Longitudinal development of theory of mind in adolescence and its associations with fiction reading experience.

May combine two claims.

Two short-term longitudinal studies in 9.5-year-olds and 13- and 16-year-olds showed small within-person changes in theory of mind performance over 12 and 18 months respectively.

“Additionally, two short-term longitudinal studies in 9.5(Lecce et al., 2017) and 13- and 16-year-olds (St ę pie ń -Nycz et al., 2021) showed small within-person changes in ToM performance over 12 and 18 months respectively.”

• cites Lecce et al., 2017 sourcePartly supported

  The study summary in passage [1] (no page given) describes a three-wave study following children from about age 9.5 to 10.5 and reports growth in ToM task performance. It does not characterize the changes as small or specifically as within-person changes. The citation to Lecce et al. is attached to the 9.5-year-old, 12-month study, not the separate study of older children. (LLM, conf=0.83)

  Experimental — accuracy not yet measured
• cites Stępień-Nycz et al., 2021Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

The Silent Film and Triangles tasks are reliable and valid indicators of theory of mind for this age range.

“the tasks are reliable measures of ToM as indicated by strong test -retest reliability (e.g., Devine & Hughes, 2016) and are valid indictors of ToM as indicated by convergent validity with other advanced measures such as the Strange Stories task (e.g., Devine et al., 2024)”

• cites Osterhaus & Bosacki, 2022 sourceNot checked — no full text of the source
• cites Devine & Hughes, 2016 sourceNot checked — no full text of the source
• cites Devine et al., 2024 sourcePartly supported

  The Theory of mind section says that longitudinal evidence supports the validity of the Silent Film and Triangles tasks for children in this age range. It does not identify convergence with the Strange Stories task as that evidence. The citing sentence attributes test–retest reliability to Devine and Hughes (2016), not to this paper. (LLM, conf=0.83)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

A small-scaled longitudinal study following theory of mind development in 26 8-year-old children showed significant improvement in performance over a 2-year interval.

“For instance, a small-scaled longitudinal study following ToM development in 26 8-year-old children (Bosacki, 2013) showed signi fi cant improvement in performance over a 2-year interval.”

• cites Bosacki, 2013 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Children and adolescents who read more classic literature showed more prosocial behavior and better social adjustment than their peers, while controlling for socioeconomic status, migration background, and nonverbal intelligence.

“They found that children and adolescents who read more ' classic literature ' ( fi ction that is not contemporary but is still widely read) showed more prosocial behavior and better social adjustment than their peers”

• cites Lenhart et al., 2023 sourceSupported

Classified as probably not a substantial claim.

The previous study showed that self-reported fiction reading behavior, but not nonfiction reading, was concurrently associated with theory of mind performance.

“Additionally, the previous study showed that self-reported fi ction reading behavior, but not non fi ction reading, was concurrently associated with ToM performance, consistent with meta-analytic evidence described above suggesting speci fi c links between reading fi ction and ToM (van der Kleij et al., 2022).”

• cites van der Kleij et al., 2022 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

More reading was associated with higher levels of later prosocial behavior.

“More reading was associated with higher levels of later prosocial behavior.”

• cites Wan Mak and Fancourt, 2020 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Differences in reading behavior or reading interest seem to increase during adolescence.

“but also a time in which differences in reading behavior or reading interest seem to increase (Nagel & Verboord, 2012).”

• cites Nagel & Verboord, 2012 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Parental reports of children's leisure reading at age 7 predicted prosocial behavior at age 11.

“Using data from the U.K. Millennium cohort study, another study by Wan Mak and Fancourt (2020) found that parental reports of children ' s leisure reading at age 7 predicted prosocial behavior at age 11.”

• cites Wan Mak and Fancourt, 2020 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Both theory of mind ability and reading experience have been linked to individual differences in verbal ability, gender, and socioeconomic status.

“Both ToM ability and reading experience have for instance been linked to individual differences in verbal ability, gender, and socioeconomic status (e.g., Devine et al., 2023; Devine & Hughes, 2018; McGeown et al., 2016; Milligan et al., 2007; van der Kleij et al., 2022).”

