Fiction Reading Has a Small Positive Impact on Social Cognition

A Meta-Analysis

Dodell-Feder, David
Tamir, Diana I.

Institute for Technology in Psychiatry, McLean Hospital, Belmont, Massachusetts, and Department of Psychiatry, Harvard Medical School
Department of Psychology, Princeton Neuroscience Institute, Princeton University

David Dodell-Feder is now with the Department of Psychology at the University of Rochester.

We thank the researchers who shared data and information—Matthijs Bal, Jessica Black, Maja Djikic, Dan Johnson, David Kidd, Emy Koopman, Raymond Mar, Micah L. Mumper, M. E. Panero, Dalya Samur, Deena Weisberg—and Steven Worthington at the Institute for Quantitative Social Sciences at Harvard University for assistance with data analysis. Both authors developed the study concept. D. Dodell-Feder collected and analyzed the data under the supervision of D. Tamir. D. Dodell-Feder wrote the first draft of the manuscript, and D. Tamir critically revised the manuscript. Both authors approved the final version of the manuscript for submission.

David Dodell-Feder, Institute for Technology in Psychiatry, McLean Hospital, 115 Mill Street, Belmont, MA 02478 [email protected]

Received July 27, 2017

revised November 1, 2017

Accepted November 13, 2017

Journal of Experimental Psychology: General 147(11):p 1713-1727, November 2018. | DOI: 10.1037/xge0000395

Scholars from both the social sciences and the humanities have credited fiction reading with a range of positive real-world social effects. Research in psychology has suggested that readers may make good citizens because fiction reading is associated with better social cognition. But does fiction reading causally improve social cognition? Here, we meta-analyze extant published and unpublished experimental data to address this question. Multilevel random-effects meta-analysis of 53 effect sizes from 14 studies demonstrated that it does: compared to nonfiction reading and no reading, fiction reading leads to a small, statistically significant improvement in social-cognitive performance (g = .15–.16). This effect is robust across sensitivity analyses and does not appear to be the result of publication bias. We recommend that in future work, researchers use more robust reading manipulations, assess whether the effects transfer to improved real-world social functioning, and investigate mechanisms.