It's certainly not total evidence, and I rather imagine there are limits to the effect if it exists at all. But 118 students isn't that small of a sample (if I

It's certainly not total evidence, and I rather imagine there are limits to the effect if it exists at all. But 118 students isn't that small of a sample (if I'm picking a methodological criticism, it'd be that it's a sample of college students rather than the general population). And the exercise seems reasonably likely to provide some of the reported effect, and those effects fit with other studies that hold reading/writing fiction improves people's ability to model others better and some degree of empathy.