The DOI matches its Crossref record, with a small difference in title or year; 94 of the 94 references it cites were found in a registry or an index; 7 of the 52 sources its claims cite were checked for support.
Do New Forms of Reading Pay Off? A Meta-Analysis on the Relationship Between Leisure Digital Reading Habits and Text Comprehension
Open accessAbstract
Previous research has evidenced a strong positive relationship between leisure print reading habits and reading comprehension across the lifespan. The rapid evolution of new forms of leisure digital reading could modify such a relationship. This meta-analysis extends previous research by analyzing the relationship between leisure digital reading habits and reading comprehension. We analyzed 40 effect sizes using multilevel analysis. Data involved 469,564 participants from studies published between 2000 and 2022. The average effect size reflects a small significant effect on reading comprehension (r = .055), which contrasts with the medium size effects found in the literature related to print reading habits and comprehension. This relationship is significantly moderated by the reader’s educational stage. At early stages (primary and middle school) negative relationships are observed between leisure digital reading and text comprehension, while at later stages (high school and university) the relationship turns positive. We highlight the different contributions that reading modalities and technological contexts have on our reading comprehension, especially across the lifespan. In sum, leisure digital reading does not seem to pay off in terms of reading comprehension, at least, as much as traditional print reading does.
Overview
CoChat Summary
TL;DR
This meta-analysis of 26 studies (40 effect sizes; 469,564 participants) found a small positive average association between leisure digital reading habits and reading comprehension (r = .055 after outlier removal), far smaller than print-reading associations. The association was moderated by educational stage, being negative in primary/middle school and positive in high school/university.
Key findings
- Overall association was significant but small: r = .055, 95% CI [.003, .107], p = .03, k = 39, n = 25 after removing one influential outlier; prediction interval was -.18 to .29. Heterogeneity was high: Q(38) = 2591.93, p < .001; I² = 98.9.
- Educational stage significantly moderated the effect (p = .047; R² = .57): middle school r = -.025, high school r = .085, primary/elementary r = -.095, undergraduate r = .070. Significant pairwise comparisons were middle vs. high school (p = .021) and primary vs. university (p = .030).
- Type of digital reading habit was not a significant moderator (p = .18): social-communicative r = .03, linear-informative r = .12, mixed r = .03.
- Data collection year was not significant in the main analysis (p = .121), but became significant when the outlier was included (b = -.014, 95% CI [-.27, -.001], p = .043). Publication bias was not indicated (Egger p = .427; published vs. unpublished p = .525).
Methods
- Systematic search in Web of Science, Scopus, PsychInfo, and Eric, plus a PISA/PIRLS-specific search, reference checks, expert contact, and forward search; publication years restricted to 2000–2022. The search yielded 1,866 records, with 26 studies meeting inclusion criteria and 40 independent effect sizes from 469,564 participants.
- Effect sizes were Pearson correlations, transformed to Fisher’s Z; a three-level random-effects multilevel meta-analysis used robust variance estimation (CR2), REML, and inverse-variance weighting, with multilevel meta-regressions for moderators and sensitivity/publication-bias analyses.
- Sample average age was 15.9 years (range 9.9–72.5), 54.84% female; most participants were middle school (35.89%) or high school (35.89%) students; 66.67% of effect sizes came from non-probability samples; digital reading habits were mostly measured by questionnaire (97.37%) with closed-answer formats (97.7%); reliability was α = .77 for comprehension tests and α = .59 for habit questionnaires.
Limitations
- The paper states that several studies came from the same research teams; most primary studies were cross-sectional, so correlations cannot inform causal effects; and most reading-habit measures were self-reports with closed-answer options, which may be biased and tied to measurement error.
- The paper also states limitations from heterogeneous scales/instruments, aggregated means that do not reflect individual differences, pseudo R² instability when fewer than 40 studies are included, restriction to native-language assessments, exclusion of readers with learning difficulties or physical impairments, and a dependent variable focused on traditional linear text comprehension rather than digital hypertext/Internet documents.
