Studie Nikals.pdf

ORIGINAL RESEARCH publi...

ORIGINAL RESEARCH published: 21 July 2020 doi: 10.3389/fpsyg.2020.01628 The Home Literacy Environment as a Mediator Between Parental Attitudes Toward Shared Reading and Children’s Linguistic Competencies Frank Niklas 1* , Astrid Wirth 1 , Sabrina Guffler 2 , Nadja Drescher 2 and Simone C. Ehmig 3 1 Department of Psychology, Ludwig-Maximilians-University, Munich, Germany, 2 Department of Educational Psychology, Julius-Maximilians-University, Würzburg, Germany, 3 German Reading Foundation, Mainz, Germany The home learning environment plays an important role for children’s early competencies development. In particular, the early home literacy environment (HLE) that consists of all literacy resources and interactions in a family that support children’s linguistic and literacy learning is closely associated with children’s language comprehension and production. A key aspect of the HLE is shared reading that should start early in children’s life and should be part of a regular routine in the family. However, parental attitudes toward Edited by: (shared) reading have hardly been analyzed. Ying Guo, In this longitudinal study, we analyzed the associations between parental attitudes University of Cincinnati, United States toward shared reading and children’s linguistic competencies and whether these Reviewed by: Christian Tarchi, associations may be mediated by the HLE. Further, we were interested in changes University of Florence, Italy of parental attitudes over time and their association with child and family background Kelly Farquharson, Florida State University, United States characteristics. The sample consisted of N = 133 children with an average age of *Correspondence: about 3 years at t1. Children were tested two more times with a 6-month period in- Frank Niklas between each assessment. Parental attitudes toward shared reading, socioeconomic [email protected] status (SES), and the HLE were assessed via parental survey. Children’s sentence Specialty section: comprehension, productive language, and grammar were measured with a standardized This article was submitted to test battery. Children whose parents had a more positive attitude toward shared reading Educational Psychology, not only lived in a greater quality HLE but also performed better in the linguistic tests. a section of the journal Frontiers in Psychology In a structural equation model, an indirect effect was found showing that the HLE Received: 03 March 2020 mediated the effect of parental attitudes on children’s linguistic competencies. Further, Accepted: 16 June 2020 parental attitudes toward shared reading did not change significantly across t1 to t3, Published: 21 July 2020 and a lower score in the SES scale was associated with a less positive attitude toward Citation: Niklas F, Wirth A, Guffler S, shared reading. Consequently, parental attitudes toward shared reading seem to be an Drescher N and Ehmig SC (2020) The important basis for individual differences in the quality of the HLE and also for children’s Home Literacy Environment as linguistic competencies. As these attitudes vary in the context of different family SES a Mediator Between Parental Attitudes Toward Shared Reading backgrounds, they may be a good target for interventions to support the quality of the and Children’s Linguistic HLE and young children’s linguistic learning. Competencies. Front. Psychol. 11:1628. Keywords: home literacy environment, parental attitude toward reading, linguistic competencies, kindergarten doi: 10.3389/fpsyg.2020.01628 children, development of early child competencies Frontiers in Psychology | www.frontiersin.org 1 July 2020 | Volume 11 | Article 1628 Niklas et al. Parental Attitudes, HLE and Linguistic Competencies INTRODUCTION production are important predictors of academic performance in school (e.g., Joshi, 2005; Juel, 2006; Claessens et al., 2009). Attitudes are of great interest for psychologists and educators as Consequently, precursors of literacy abilities and children’s later attitudes influence our perception and may have an impact on literacy competencies lie on a continuum (e.g., Torppa et al., our behavior (cf. Eysenck, 2004; Schwarz, 2007). In the family 2007; Lepola et al., 2016). Further, individual differences in context, parental attitudes play a major role for young children vocabulary and language comprehension skills in early years as parents are very attractive role models for their children predict not only later reading abilities but also