Improving compassion measurement in the workforce by analyzing users’ mouse-cursor movements

Compassion is often an important characteristic of effective employees in the workplace. To measure improvements in employee compassion, identify shortcomings, and even screen applicants, researchers and practitioners have sought reliable measures of compassion. Measuring compassion, however, can be very difficult because people are hesitant reporting a lack of compassion, due to their desire to be viewed favorably by others. We draw on the domain of human-computer interaction to help address this issue. Namely, we propose a theoretically-grounded methodology for helping differentiate between high and low compassion by monitoring users’ mouse-cursor movements in specialized online surveys. We find that people with lower compassion show greater deviation in their mouse-movement trajectories when answering affirmatively that they are compassionate. Based on this finding, we suggest future research to use this deviation measure to help correct for bias in self-reported compassion levels both in a research and professional setting.