- Evident from the text: heterogeneity was very high (I² = 98.9) with a wide prediction interval, and some moderator categories were small (e.g., primary/elementary k = 2; undergraduate k = 6), limiting precision for moderator conclusions.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
6 Valid · 1 Not valid · 31 Not checked · 64 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Some studies evidenced positive associations between digital reading frequency and reading comprehension in adolescents (Botello-Peñaloza & López-Alba, 2014; Duncan et al., 2016; Islam et al., 2020; Jackson et al., 2011; Johnson, 2013; Kong et al., 2022; Wu & Peng, 2017) as well as in college students (Amiama-Espaillat & Mayor-Ruiz, 2017; Chen & Fang, 2013; Mežek, 2013).”
What the sources say
- Partly supportedcites Botello-Peñaloza & López-Alba, 2014 sourceExperimental — accuracy not yet measured
Our copy does support a positive association between the frequency of digital reading activity and reading-test performance: the Resultados y análisis bullet 'El uso del PC para lectura' reports that children who do their reading and writing tasks through digital media daily in class score on average 24% higher than those doing so one to two times per week (22% vs. monthly, 32% vs. never), and the Abstract reports a 21%-31% increase from ICT use for reading digital content. However, the population is not adolescents: the Fuentes de datos and Metodología sections state the analysis uses PIRLS 2011 data for Colombia restricted to fourth-grade primary school children (3,966 pupils, 150 schools), i.e. roughly 9-10 year-olds, so attributing this study to findings 'in adolescents' extends it to a different population than it studied. (LLM, conf=0.80)
- Not checked — no full text of the sourcecites Duncan et al., 2016 source
- Partly supportedcites Islam et al., 2020 sourceExperimental — accuracy not yet measured
The Results (‘Internet use and academic performance’; page not provided) report that adolescents who spent 2–4 hours online during weekends were more likely to have higher reading scores. The study measured time spent using the internet, not digital reading frequency, and it does not identify its reading-score measure as reading comprehension; it also reports a negative association for heavy weekday internet use. (LLM, conf=0.86)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“pseudo R 2 coefficient used in multilevel meta-regression analysis cannot be considered fully stable when the number of introduced studies is below 40 (Van den Noortgate et al., 2013).”
What the sources say
- Partly supportedcites Van den Noortgate et al., 2013 sourceExperimental — accuracy not yet measured
The Conclusions (page unavailable) warn that small numbers of studies may bias study-level variance estimates, which lends support to a general concern about stability. But the paper does not identify the pseudo R² coefficient as unstable or give a 40-study cutoff; it says 30 studies is often regarded as the smallest acceptable number and did not simulate fewer than 30. (LLM, conf=0.83)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“This pattern concurs with recent evidence showing a stronger negative association between the frequency of use of digital devices in the classroom and reading comprehension in US representative samples of fourth grade, as compared to eighth grade (Salmerón et al., 2023).”
What the sources say
- Supportedcites Salmerón et al., 2023 source
The Discussion states that the study used representative samples of US fourth- and eighth-grade students and found the negative association between classroom digital-device use and reading comprehension was almost twice as large in fourth grade. The Results (Model testing analyses) report estimates of −4.03 and −2.07, respectively, for students’ frequency of digital-device use in language arts class. (LLM, conf=0.94)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“a recent systematic review has evidenced that the correspondence between self-reported and logged social media use is far from perfect (Parry et al., 2021).”
What the sources say
- Supportedcites Parry et al., 2021 source
The moderator analysis (page unavailable) reports a correlation of r = 0.35 between self-reported and logged social media use. The Discussion also describes the overall correspondence between self-reports and usage logs as modest, supporting the citing sentence’s characterization. (LLM, conf=0.95)
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This relationship is significantly moderated by the readers' educational stage.”
plainstates an associationn = 469,564- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
35 more claims
Create a free account to see all 102 claims and every source verdict.
- Probably not a substantial claim. 39
- It may state two separate claims as one. 34
- The quote only partly states this claim. 2
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Acknowledgements
“Main Analysis | 0.073”
§ Acknowledgements
“[.012, .013]”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Descriptive characteristics of the studies
“including primary and secondary school students ( p = .03; R 2 = .38) and revealed higher effect sizes as the mean age of the students increased ( β = .012)”
§ Acknowledgements
“| Mean age of student sample from Primary and Secondary school | 30 | 18 | .028 | .011 | [.002, .05] | 6.04(1, 9.12) 2.53(1, 4.35) | 1289.91(28)*** | .36 | .036 |”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Descriptive characteristics of the studies
“the mean percentage of female participants was 54.84% (range: 41.592.0).”