motivational and (cf. Niklas, 2015). Further, parents create the environment their behavioral outcomes in children (Laitinen et al., 2017). children experience, and thus parental attitudes are most likely to influence the home learning environment and children’s The Home Literacy Environment and learning within this context (e.g., Bingham, 2007; Park, 2008; Early Linguistic Competencies Skibbe et al., 2008). Children develop early linguistic competencies during the Shared reading with children is a key aspect of the home interaction with their parents (Vygotsky, 1978). Consequently, literacy environment (HLE) that supports children’s development the HLE provides numerous opportunities for teaching and of linguistic and literacy competencies (Niklas et al., 2016b). learning activities that support the development of children’s However, although shared reading is deemed important by most linguistic and literacy abilities (Niklas and Schneider, 2017a). parents in Germany, some children, and in particular, children The HLE is a multifaceted construct comprising current parental from families with a low socioeconomic status (SES), are rarely reading habits, shared reading habits in the family, and more read to (German Reading Foundation, 2010). As maternal literacy general aspects of family literacy such as the frequency of library beliefs are closely associated with the HLE and child outcomes visits and the number of books in a household. These aspects (Weigel et al., 2006), such attitudes may be a good target can be further differentiated into a cultural capital and a cultural for interventions. praxis (e.g., Niklas et al., 2013). Whereas in the context of the HLE In this study, we analyze the associations of parental attitudes cultural capital refers to the number of books and children’s books toward shared reading, the quality of the HLE, and young in a household, cultural praxis consists of all literacy activities children’s linguistic outcomes in a longitudinal design. Further, in the family such as shared reading. Both aspects are closely we were interested in whether parental attitudes change across 1 associated; however, they may still differ in the role they play year and whether these attitudes were associated with child and for the development of children’s linguistic competencies (e.g., family characteristics. McElvany et al., 2009). The association between the HLE and children’s linguistic The Development of Children’s Early and literacy competencies is also evident in intervention studies that try to enhance the quality of the HLE to support Linguistic Competencies children’s competency development. Indeed, various family An important early linguistic ability is the ability to understand literacy programs have demonstrated small to large effects (e.g., spoken language, often referred to as language comprehension Harper et al., 2011; Lever and Sénéchal, 2011). For instance, skills. Language comprehension skills consist of basic abilities Niklas and Schneider (2015, 2017b) showed that even non- such as the activation of word meanings and understanding intensive interventions that just comprised one parent evening sentences, of receptive vocabulary, the knowledge of text and and one individual session may change the HLE and subsequently sentence structures, and language production skills such as impact on children’s development of their vocabulary and children’s expressive vocabulary (Lepola et al., 2016; Niklas et al., phonological awareness. 2016a). Both receptive and expressive language skills are closely The observation that the HLE and subsequent child related (e.g., Cutting and Dunn, 1999). Further, these abilities competencies can be improved by interventions has been also are highly stable competencies from kindergarten age onward confirmed in comprehensive meta-analyses. Sénéchal and Young (Whitehurst and Lonigan, 2001). (2008) and Mol et al. (2008) each analyzed 16 intervention studies In their concept of emergent literacy, Whitehurst and Lonigan that focused either on parental involvement in kindergarten and (1998) differentiate between such language competencies as primary school children’s development of reading and spelling outside-in skills and inside-out skills such as phonological abilities or on dialogic reading (for more information on dialogic awareness and letter knowledge. Indeed, inside-out skills are reading, see Cohrssen et al., 2016) and its effect on children’s also important predictors of later reading and writing abilities; vocabulary. Mean