People Analytics: Using HR Data Science to Predict Care Quality

Originally posted in 2019

On 18 June, we will host @Andy Charlwood (Professor of HRM and ThinkTank Academic Lead) and Danat Valizade as they talk us through data science in HR, giving real examples based on their recent work for the National Institute for Health Research.

To sign up, please click here. If you have participated in the ThinkTank research, you will be sent a copy of the report after the Launch webinar. If you would like to take part in the research, please register here.

There is a huge amount of hype about the potential of predictive analytics and data science to make HR more data driven and evidence based. At the same time, a range of evidence suggests that the use of predictive analytics in HR is quite limited. Although many organisations are now using predictive analytics to understand which employees are more likely to quit I haven’t come across many other interesting examples of predictive analytics in action. I have talked to a number of HR analytics leaders who have experimented with predictive analytics in projects but found the results underwhelming because the actionable insight from the analysis was limited.

"Does having more and more highly skilled care staff result in better quality care?"

     — Andy Charlwood

Over the last year, my colleague Danat Valizade and I have been working on a research project (funded by the National Institute for Health Research) with a care home operator to try to understand the relationship between care home staffing and indicators of the quality of care. Broadly, we have been asking the question, does having more and more highly skilled care staff result in better quality care? This is an important question, because the care home sector expected to provide high quality care for residents despite low levels of funding from the UK government (if we contrast the UK with countries like the Netherlands, UK funding is much lower). The care home operator provided us with data from their administrative systems detailing staffing levels for different job grades (care assistant, senior care assistant and registered nurse) week-by-week along with data on the characteristics of the care homes and their residents. This project has allowed us to experiment with predictive analytics techniques, comparing and contrasting these with more traditional forms of statistical analysis.

Danat and I are going to be sharing the provisional results of this project in a webinar on Tuesday 18th June. The exciting thing about this is that it will give us the opportunity to share what we have learnt about applying predictive analytics to real world people management problems. We’ll be talking about the importance of thinking hard about causality before starting any analysis, how traditional regression results compare to machine learning techniques and the strengths and limitations of different machine learning methods for addressing people management problems. If you are interested in finding out more about how to apply predictive analytics in HR, or if you have yourself used predictive analytics and want to tell us about your experiences and how they compare with ours, we hope you can join us for the webinar.

To register for the webinar on 18th June please register here.