§ Acknowledgements
“| Johnson (2013) | .207 | 90 | Yes: Research paper | NR | 10.63 (1.3) | 34.50 | NA”
Citations
Cites
The PRISMA 2020 statement: An updated guideline for reporting systematic reviews
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., et al. (2021) · BMJ · doi:10.1136/bmj.n71 · Cited by 108,283
Conducting Meta-Analyses in R with the metafor Package
Viechtbauer, W. (2010) · Journal of Statistical Software · doi:10.18637/jss.v036.i03 · Cited by 20,464 · cited 1×
“The statistical analyses were performed using the R 4.2.1 version with Metafor (Viechtbauer, 2010) and clubSandwich (Pustejovsky, 2020) packages.”
InfluentialStanovich, K. E. (1986) · Reading Research Quarterly · doi:10.1598/rrq.21.4.1 · Cited by 5,222
Outlier and influence diagnostics for meta-analysis
Viechtbauer, W., & Cheung, M. W.-L. (2010) · Research Synthesis Methods · doi:10.1002/jrsm.11 · Cited by 2,949 · cited 1×
“Furthermore, outliers’ cases were assessed with standardized deleted residuals larger than 1.96 (Viechtbauer, 2021; Viechtbauer & Cheung, 2010).”
InfluentialPISA 2018 Results (Volume I): What Students Know and Can Do
OECD (2019) · PISA · doi:10.1787/5f07c754-en · Cited by 2,350
Reading Ability: Lexical Quality to Comprehension
Perfetti, C. (2007) · Scientific Studies of Reading · doi:10.1080/10888430701530730 · Cited by 2,095
National Center for Educational Statistics
(2007) · Reference Reviews · doi:10.1108/09504120710838778 · Cited by 2,030
Doing Meta-Analysis with R: A Hands-On Guide
Harrer, M., Cuijpers, P., Furukawa, T. A., & Ebert, D. D. (2021) · doi:10.1201/9781003107347 · Cited by 1,850 · cited 1×
“Instead of aggregating the information for each sub-sample, with this multilevel analysis we were able to work with individual effect sizes while controlling dependence at the study-level, increasing statistical power as well (Harrer et al., 2021).”
On the Use of Beta Coefficients in Meta-Analysis.
Peterson, R. A., & Brown, S. P. (2005) · Journal of Applied Psychology · doi:10.1037/0021-9010.90.1.175 · Cited by 1,643 · cited 1×
“This procedure was developed by Peterson and Brown (2005) who proposed and evidenced an appropriate approach for imputing r-based meta-analytic effect sizes in behavioral studies using knowledge of corresponding beta coefficients.”
PISA 2018 Assessment and Analytical Framework
OECD (2019) · PISA · doi:10.1787/b25efab8-en · Cited by 1,614
Research works printed but not linked: 20
- *Altamura L. (2019). Digital reading habits and reading comprehension. [Unpublished bachellor dissertation, University of Valencia]. https://go.uv.es/D4FyR5I
- Borenstein M., 2019, Common mistakes in meta-analysis and how to avoid them
- Chen J. L., 2010, 20th Annual meeting of the Society for Text & Discourse
- Cuetos E., 2014, Batería de evaluación de los procesos lectores, revisada (5a ed.) [Battery for the reading process assessment, reviewed (5th ed.)]
- Cuetos F., 2016, Batería de evaluación de los procesos lectores en secundaria y bachillerato, revisada [Battery for the reading process assessment in middle and high school, reviewed]
- Duchêne M. C. A., 2010, Protocole Emilie—Evaluation de la compréhension de textes chez les collégiens [Emilie protocol—Text comprehension assessment for secondary school students]
- Dunn L. M., 1970, Peabody individual achievement test
- García V. (2019). Construcción de mapas conceptuales navegables y comprensión lectora: análisis de procesos y diseño de instrucción [Navigable concept maps and Reading comprehension: process analysis and instruction design] [Doctoral dissertation, University of Valencia]. https://roderic.uv.es/handle/10550/70352
- *Heesbeen S.E.A., Schaars M.M.H. (2021). Leisure reading and reading comprehension: The moderating role of executive functions [Unpublished master’s thesis]. Utrecht University.