effect sizes of Cohen’s d = 0.65 and 0.42, and however, these skills develop at a later age and are not formally thus small to medium effects were found. Consequently, the HLE taught in German kindergartens and preschools (Niklas and is a very important factor in children’s development of linguistic Schneider, 2015). As our analytic sample consists of 3- to 4-year- and literacy competencies. old children, we only focus on outside-in skills. Early linguistic and literacy competencies are essential for a successful school career, and precursors of these abilities Parental Attitudes Toward Shared develop long before children enter school. An early assessment Reading of these skills is preferable, as specific precursors of later Some studies explicitly regard parental attitudes toward literacy literacy competencies such as language comprehension and as an aspect of a broader construct of the HLE (e.g., Frontiers in Psychology | www.frontiersin.org 2 July 2020 | Volume 11 | Article 1628 Niklas et al. Parental Attitudes, HLE and Linguistic Competencies Niklas and Schneider, 2017b) or differentiate between the HLE with child and family characteristics as many studies in and these attitudes as separate variables (e.g., Park, 2008), whereas this context only used cross-sectional data (e.g., Park, 2008; in other studies on the HLE, attitudes are not taken into Hemmerechts et al., 2017). account (e.g., Niklas and Schneider, 2013). As children learn by We analyzed the development of child competencies across interacting with and observing more knowledgeable others, in the a 1-year period and assessed parental attitudes toward shared early years often their parents (Vygotsky, 1978), they also take reading, the HLE, and linguistic outcomes. Here, we expected the notice of parental attitudes displayed during these interactions parental attitudes toward reading and the quality of the HLE to and observations. Parents act as important role models for be stable across the 1-year period (cf. Niklas, 2015). Further, we their young children (Bandura, 1977), and their attitudes are expected that a more positive attitude toward shared reading and very likely to impact on children’s own attitudes and interests. a greater quality in the HLE should be associated with greater Consequently, it is to be expected that parental attitudes toward linguistic competencies in children. Finally, we assumed that shared reading shape children’s interest in literacy and books the HLE should act as a mediator between parental attitudes and in turn may also impact on children’s linguistic and literacy and child outcomes. competencies (Bingham, 2007; Skibbe et al., 2008). Therefore, parental attitudes toward reading and literacy in general, and in families with young children, the attitudes toward shared reading, MATERIALS AND METHODS specifically, may be important for children’s development (Weigel et al., 2006; cf. Niklas, 2015). Sample According to the model of Zanna and Rempel (1988), objects In total, N = 133 children were assessed using a longitudinal are evaluated according to three different components: (1) research design with three measurement points (t1–t3) across cognitive, (2) affective, and (3) behavioral. In regard to shared 12 months (6 months in-between each measurement). Power reading, this model implies that parents will put a certain value analysis with G∗ Power (Faul et al., 2007) indicated a sample on shared reading, will feel more or less positive about it, size above N = 129 participants to be sufficient for the planned and finally, initiate shared reading session more or less often analyses. At t1, children were between 26 and 45 months old with their children and in a way that triggers more or less (M = 36.6, SD = 4.1). In the sample, gender was almost equally reading motivation. The attitude toward shared reading develops distributed, with 46% girls (N = 61). More than a third of the over time, may change from situation to situation, and will be children (37.6%, N = 50) had a migration background with at closer associated with actual behavior when specific and concrete least one parent being born outside of Germany. attitudes are assessed (cf. Schwarz and Bohner, 2001). However, All participating parents were asked about their occupation given that attitudes also comprise a behavioral component, it is and their partner’s occupation to assign prestige values to these likely that parental attitudes toward shared reading will be closely occupations (Wegener, 1988; cf. Christoph, 