• cites Devine et al., 2023 sourceNot checked — no full text of the source
• cites Devine & Hughes, 2018 sourcePartly supported

  The sections on socioeconomic status and false-belief understanding report a modest association between family SES and children’s false-belief understanding, and describe language ability as closely related to false-belief performance. The Theoretical Implications section notes that the sibling association varied with the proportion of girls in a study, but does not establish a general link between gender and theory-of-mind ability. This paper does not examine children’s reading experience or its links to verbal ability, gender, and SES, so it supports only part of the cited sentence. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites McGeown et al., 2016 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Narrative fiction reading experience is associated with later verbal ability, where children who read more books or novels perform better on tasks measuring children's verbal ability.

“It is noteworthy that narrative fi ction reading experience is also associated with later verbal ability (Cain & Oakhill, 2011; Sullivan & Brown, 2015; van der Kleij et al., 2022a), where children who read more books or novels perform better on tasks measuring children ' s verbal ability.”

• cites Cain & Oakhill, 2011 sourcePartly supported

  The Results (‘Relations between reading habits and reading ability,’ page unavailable) report that reading habits were positively correlated with vocabulary knowledge, and the vocabulary-growth analyses found that reading experience predicted later vocabulary. The study measured general reading habits, including reading for pleasure and books read, rather than isolating narrative fiction or novels; it tested vocabulary rather than verbal ability broadly. (LLM, conf=0.86)

  Experimental — accuracy not yet measured
• cites Sullivan & Brown, 2015 sourcePartly supported

  The TABLE 5 discussion (page not provided) links more frequent book reading at ages 10 and 16 with higher vocabulary scores, supporting the association with later verbal ability. The paper measures book reading generally, however, and does not establish that the books were narrative fiction or novels. (LLM, conf=0.87)

  Experimental — accuracy not yet measured
• cites van der Kleij et al., 2022aNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Classified as probably not a substantial claim. May combine two claims.

ToM tasks measure the ability to employ mental states in explanations and predictions of behavior and can be distinguished from tasks that may only require emotion recognition or only require empathic sharing of mental states.

“These ToM tasks measure the ability to employ mental states in explanations and predictions of behavior and as such can be distinguished from tasks that may only require emotion recognition (e.g., Baron-Cohen et al., 2001; Oakley et al., 2016), or only require empathic sharing of mental states (e.g., Quesque et al., 2024; Reniers et al., 2011).”

• cites Baron-Cohen et al., 2001 sourceNot checked — no full text of the source
• cites Oakley et al., 2016 sourceSupported
• cites Quesque et al., 2024 sourceSupported

May combine two claims.

There were bidirectional associations between ToM and reading experience between ages 12.5 and 15, where more fiction reading experience at age 12.5 was associated with better performance on the ToM task at age 15, and better ToM performance at age 12.5 was associated with more fiction reading at age 15.

“We also found bidirectional associations between ToMand reading experience between ages 12.5 and 15, where more fi ction reading experience at 12.5 was associated with better performance on the ToM task at 15, and better ToM performance at 12.5 was associated with more fi ction reading at 15.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Time spent reading fiction was correlated with all reading experience measures and the vocabulary measure.

“The time spent reading fi ction was correlated to all reading experience and the vocabulary measure.”

May combine two claims.

In the mediation model, there were significant effects from reading experience at age 12.5 to verbal ability at age 13.5, and significant links from verbal ability at age 13.5 to ToM at age 14.

“we found signi fi cant effects from reading experience at 12.5 to verbal ability at age 13.5, as well as signi fi cant links from verbal ability at age 13.5 to ToM at age 14.”

• Also stated in: § Figure 3 · p. 9 · plain

  “Additionally, it showed that earlier reading experience was associated with later verbal ability and that verbal ability was associated with later ToM ability.”

May combine two claims.

Constraining the cross-lagged paths from age 14 to age 15 to be equal did not decrease model fit, χ2(1)=0.13, p=.71, indicating that the relative strength of the paths from ToM to Fiction Reading and from Fiction Reading to ToM did not differ for both time points.

“Second, constraining the cross-lagged paths from age 14 to age 15 years to be equal did not decrease the model fi t, χ 2 (1) = 0.13, p = .71, indicating that the relative strength of the paths from ToM to Fiction Reading and from Fiction Reading to ToM did not differ for both time points.”

