Reading from paper compared to screens: A systematic review and meta‐analysis

Background Given the increasing popularity of reading from screens, it is not surprising that numerous studies have been conducted comparing reading from paper and electronic sources. The purpose of this systematic review and meta‐analysis is to consolidate the findings on reading performance, reading times and calibration of performance (metacognition) between reading text from paper compared to screens. Methods A systematic literature search of reports of studies comparing reading from paper and screens was conducted in seven databases. Additional studies were identified by contacting researchers who have published on the topic, by a backwards search of the references of found reports and by a snowball search of reports citing what was initially found. Only studies that were experiments with random assignment and with participants who had fundamental reading skills and disseminated between 2008 and 2018 were included. Twenty‐nine reports with 33 identified studies met inclusion criteria experimentally comparing reading performance ( k = 33; n = 2,799), reading time ( k = 14; n = 1,233) and/or calibration ( k = 11; n = 698) from paper and screens. Results Based on random effects models, reading from screens had a negative effect on reading performance relative to paper ( g = −.25). Based on moderator analyses, this may have been limited to expository texts ( g = −.32) as there was no difference with narrative texts ( g = −.04). The findings were similar when analysing literal and inferential reading performance separately ( g = −.33 and g = −.26, respectively). No reliable differences were found for reading time ( g = .08). Readers had better calibrated (more accurate) judgement of their performance from paper compared to screens ( g = .20). Conclusions Readers may be more efficient and aware of their performance when reading from paper compared to screens.