2005). Here, values associated with the HLE, in particular with the cultural practice ranged from 20 (an unskilled laborer) to 186.8 (a physician), and (cf. Niklas et al., 2013). Actually, in a study by Tambyraja et al. for the analyses, the highest prestige score in the household was (2017), caregivers’ own reading habits were a predictor of the used. Information about the SES could be obtained from N = 122 general HLE in the family. families with a mean of M = 86.86 (SD = 40.53), a value assigned The development of parental attitudes toward shared reading to the occupation of a salesman. depends on various experiences the parents had encountered such as their own shared reading experiences as children and Procedure in general their socialization (cf. Eysenck, 2004). Consequently, Formal consent to conduct the study was obtained from it is to be expected that the attitude toward shared reading the center coordinators and parents, and ethics approval was should be associated with the socioeconomic status (SES) of obtained from the University of Würzburg, Germany. Randomly the family (e.g., Park, 2008; Skibbe et al., 2008; Becker and selected kindergartens in two German states were contacted McElvany, 2018), similar to the association of the SES with the and invited to participate in our study. In Germany, most HLE (Aikens and Barbarin, 2008; Niklas et al., 2013). When children are enrolled in kindergarten from 2 to 3 years trying to tackle different linguistic and literacy outcomes of of age until the beginning of formal schooling at the age children from different family backgrounds, parental attitudes of 6. Kindergarten refers to a nursery school or preschool might be a worthwhile target. setting, with a focus on playing and practical activities (see further Niklas et al., 2018). N = 21 kindergartens agreed to Research Focus participate and handed out information and consent forms The association between the HLE and children’s early and later for all parents with children in the age group between 26 linguistic and literacy outcomes is well established (e.g., Sénéchal and 45 months. In each participating kindergarten, between and LeFevre, 2002; Niklas and Schneider, 2013; Hemmerechts N = 4 and N = 13 children (and their parents) participated et al., 2017). However, less is known about the role parental in our study. At each measurement point, trained psychologists attitudes toward shared reading play in this association, in assessed children’s competencies in their kindergartens, whereas particular for younger children (for an example, see Bingham, parents were asked to fill in surveys. Parental response rates 2007). Further, it is still not clear whether we see changes in lay between 84.1% and 75.8% for each measurement point these attitudes across time and whether they are associated (between N = 21 and N = 32 parents did not return the Frontiers in Psychology | www.frontiersin.org 3 July 2020 | Volume 11 | Article 1628 Niklas et al. Parental Attitudes, HLE and Linguistic Competencies survey). Between N = 12 and N = 21 of the study children Linguistic Abilities could not be tested at least once due to absence or refusal to Children’s level of linguistic abilities was assessed with the participate. We address the handling of missing data in our standardized German language development test instrument analytic approach. SETK 3-5 (Grimm et al., 2010) that comprises subscales for language comprehension and language production. Reliability Surveys and Test Instruments (Cronbach’s α) was at least α = 0.70 for each subscale (Neugebauer At each measurement point, parents were asked about their and Becker-Mrotzek, 2013). Each subtest started with a sample family’s HLE and their attitude toward shared reading. item to demonstrate how to approach the question and to Further, they were asked to provide information about their provide feedback for the child. During the test phase, no further family background. feedback was given. At t1 and t2, the language comprehension scale consisted of Home Literacy Environment three subtests. In the first one, children were asked to select a The HLE survey was an adapted version of a survey used picture out of four similar pictures, matching the sentence that by Niklas et al. (2016a). This survey contained seven items had been read out to them (nine items, maximum attainable score covering different facets of the HLE: the number of books at of 9). In the following two subtests, children were asked to act home, the number of children’s