- Lerkkanen M. K., 2018, YKÄ—Luku- ja kirjoitustaidon arviointimenetelmä yläkouluun [YKÄ—Reading test for lower secondary school]
- Leslie L., 2001, Qualitative reading inventory
- Macmillan Test Unit, 2000, Group Reading Test II, 6–14
- National Assessment Program Literacy and Numeracy, 2013, NAPLAN achievement in Reading, Persuasive Writing, Language Conventions and Numeracy: National Report for 2013
- *Norris T. (2010). Adolescent academic achievement, bullying behavior, and the frequency of Internet use [Doctoral dissertation, Kent State University]. ProQuest Dissertations and Theses Global. https://www.proquest.com/docview/193947450
- *Park S., Weng W. (2020). The relationship between ICT-related factors and student academic achievement and the moderating effect of country economic index across 39 countries. Educational Technology & Society, 23(3), 1–15. https://bit.ly/3rVyTc2
- Stothard S. E., 2010, York assessment of reading for comprehension. Secondary test
- SweSAT-R, 2006, Swedish scholastic aptitude test reviewed
- Wiederholt J. L., 2000, Gray silent reading tests GSRT
- Wilkinson G. S., 1993, Wilmington, DE: Wide Range, Inc
- Wilkinson G. S., 2006, Lutz, FL: Psychological Assessment Resources
Cited by
Vargas, C., Altamura, L., Blanco-Gandía, M. C., Gil, L., Mañá, A., Montagud, S., & Salmerón, L. (2024) · Research in Developmental Disabilities · doi:10.1016/j.ridd.2024.104675 · Cited by 50
Self-citationOrellana, P., Silva, M., & Iglesias, V. (2024) · Frontiers in Psychology · doi:10.3389/fpsyg.2024.1324055 · Cited by 59
The decline in reading for pleasure over 20 years of the American Time Use Survey
Bone, J. K., Bu, F., Sonke, J. K., & Fancourt, D. E. (2025) · iScience · doi:10.1016/j.isci.2025.113288 · Cited by 23
Cummins, J. (2024) · Language Culture and Curriculum · doi:10.1080/07908318.2024.2306286 · Cited by 9
Naumann, J., Brandhuber, L., Goldhammer, F., Eichmann, B., & Greiff, S. (2025) · Learning and Individual Differences · doi:10.1016/j.lindif.2025.102745 · Cited by 8
Metadata
- Authors
- Altamura, L., Vargas, C., & Salmerón, L.
- Affiliations
- Universitat de València (ES)
- Corresponding author
- Salmerón, Ladislao
- Venue
- Review of Educational Research
- Published
- 2023-12-12
- Volume
- 95
- Issue
- 1
- Pages
- 53-88
- ISSN
- 0034-6543 (ISSN-L)
- 1935-1046
- OpenAlex
- W4389639978
- Semantic Scholar
- 81c60f35219a12a1636d53b1eefd325967c2a809
- Cited by
- 59
- References
- 94
- Citations by year
- 2026: 27 · 2025: 21 · 2024: 10 · 2023: 1
- FWCI
- 3.63
- Type
- journal article
- Language
- en
- Open access
- green
- hdl.handle.net
- Licence
- cc-by-nc
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
- Locations
- Review of Educational Research (publishedVersion)
- Repository of Digital Objects for Teaching Research and Culture (University of Valencia) (submittedVersion, cc-by-nc)
- Funders
- Ministerio de Ciencia e Innovación (PID2020-118512GB-I00)
- Topics
- Social Sciences › Education › Literacy and Educational Practices
- Social Sciences › Psychology › Developmental and Educational Psychology › Reading and Literacy Development
- Social Sciences › Library and Information Sciences › Library Collection Development and Digital Resources
- Keywords
- reading for pleasure, digital reading, print reading, reading comprehension, reading habits, educational stages, lifespan development, e-books
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash and gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Do New Forms of Reading Pay Off? A Meta-Analysis on the Relationship Between Leisure Digital Reading Habits and Text Comprehension. https://app.cochat.ai/papers/10.3102/00346543231216463
- The paper
- Altamura, L., Vargas, C., & Salmerón, L. (2025). Do New Forms of Reading Pay Off? A Meta-Analysis on the Relationship Between Leisure Digital Reading Habits and Text Comprehension. Review of Educational Research. https://doi.org/10.3102/00346543231216463
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.