validity and reliability (if it is repeatable). Algorithms from these digital interview platforms can flag and interpret relevant talent signals (facial expressions, tones, emotions, language) replacing human observation with ‘computer vision’ to make inferences in creating a psychological profile – an estimate of their potential fit for the role. The process of digital interviews provides standardization of the interview process, making it more objective and cost-efficient whilst reducing interviewer biases.
A common question in the research field for these platforms is their tendency to reinforce certain biases that are common in any interviewing process. A certain limitation researchers have noted is that people responsible for making hiring decisions are themselves biased, so “we should not expect AI to erase that problem” (Chamorro). A way to address this issue is to focus less on individual traits but more on group outcomes (e.g. 360-degree reviews).
Gamification
Companies provide psychometric tests with the goal in mind of enhancing candidates’ experiences, including, applying game-like features and interactive and immersive scenarios (Chamarro-Premuzic, 2017). The digital revolution has undeniably fuelled a proliferation of game-based assessments. For example, Pymetrics leverages behavioural science and AI with tests such as the Balloon Analogue Risk Task, which assesses candidates’ risk-taking tendencies by examining how far they allow a self-inflating balloon to expand. A disadvantage to consider is that the more interesting and enjoyable the assessment experiences are, the less predictive it may seem to be (Chamarro-Premuzic et al., 2018).
Below we have synthesized a few benefits of each innovation and highlighted the overarching ethical considerations
Why does this matter for organizations?
Recent technological developments have provided organizations with new tools and assessments for talent management. Most organizations understand that high talent density is imperative in building a fiercely competitive corporate culture, thus a perennial challenge is to use the right tools that provide high validity and reliability. The authors highlight three main technologies which are circulating through recruitment areas, providing insight on how it is beneficial and also what are some disadvantages associated with them; in hopes that HR professionals are able to utilize this knowledge to accurately implement new data-led assessments – whilst taking into consideration the ethics surrounding data rights.
References
Chamorro-Premuzic, T., Akhtar, R., Winsborough, D., & Sherman, R. A. (2017). The datafication of talent: How technology is advancing the science of human potential at work. Current Opinion in Behavioral Sciences, 18, 13-16.
Chamorro-Premuzic, T., Winsborough, D., Sherman, R. A., & Hogan, R. (2016). New talent signals: Shiny new objects or a brave new world?. Industrial and Organizational Psychology, 9(3), 621-640.