books at home, the frequency according to short instructions (for example, “Put the red buttons of reading to the child and the frequency of both parents’ own on the box”). Both subtests consisted of five items each, with a reading, the frequency of the child looking at picture books, maximum attainable score of 5, respectively. At t3, subtest 1 was and the frequency of library visits with the child. Each item omitted due to children’s age and according to the test manual had a range from 0 to 4. Both items concerning the number of and instead another five items were added in which children (children’s) books at home were used to estimate the cultural were asked to act according to more complex instructions (for literacy capital in a family (example: “How many children’s example, “The yellow ball, that is pushed by the white ball, falls books does your child own?”), whereas the remaining five items from the table”). Consequently, the attainable maximum score assessed a family’s cultural literacy praxis (example: “How often was lower at t3 compared to t1 and t2. do you read to your child?”; cf. Niklas et al., 2013). The cultural Language production consisted of two subtests for t1 and t2, capital scale had a maximum attainable sum score of 8, with assessing the encoding of semantic relations and morphological Cronbach’s α = 0.78 at t1, 0.86 at t2, and 0.84 at t3. The rule-making. Both subtests were z-transformed and summed cultural praxis scale had a maximum attainable sum score of up into the language production scale at t1 and t2. At t3, 20, with Cronbach’s α = 0.67 at t1, 0.60 at t2, and 0.65 at t3. when all participating children were older than 3 years old, The sum score of the global HLE scale was a reliable measure language production was assessed with a more comprehensive (Cronbach’s α = 0.78 at t1, 0.74 at t2, and 0.71 at t3) with morphological rule-making test. To be consistent, this test was a maximum attainable score of 28. Retest reliability for the also z-transformed. global HLE scale was high, with r12 = 0.80, r13 = 0.71, and In the subtest “encoding of semantic relations,” children were r23 = 0.79. asked to describe 11 pictures to assess their use of prepositions (for example, “The children walk across the street.”). There is Parental Attitudes no maximum attainable score as children were free to describe In the parent survey, 11 items assessed attitudes, behaviors, and pictures with an unlimited number of words, which were family situations in the context of shared reading. We conducted counted for each child individually. In the subtest “morphological an exploratory factor analysis in SPSS to identify common factors. rule-making,” children were asked to say plural forms of Only the first factor explained a major part of the variance, and different nouns (for example, “car–cars”). Here, the maximum four items on cognitive attitudes loaded on this factor. All other attainable score was 20 at t1 and t2 and 36 at t3 due to eight items either did not load on a specific factor or were the only additional test items. items to load on an additional factor. Consequently, we assessed We created an index score combining both z-transformed parental attitudes toward shared reading with four items on 5- language comprehension and language production scales into a point Likert scales. The items ranged from 0 (I do not agree) general linguistic abilities scale. Retest reliability for the general to 4 (I agree completely). Here, all items focused on cognitive linguistic abilities was very high, with r12 = 0.85, r13 = 0.75, and attitudes toward shared reading and assessed the value attached r23 = 0.88. to reading at home, perceived interest in reading by the child, and parental motivation toward reading and shared reading (item Non-verbal Intelligence example: “Reading is regarded as an important activity at our In addition to children’s age, sex, and their family’s SES, all home”). The parental attitudes toward shared reading subscale analyses were controlled for children’s non-verbal intelligence. had a maximum attainable sum score of 16, with Cronbach’s Children’s non-verbal intelligence was assessed at t1 and t3 α = 0.90 at t1, 0.90 at t2, and 0.86 at t3. Parental attitudes toward with the Columbia Mental Maturity Scale (CMM; Burgemeister shared reading were fairly stable across t1 to t3, with r12 = 0.60, et al., 1972), assessing 3- to 5-year-old children’s capability r13 = 0.73, and r23 = 0.71. The four items assessing cognitive for abstraction and logical