Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Letters from home: Parent-child correspondence in reading achievement
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
Deborah Schneider1, Florence Bouhali2, Zhichao Xia1, Jillian O'Malley1, Chanyuan Gu1, Jie Luo1, Fumiko Hoeft1; 1University of Connecticut, 2Aix-Marseille University, CNRS, CRPN
Background/Goals: The intergenerational transmission (IGT) of linguistic and cognitive skills is influenced by diverse genetic and environmental factors and their interactions. Neurobiological and cognitive research has identified sex-dependent and sex-independent patterns of parent-child transmission. Drawing on data from the NIH-supported Intergenerational Neuroimaging of Language and Reading Networks project, we sought to examine whether parent-child associations in reading vary by component skill; whether associations differ by parent and child sex; and whether shared reading (SR) mediates or moderates these relationships. Methods: Our sample comprised 76 parent-child triads (mother, father, child[ren] aged 8-12). Phonetic Decoding Efficiency (PDE; pseudoword reading) and Sight Word Efficiency (SWE; word reading) were assessed using the Test of Word Reading Efficiency, Second Edition (TOWRE-2). SR was indexed by frequency of shared reading. We used path analysis, linear mixed models (LMMs), and permutation testing to provide converging evidence for parent-child resemblance and the role of SR. Results: Path analyses revealed robust direct effects, particularly for PDE. Fathers' scores predicted children's PDE in the both-parents (β = 0.40, pFDR < .001) and father-daughter models (β = 0.63, pFDR < .001); mothers' scores showed moderate effects in the both-parents (β = 0.21, pFDR = .004) and mother-son models (β = 0.37, pFDR = .001). SWE showed weaker effects, with only the father-daughter model achieving bootstrap significance (β = 0.42, 95% CI [0.06, 0.78]), though p-value significance did not survive FDR correction (pFDR = .12). A significant SR-to-child path emerged in the mother-son PDE model (β = -4.94, pFDR = .02); other mediation paths were non-significant. Given the near absence of mediation effects in our path models, we used LMMs to test SR as a potential moderator. For PDE, parent-child resemblance was strongest in cross-sex dyads, particularly father-daughter (β = 0.76, pFDR < .001) and mother-son (β = 0.69, pFDR = .01) pairs, with non-significant effects in same-sex dyads (father-son: pFDR = .36; mother-daughter: pFDR = .40). SR did not moderate PDE transmission in any model. For SWE, parent-child effects were weaker, with only the father-daughter association achieving significance (β = 0.40, pFDR = .05). SR significantly moderated paternal transmission, with SR × father interactions in the both-parents model (β = 0.43, pFDR = .04) and father-only model (β = 0.56, pFDR = .04), indicating stronger resemblance in families with higher SR. Permutation tests corroborated key findings. For PDE, correlations were significantly larger in related than unrelated pairs, and the strongest correlations were observed in cross-sex dyads (father-daughter: r = 0.58, p < .001; mother-son: r = 0.38, p = .01); same-sex dyads yielded non-significant correlations. For SWE, only the father-daughter correlation approached significance (r = 0.32, p = .05). Discussion: Our findings provide preliminary support for potential domain- and sex-specific transmission pathways in reading achievement, with PDE exhibiting stronger familial resemblance, particularly in cross-sex pairings. SR moderated but did not mediate parent-child similarity, suggesting a limited role as an intergenerational transmission pathway in our sample. Results are presented as observational, rather than mechanistic, and require replication in larger, more diverse samples.
Topic Areas: Reading, Genetics