Assessing empathy would be inaccurate and subjective. The act of assessment may bias the final result. The random selection into two groups should do better tha

Assessing empathy would be inaccurate and subjective. The act of assessment may bias the final result. The random selection into two groups should do better than that, assuming the sample size is statistically significant. This experiment can simply be repeated to build more confidence in the result.