reasoning. Here, children had to attitudes toward shared reading had been used in previous studies identify the odd picture in an array of three to five pictures (e.g., Park, 2008; Niklas et al., 2016a; Wirth et al., 2019). (e.g., four identical dogs and one cat), and a maximum Frontiers in Psychology | www.frontiersin.org 4 July 2020 | Volume 11 | Article 1628 Niklas et al. Parental Attitudes, HLE and Linguistic Competencies of 57 points was attainable. Split-half reliability in German size correlations of the HLE with parental attitudes toward contexts ranges from 0.92 to 0.96, and the CMM proved to shared reading (r = 0.49–0.65). Further, linguistic abilities be a good indicator of children’s general cognitive abilities in were also significantly correlated with parental attitudes toward recent German studies (Esser, 2002; Hasselhorn et al., 2012; shared reading at all three measurement points (r = 0.37– Niklas and Schneider, 2017a). 0.39). Children’s level of linguistic skills seemed to be partly All descriptive data and the sample sizes for all variables are dependent on other influencing variables, such as age, gender, shown in Table 1. intelligence, and the family’s socioeconomic background (mainly small to medium effect size correlations). Whereas the HLE and Analytic Approach parental attitudes toward reading to their children were strongly Data analyses were conducted using SPSS 24 (Ibm Corp, 2016) associated with families’ SES (r = 0.40–0.58), the correlations for descriptive and correlative analyses and Mplus 7 (Muthén with the other control variables were much smaller and mostly and Muthén, 2012) for structural equation modeling (SEM). As not significant. some parental surveys were not or only partially completed, and some children’s test scores were missing, several cases were Change in Parental Attitudes Toward incomplete. After analyzing the missing data for patterns, we estimated missing data using the full information maximum Shared Reading and the Quality of the likelihood option (MLR estimator) in Mplus. Home Literacy Across t1 to t3 First, results of bivariate correlational analyses (Pearson’s In order to investigate whether the necessary conditions r) of all study variables for the three measurement points for rm ANOVA had been met, Mauchly’s test of sphericity are presented. Here, we analyzed whether parental attitudes was conducted. The results indicated that the assumption and HLE were associated with the control variables and the of sphericity had been violated for parental attitudes, with linguistic outcomes. In a second step, we carried out univariate χ2 (2) = 11.67, p < 0.05. Therefore, a Huynh–Feldt correction variance analyses with repeated measurement (rm ANOVA) to was applied. For the HLE, Mauchly’s test of sphericity was test whether parental attitudes toward shared reading and the not violated, with χ2 (2) = 2.02, p < 0.05. We did not find quality of the HLE varied across t1 to t3. significant effects of time on parental attitudes and the HLE, with Finally, we used SEM to analyze the association between F(2,130) = 2.11, p = 0.13, η2 = 0.03 and F(2,138) = 0.01, p = 0.99, parental attitudes toward shared reading and children’s linguistic η2 = 0.00. Consequently, no significant and meaningful change competencies and whether this association may be mediated by in the quality of the HLE and parental attitudes toward shared the HLE. We examined this association controlling for various reading was observed between t1 and t3. child and family characteristics. The Association of Parental Attitudes RESULTS Toward Shared Reading, the Home Literacy Environment, and Children’s Correlational Analyses Linguistic Competencies Table 2 shows the cross-sectional results for the correlational We used SEM to address the main research question concerning analyses at t1, t2, and t3. the associations between parental attitudes toward shared As expected, significant medium effect size correlations were reading and children’s linguistic competencies and whether this observed between the HLE with its subscales cultural praxis association may be mediated by the HLE. In a first model, we and cultural capital and children’s linguistic abilities at all three tested the direct prediction of linguistic competencies by parental measurement points (r = 0.37–0.55) as well as large effect attitudes toward shared reading (Figure 1). Latent variables were TABLE 1 | Descriptive statistics for the study variables at t1, t2, and t3 (sample sizes, means, standard deviations, observed