May combine two claims. The quote only partly states this claim.

In the cross-lagged model without control variables, there were bidirectional associations from amount of reading fiction at age 12.5 to ToM ability at age 14, and from ToM ability at age 14 to amount of fiction reading at age 15.

“We also found bidirectional associations from the amount of reading fi ction at age 12.5 to ToM ability at age 14, and from ToM ability at age 14 to the amount of fi ction reading at age 15.”

• Also stated in: § Model Outcomes Testing Bidirectional Associations · p. 6 · hedged

  “This positive relation indicates that participants who read more fi ction at age 12.5 performed better on the ToM tasks at age 14, and better ToM ability at age 14 was associated with more fi ction reading at age 15.”

May combine two claims.

In the model for time spent reading fiction, the main effects for parent education and gender were not significant.

“The main effects for parent education and gender were not signi fi cant, p > .05.”

May combine two claims. The quote only partly states this claim.

Girls scored higher on fiction reading experience at age 14, β=.23, p=.006, and age 15, β=.20, p<.001.

“age 14, β = .23, p = .006, and age 15, β = .20, p &lt; .001.”

The quote only partly states this claim.

There were bidirectional associations between theory of mind and reading experience.

“There were bidirectional associations between ToM and reading experience, but these effects disappeared after controlling for verbal ability, gender, and parent education.”

May combine two claims.

The effect of gender on Silent Film scores was only marginally significant, with a trend toward better performance among girls compared to boys.

“The effect of gender was only marginally signi fi cant, estimate = -.51, SE = .27, t = -1.93, p = .06, with a trend toward better performance among girls compared to boys.”

May combine two claims.

Verbal ability predicted variance in ToM at age 14, β=.24, p=.04, and fiction reading at age 14, β=.32, p=.006, and age 15, β=.29, p=.002.

“Verbal ability predicted variance in ToM at age 14, β = .24, p = .04, and fi ction reading at age 14, β = .32, p = .006, and age 15, β = .29, p = .002.”

May combine two claims.

Although children showed overall improvement in ToM, they did not vary in their rate of growth.

“This showed that, although children showed overall improvement, they did not vary in their rate of growth.”

May combine two claims.

There was initial evidence for a developmental benefit of narrative exposure for children's ToM, as well as an effect in the opposite direction.

“We found initial evidence for a developmental bene fi t of narrative exposure for children ' s ToM, as well as an effect in the opposite direction.”

May combine two claims.

Even in middle childhood and early adolescence there is still meaningful development of ToM, although children no longer vary in their rate of growth.

“We showed that even in middle childhood and early adolescence there is still meaningful development of ToM, although children no longer vary in their rate of growth.”

Classified as probably not a substantial claim.

In middle childhood and adolescence, ToM is moderately associated with language skills, regardless of the type of ToM task.

“Similarly, in middle childhood and adolescence, ToM is moderately associated with language skills, regardless of the type of ToM task (Osterhaus &amp; Bosacki, 2022).”

• cites Osterhaus & Bosacki, 2022 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Exposure to fiction texts is associated with better performance on theory of mind tasks directly after this exposure.

“Although some experimental studies show that exposure to fi ction texts is associated with better performance on ToM tasks directly after this exposure (Dodell-Feder &amp; Tamir, 2018), this does not necessarily mean that there is also a long-term or developmental bene fi t.”

• cites Dodell-Feder & Tamir, 2018 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

The positive association between reading and social cognition seems particularly found for fiction reading and measures of social cognition.

“Although it has been suggested that reading in general could be bene fi cial, this positive association seems to be particularly found for fi ction reading and measures of social cognition (Wimmer et al., 2024).”

• cites Wimmer et al., 2024 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Difficulties with reading could in turn impact motivation and enjoyment of reading in middle childhood.

“It has also been shown that dif fi culties with reading could in turn impact motivation and enjoyment of reading in middle childhood (Torppa et al., 2020; van Bergen et al., 2021).”

• cites Torppa et al., 2020 sourceNot checked — no full text of the source
• cites van Bergen et al., 2021 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

There is a moderate association between verbal ability and children's false belief understanding before age 7, independent of age and regardless of the type of false belief task.

