Early Literacy Research Library (ELRL) - Article

Predictive Brain Signals Mediate Association Between Shared Reading and Expressive Vocabulary in Infants

Wang, S., Tzeng, O.J.L., Aslin, R.N. (2022) Predictive Brain Signals Mediate Association Between Shared Reading and Expressive Vocabulary in Infants. PLoS One, 17(8), e0272438.,

Access: FREE/Open Access

Publication year


study description


core topic(s)

Shared Reading

Population Characteristics

Infant/Newborn , International

Exposures, Outcomes, Other

Brain/Neurocognitive , Child Behaviors and Skills , Child Development (general) , Language and Literacy Development , Parent-Child Relationships/Interactions


The present study explored how shared reading experience influenced predictive brain signals and expressive vocabulary of 12-month-old infants.


Shared reading

outcomes evaluated

Brain signals and expressive vocabulary


Participants were recruited through online advertisement in 2020 and the database of interested participants for the NTNU-Haskins Joint Laboratory of Brain Development and Learning.


To achieve these aims, we tested 12-month-old infants’ predictive brain signals using functional near-infrared spectroscopy (fNIRS), their parent-child shared reading experience as assessed by the StimQ-READ subscale-Infant [31, 32] translated into Chinese, and their expressive vocabulary as assessed by the Infant form of the Mandarin-Chinese Communicative Development Inventory (Taiwan) (hereafter MCDI-T) [33]. We hypothesized that the parent-child shared reading experience would be positively correlated with the strength of the predictive brain signal and that the predictive brain signal would serve as a mediator that explains the underlying mechanism of the relation between parent-child shared reading experience and infants’ expressive vocabulary development.

sample size

n=23 (dyads)


Measure of Expressive Vocabulary: Mandarin-Chinese Communicative Development Inventory (Taiwan) (hereafter MCDI-T)


Measure of Shared Reading Experience: Infant version of StimQ-READ subscale


Measure of Predictive Brain Signals: functional near infrared spectroscopy (fNIRS) task


The amount of shared reading experience was correlated with the strength of this predictive brain signal and with infants' expressive vocabulary. Importantly, the predictive brain signal explained unique variance of expressive vocabulary beyond shared reading experience and maternal education. A further mediation analysis showed that the effect of shared reading experience on expressive vocabulary was explained by the infants' predictive brain signal.


This is the first evidence indicating that richer shared reading experience strengthens predictive signals in the infant brain and in turn facilitates expressive vocabulary acquisition.


A substantial proportion of collected data were excluded from final analyses due to a range of reasons such as infants’ shorter attention span [54]. In the current study, the attrition rate is 43.9% (18 out of 41 infants’ collected data was excluded due to excessive movement or failure to watch a sufficient number of trials). Although this rate compares well with the average attrition rate of 34.2% reported in a recent meta-analysis with 272 experiments across 182 publications in infants aged under 24 months [54], future studies should attempt to modify the study design to reduce the attrition rate and increase generalizability. We would also note that the final sample consisted of relatively high SES families which may limit the generalizability of the current findings. Future studies can consider to include samples from more diverse SES backgrounds to see whether the current results can be replicated in samples from different SES backgrounds. StimQ-READ Limitations: One is that its reliance on parent report made it subjective to bias due to memory inaccuracy or social desirability (i.e., the tendency to report in a generally favor fashion). The other is that it did not measure specific interaction strategies that parents used when reading to their 12-month-old infants. As a consequence, it is not possible to make a strong inference between interactive strategies used by parents and how they are linked to the formation of predictive brain signal.