Text Analysis in Human Resources: Competencies and Open-Ended Question Responses

Text Analysis and Human Resources

Text analytics and text mining have many uses in human resources. There is one main reason, there is a lot of text-based communication in almost every organization; employee surveys, performance reviews general employee feedback, resumes, “town hall” meetings, social media, and so on. HR departments realize that regularly analyzing and tracking this communication can improve morale, increase efficiency, and early identify problem areas to adjust or redeploy resources. Text analytics can be an important operational tool in the hiring process and in crisis management. It can also be an important research tool to help HR find out what it is doing well, what it isn’t doing well, and how it can improve. Text analytics HR research can also open windows into areas that merit further research to discover new ways of looking at work, employees, and their relationship to it.

A recent Lunch and Learn is one example. John Ford is a Senior Research Psychologist in the Office of Policy and Evaluation at the U.S. Merit Systems Protection Board. His presentation Competencies Needed for Success by U.S. Federal Employees shows how one can use text analysis to dig deeper into areas of human resources and how research in one area can lead to discovery in another.

The presentation looks at how to use and analyze open-ended question responses to measure if current training programs are hitting the right themes if training is concentrating on competencies that can be trained, and how training matches with employee needs. It also looks at how to detect evidence of emotional labor in job task descriptions to show that some jobs seem to inherently involve a higher emotional component than others. Recognizing this allows organizations to be mindful of the added stress in such jobs and be prepared to help employees (Guy, Newman, & Mastracci, 2014).

Finally, John presents text mining research that looked at how engaged employees described their jobs differently than less engaged employees. For example, engaged employees often claim lack of resources is a barrier to greater performance. “Give me additional resources and I can do better work.” Less engaged employees more often cite management or a poor work environment as reasons for low performance. This information was used to craft additional closed-ended questions to measure employee engagement (Ford, J., Nierle, D., Leeds, P., & Stetz, T., 2018).

The specific, actionable takeaways for HR are many.  The research shows how using good research techniques combined with text analytics you can go beyond the obvious and get deeper insights into your organization and your employees. The presentation also explains how you can use different features in QDA Miner and WordStat: coding, clustering, text mining, dictionaries, dictionary building, keyword-in-context, and other elements to analyze closed-ended and open-ended question responses. These techniques can be used to discover elements in HR policies that might elicit new training methods, make better use of patterns in answers from different groups of employees, and open up avenues for future research.

Some References

U.S. Merit Systems Protection Board. (2011). Making the right connections: Targeting the best competencies for training. Available at

Guy, M. E., Newman, M. A., & Mastracci, S. H. (2014). Emotional labor: Putting the service in public service. Routledge.

Ford, J., Nierle, D., Leeds, P., & Stetz, T. (2018, January). Text Mining Narrative Survey Responses to Develop Engagement Scale Items. In Proceedings of the 51st Hawaii International Conference on System Sciences.