“We know from previous research that there is a moderate association between verbal ability and children ' s false belief understanding before age 7, independent of age and regardless of the type of false belief task (Milligan et al., 2007).”

• cites Milligan et al., 2007 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Preregistration of hypotheses, analyses scripts, and full model outputs are available on the Open Science Framework.

“Preregistration of hypotheses, analyses scripts, and full model outputs are available on the OSF at https://osf.io/ebjm9/ (van der Kleij et al., 2022b).”

• cites van der Kleij et al., 2022bNot matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

Concurrent evidence does not shed light on the direction of the association or potential developmental benefits of reading fiction for understanding others' minds.

“However, this concurrent evidence does not shed light on the direction of this association or the potential developmental bene fi ts of reading fi ction for understanding others ' minds.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Cronbach's α for the British Picture Vocabulary Scale in this sample ranged from .89 to .94 across time points.

“Cronbach ' s α for this sample ranged from .89 to .94 across time points.”

Classified as probably not a substantial claim.

Better theory of mind performance is longitudinally associated with better social outcomes.

“Given that better ToM performance is longitudinally associated with better social outcomes (Devine et al., 2016), it is possible that ToM mediated these effects.”

• cites Devine et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Individual differences in theory of mind in middle childhood are stable and relate to social outcomes, over and above confounding variables such as verbal ability, executive functions, or demographic characteristics.

“Individual differences in ToM in middle childhood are also found to be stable and relate to social outcomes, over and above confounding variables such as verbal ability, executive functions, or demographic characteristics (e.g., Devine et al., 2016).”

• cites Devine et al., 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

The findings extend the existing literature by showing further evidence for rank-order stability in individual differences in ToM, suggesting that variation in theory-of-mind performance is genuine.

“The fi ndings also extend the existing literature by showing further evidence for rank-order stability in individual differences in ToM (e.g., Devine, 2021), suggesting that variation in theory-of-mind performance is genuine.”

• cites Devine, 2021 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Theory of mind may indirectly impact reading behavior, for instance via reading ability.

“Alternatively, ToM may indirectly impact reading behavior, for instance via reading ability.”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Better understanding of characters' intentions or emotions could aid forming a mental model of the text or help draw inferences important for comprehension.

“It is argued that better understanding of character ' s intentions or emotions could aid forming a mental model of the text or help draw inferences that are important for comprehension (Dore et al., 2018; Kim, 2017).”

• cites Dore et al., 2018 sourceNot checked — no full text of the source
• cites Kim, 2017 sourceNot checked — no full text of the source

Classified as probably not a substantial claim.

Short-term training designs do not provide prolonged exposure or test for sustained benefits.

“Existing work has used either cross-sectional designs that cannot provide evidence about causality or short-term training designs that do not provide prolonged exposure or test for sustained bene fi ts.”

Classified as probably not a substantial claim.

Many existing studies do not control for potential confounding variables.

“Furthermore, many existing studies do not control for potential confounding variables.”

Classified as probably not a substantial claim.

The study deviated from the preregistration by using observed variables for the models rather than latent variables, due to the smaller sample size.

“We deviated from the preregistration by using observed variables for the models rather than latent variables, due to the smaller sample size.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

Children with better theory of mind abilities may prefer to read more narrative fiction over other genres or types of reading materials, due to its relatively rich social content, compared with children who struggle more with understanding other people's intentions and emotions.

“An association in this direction could for instance follow from children with better ToM abilities preferring to read more narrative fi ction over other genres or types of reading materials, due to the relatively rich social content, compared with children who struggle more with understanding other people ' s intentions and emotions (e.g., Dore et al., 2018).”

• cites Dore et al., 2018 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based).

No previous studies have examined direct associations between theory of mind and reading behavior in adolescence.

“However, no previous studies have examined direct associations between ToM and reading behavior in adolescence.”

May combine two claims.

Reading boosts language skills needed to express or understand others' intentions or emotions, rather than having a direct link to theory of mind.

“Our fi ndings suggest that rather than a direct link, reading boosts language skills needed to express or understand others ' intentions or emotions.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Parent education, gender, and verbal ability at age 12.5 years were added as control variables.

