About

Welcome to the website Predictive Psychometrics!

We believe data science has a lot to contribute to the field of organisational behaviour and HRM and has become a very popular topic in contemporary management. People Analytics and modern psychometric techniques have increasingly helped managers and executives make informed decisions about their employees and organisational processes, being therefore associated with best practices for business. However, in order to fulfil its promise, standard barriers of sources of information need to be expanded.

Psychometrics is at the centre of the intersection between Statistics and Psychology, and aims primarily at quantifying elements which are not directly observable. Sentiment, feelings, beliefs, cognitive skills, knowledge, leadership and decision-making biases are among the intangible aspects that Psychometrics specializes in transforming into predictors. Each with their associated levels of uncertainty, but nonetheless providing an increase in predictive power.

Although Psychometrics has continuously developed its methods for more than a century, new techniques from Data Science have shown a great potential for understanding human behaviours in organisational settings. Hence, few initiatives exist in order to systematically investigate organisational behaviours through the combination of psychometric techniques and machine learning algorithms, which can in turn offer more accuracy for the prediction of organisational outcomes.

Limitations in predictive tasks usually concern modeling the chaotic nature of the human element, and we provide a scientific approach to resolve this challenge. Most significant increases in predictive power comes from opening new dimensions in the predictor spectrum, either by clever feature engineering or by tapping into novel sources of information.

The main goal of our work is thus to approximate these major fields of research (Psychometrics, Statistics, and People Analytics) into a cohesive framework, providing descriptive, predictive and prescriptive solutions to business.

We have currently applied the modelling techniques aforementioned to predict individual performance and staff turnover as well as to estimate organisational climate networks that can improve the quality of its measurement. Several products and publications are already underway as you can see from the navigation menu above.

Although this site introduces advanced statistical and computational research methods, it has been designed mostly using “plain language” in order to reach out to practitioner, young researchers and students who are not familiar with more complex techniques. For more detailed information, please refer to our publication list.