ranges). Variables N M SD Observed range Intelligence 121/-/112 27.8/-/38.2 13.4/-/10.7 0.0–50.0/–/0.0–54.0 SES1 122 86.89 40.84 20.00–186.80 HLE 108/100/94 19.2/19.5/19.4 4.5/4.2/4.2 5.0–26.0/6.0–26.0/7.0–26.0 HLE–Cultural capital 111/102/96 7.02/7.15/7.29 1.56/1.45/1.37 2.0–8.0/2.0–8.0/2.0–8.0 HLE–Cultural praxis 110/103/103 12.07/12.41/12.41 3.47/3.18/3.18 3.0–18.0/3.0–18.0/3.0–18.0 Parental attitudes 109/101/94 12.43/12.89/12.84 3.55/3.19/2.73 0.0–16.0/3.0–16.0/5.0–16.0 Linguistic abilities2 114/112/101 0.53/0.01/0.78 3.19/3.19/3.78 -5.5 to 7.8/-7.6 to 5.5/-8.9 to 7.3 Language comprehension3 118/116/112 7.94/10.99/8.00 4.67/5.06/4.28 0.0–19.0/0.0–18.0/0.0–15.0 Language production4 115/113/112 -0.01/0.01/0.00 1.77/1.85/1.00 -2.9 to 4.1/-3.7 to 2.9/-2.0–1.9 1 SES, highest family occupational prestige; 2 combined index of the z-transformed language comprehension, and language production scale of the SETK; 3 the subtests changed between t2 and t3; 4 z-transformed subscale with two subscales at t1 and t2 and one subtest at t3. HLE, home literacy environment; SES, socioeconomic status. Frontiers in Psychology | www.frontiersin.org 5 July 2020 | Volume 11 | Article 1628 Niklas et al. Parental Attitudes, HLE and Linguistic Competencies -0.09/-0.11/-0.10 -0.12/-0.11/-0.12 -0.05/-0.09/-0.09 -0.11/-0.06/-0.08 -0.25/-0.24/-0.17 -0.20/-0.18/-0.16 -0.28/-0.21/-0.15 Pearson’s r correlation coefficients; p < 0.05 in bold characters, N = 78–133. HLE, home literacy environment; SES, socioeconomic status. 1 Children’s non-verbal intelligence (CMM) was assessed at t1 and t3; modeled for parental attitudes and children’s level of linguistic 0.37/0.29/0.12 0.47/0.38/0.44 0.40/0.44/0.55 0.45/0.46/0.52 0.31/0.42/0.54 0.31/0.40/0.48 0.92/0.91/0.82 0.75/0.79/0.74 0.36/-/0.47 abilities. Here, parental attitudes were modeled using the four items for cognitive attitudes toward reading, whereas the latent 11 – linguistic abilities variable included the language comprehension and language production scales of the SETK. In addition, we controlled for children’s age, sex, intelligence, and family’s SES. 0.48/0.50/0.49 0.41/0.48/0.55 0.43/0.41/0.42 0.33/0.43/0.53 0.38/0.41/0.44 0.89/0.88/0.79 0.33/0.28/0.18 0.37/-/0.46 Parental attitudes toward shared reading at t1 were a significant predictor of children’s linguistic abilities at t3. The proposed 10 – – model fit the data well, with χ2 (23) = 37.92, p > 0.05, root mean square error of approximation (RMSEA) = 0.07, comparative fit index (CFI) = 0.97, Tucker-Lewis index (TLI) = 0.94, 0.46/0.43/0.41 0.44/0.42/0.40 0.33/0.41/0.47 0.37/0.39/0.39 0.40/0.29/0.22 0.37/0.46/0.49 0.45/-/0.24 standardized root mean residual (SRMR) = 0.05. = 0, male = 1; 3 highest family occupational prestige; 4 combined index of the z-transformed language comprehension, and language production scales of the SETK. – In the final model (Figure 2), the HLE at t2 was added as – – 9 a latent variable. The HLE comprised the cultural capital and cultural praxis subscales, and again we controlled for children’s age, sex, intelligence, and family’s SES. When the HLE was 0.22/-/0.19 -0.14/-0.13/-0.08 -0.08/-0.06/-0.04 -0.09/-0.10/-0.09 -0.02/-0.13/-0.21 -0.13/-0.08/-0.13 0.01/-0.05/-0.19 0.50/0.40/0.54 0.69/0.59/0.57 0.61/0.62/0.49 0.65/0.49/0.52 0.07/-/0.23 added as a mediator to the model, parental attitudes toward shared reading were no longer direct significant predictors of – – – – 8 children’s linguistic abilities. Instead, parental attitudes toward shared reading predicted the HLE, which, in turn, predicted children’s linguistic abilities. Consequently, in this full mediation, 0.55/0.44/0.44 0.96/0.96/0.75 0.60/0.57/0.38 0.02/-/0.14 parents’ attitudes toward shared reading at t1 were significant predictors of children’s linguistic abilities at t3 only indirectly via – – – – – 7 the HLE at t2 with a total standardized indirect effect of 0.28 (p < 0.05). In addition, children’s age was predictive of their linguistic 0.80/0.78/0.64 0.53/0.48/0.41 abilities with a standardized beta coefficient of 0.26 (p < 0.01), but 0.05/-/0.21 neither children’s sex, intelligence, nor the family’s SES. However, – – – – – – 6 the family’s SES significantly predicted children’s HLE with a standardized beta coefficient of 0.37, p < 0.01, and parent’s attitudes toward reading with a standardized beta coefficient of 0.58/0.51/0.53 0.57, p < 0.001. No other control variables were significantly 0.03/-/0.26 associated with the HLE and parent’s attitudes. The proposed – – – – – – – 5 model fit t