“Parent education (i.e., low vs. high), gender (i.e., girls vs. boys), and verbal ability at age 12.5 years were added as control variables.”

May combine two claims.

Only a small number of studies provide direct evidence of developmental change by employing longitudinal designs following within-person change in adolescence, and these studies have tracked change over a relatively short time frame.

“However, despite valuable insights from cross-sectional studies, only a small number of studies provide direct evidence of developmental change by employing longitudinal designs following within-person change in adolescence, and these studies have tracked change over a relatively short time frame.”

Probably not a substantial claim: may not pass S4 (based).

Cross-sectional designs cannot provide evidence about causality.

“Existing work has used either cross-sectional designs that cannot provide evidence about causality or short-term training designs that do not provide prolonged exposure or test for sustained bene fi ts.”

May combine two claims.

The majority of studies examining the relation between reading experience and theory of mind have been cross-sectional, showing at least a small concurrent association in adults and children, where people who read more fiction perform better on tasks measuring theory of mind or emotion understanding.

“The majority of the studies examining the relation between reading experience and ToM have also been cross-sectional, showing that there seems to be at least a small concurrent association in adults and children, where people who read more fi ction perform better on tasks measuring ToM or emotion understanding”

• cites Dodell-Feder & Tamir, 2018 sourceNot checked — no full text of the source
• cites Mumper & Gerrig, 2017 sourceNot checked — no full text of the source
• cites Wimmer et al., 2024 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S4 (based).

A growing body of literature has reported associations between reading experience and the ability to reason about others' mental states, called theory of mind.

“A growing body of literature has reported associations between reading experience and the ability to reason about others ' mental states, called ' theory of mind ' (ToM).”

Probably not a substantial claim: may not pass S4 (based).

This study is the first to examine developmental benefits of reading narrative fiction for social understanding in a longitudinal design.

“This study is the fi rst to examine developmental bene fi ts of reading narrative fi ction for our social understanding in a longitudinal design.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing), S4 (based).

Better theory of mind ability may lead to a preference for reading narrative fiction.

“Moreover, these studies also do not examine the possibility that this association could be in the opposite direction: Better ToM ability may lead to a preference for reading narrative fi ction.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Full information maximum likelihood was used to handle missing data.

“Full information maximum likelihood (ML) was used to handle missing data.”

May combine two claims.

The lack of variation in rate of growth is likely because improvements were very small, even though children did not show ceiling effects on the task.

“This is likely because, even though children did not show ceiling effects on this task, improvements were very small.”

Probably not a substantial claim: may not pass S3 (load-bearing).

It is possible that theory of mind mediated the effects of reading on social outcomes.

“Given that better ToM performance is longitudinally associated with better social outcomes (Devine et al., 2016), it is possible that ToM mediated these effects.”

May combine two claims.

The data suggest that reading supports the acquisition of language that helps express more complex thoughts and emotions, and that reading-supported language may act as a confound or mediator of effects of fiction upon ToM.

“However, our data suggest that we should also consider that reading supports the acquisition of language that helps us express more complex thoughts and emotions and that reading-supported language may therefore be acting as a confound or, more interestingly, as a mediator of effects of fi ction upon ToM.”

May combine two claims.

The study design did not allow for examining differences in the type of narrative exposure or the potential mediating role of comprehension ability.

“Second, our study design also did not allow for examining differences in the type of narrative exposure or the potential mediating role of comprehension ability.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The study was not able to fit the model as originally preregistered; due to a smaller sample size than expected, the model was simplified and observed variables were used instead of latent variables.

“However, we were not able to fi t our model as originally preregistered. Due to the smaller sample size than expected, we had to simplify the model and use observed variables instead of latent variables.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Previous research has shown associations between fiction reading and measures of empathy and mentalizing abilities, but not with moral cognition, suggesting that different aspects of social abilities could be differentially associated with exposure to narrative fiction.

“Previous research on this topic has for instance shown associations between fi ction reading and measures of empathy and mentalizing abilities, but not with moral cognition, suggesting that different aspects of our social abilities could be differentially associated with exposure to narrative fi ction (Wimmer et al., 2024).”

• cites Wimmer et al., 2024 sourceNot checked — no full text of the source