The literature on parental involvement in early language and literacy pinpoints shared book reading as a potential tool for nurturing children’s linguistic and cognitive development (Inoue et al., 2018; Torppa et al., 2022). This practice, supported by families’ broader involvement in education, has been instrumental in enhancing children’s vocabulary, syntax, and code-related knowledge (Hindman et al., 2008; Wasik et al., 2016). Specifically, the home literacy environment is enriched by several facets of book-reading-related activities: the number of books (Niklas et al., 2013; Riser et al., 2019; Villiger, 2019), time spent reading (Kuo et al., 2004), and the quality of interactions surrounding reading activities (Meng, 2021). In turn, the home learning environment plays a significant role in child literacy. For instance, both classic studies (Scarborough & Dobrich, 1994) and more recent work (Wasik et al., 2016) find that the frequency of book reading explains 8-10% of the variability in children’s vocabulary skills.

While the value of book reading in families is widely acknowledged, how it is practiced shows considerable diversity, particularly across different socioeconomic and ethnic groups. Research indicates that socioeconomic status (SES), and particularly the level of parental education, are pivotal, demonstrating high correlations with the regularity and richness of reading activities at home (Chen et al., 2018). Typically, research has found that families with higher income and educational levels (compared with less economically advantaged families, a disproportionate number of whom hail from ethnically and linguistically minoritized backgrounds) engage children in more literacy activities at home, which in turn enhances reading outcomes for their children (Hemmerechts et al., 2016; Neumann, 2016; Wirth et al., 2021). However, it is crucial to delve deeper into these dynamics. For instance, despite often facing economic constraints, there is evidence that many low-SES parents regularly read with their children and engage in other ways as well (Corcoran & Steinley, 2017; Phillips & Lonigan, 2009). Moreover, challenging deficit perspectives, recent findings reveal the extensive, proactive participation of families of color or families from marginalized communities in nurturing literacy (Brinkley et al., 2022; Burris et al., 2019; Davis et al., 2015), prompting a re-evaluation of longstanding assumptions.

Buffering against a deficit model, we incorporate an understudied construct – parental beliefs – into analyses exploring parental involvement. We aim to highlight, particularly among families with fewer economic resources and families of color, the diversity of their literacy-related beliefs, the links between beliefs and home reading practices, and the unique associations of both beliefs and practices with children’s literacy development in the essential kindergarten year. By leveraging the Early Childhood Longitudinal Study-Kindergarten 2011 (ECLS-K), a large and nationally representative dataset, we seek to contribute to the literature by providing a more detailed portrait of the nature and importance of parents’ literacy beliefs and the unique role that beliefs play in the home literacy involvement during the kindergarten year among families of diverse backgrounds.

Predictors of Home Literacy Environment

Given the importance of parental involvement, researchers ask why parents become involved in their children’s education, particularly in early childhood. Much of the literature to date has focused on a subset of critical factors as predictors of this engagement.

Demographic Factors

Research consistently shows that various parental demographics – such as SES, level of education, and family size – are significantly related to parents’ home reading involvement. First, a common narrative observes that families with more resources, higher levels of parental education, and smaller family sizes tend to participate more frequently in reading activities with their children (Price & Kalil, 2018; Romeo et al., 2022). However, in striking contrast to expectations, Corcoran and Steinley (2017) report that 71% of parents living under the federal poverty threshold read with their children at least three times per week. Similarly, Hindman and colleagues (2012) found that, among Head Start families, many frequently read books with their young children and also engaged in other important activities (e.g., talking, playing). This finding hints at the possible diversity in home literacy practices among lower SES families that may translate to children’s emergent literacy outcomes.

Second, the relationship between ethnicity and parental involvement in home reading for young children is complex, with research presenting varied findings. Early studies indicated lower frequencies of reading among African American and Latine families, attributed to a range of barriers (Sosa, 1996; Raikes et al., 2006). Similarly, more recent research by Sonnenschein and Sun (2016) suggests that home literacy activities are less common among African American or Latine families compared to their White and Asian counterparts, with White families also partaking in more enrichment activities than other groups. However, other studies recognize considerable variability within families of color, offering a reminder that “families of color” are an exceedingly complex, rather than homogenous, group (Brinkley et al., 2022; Burris et al., 2019). Further, Brinkley et al. (2022) disaggregated the population of parents of color and found that African American mothers reported more traditional literacy practices such as reading to their young children than Latine mothers did. These findings challenge deficit-model perspectives and call for closer examination of reading and other facets of literacy involvement across (socioeconomic and/or ethnically) diverse families.

Parents Beliefs About Language and Literacy

More proximal to practices than demographic factors, however, are parents’ beliefs about early language and literacy (Peixoto et al., 2022). Parental literacy beliefs encompass parents’ perceptions, attitudes, or dispositions regarding a child’s literacy development and the parent’s own role in fostering this growth (Weigel et al., 2005). Parental literacy beliefs are not widely researched, although existing evidence shows that parents’ literacy beliefs are associated with their practices to help young children’s reading (Bingham, 2007; Elliott & Bachman, 2018; Puccioni, 2018), including their phonological awareness (Cottone, 2012), their expressive and receptive vocabulary (Davis et al., 2015), and their knowledge of reading conventions (Curenton & Justice, 2008). Our study will thus concentrate on parents’ beliefs about their child’s literacy development.

One framework for examining parents’ beliefs about early literacy distinguishes between those who target specific skills (often letters and sounds) and those who take a more holistic perspective focused on making meaning, frequently using the classic Parent Reading Belief Inventory (DeBaryshe & Binder, 1994). For example, Lonigan et al. (2013) and Sénéchal et al. (2017) both coded parents’ literacy beliefs into more skill-based beliefs (e.g., “I think my child should learn to read by learning letters”) versus more holistic, meaning-oriented beliefs (e.g., “I think my child should benefit from hearing stories that she has memorized”). Interestingly, Krijnen et al.’s study (2020) found that, not only did parents differ from one another in their perspectives on skills and meaning, but that while most parents regarded the enhancement of oral language skills as a significant objective in home literacy practices, over one-third of the participants believed that the primary responsibility for developing coding skills should primarily lie with the school.

In turn, families’ beliefs about the importance of letter/sound skills and meaning/language/comprehension matters for children’s learning. For example, Lynch et al.’s (2006) research also showed that parents with more skill-based beliefs offered more direct teaching activities (e.g., teaching the alphabet, helping the child write letters), while those with more holistic beliefs engaged in more indirect teaching activities (e.g., providing materials, letting children see parents reading). Similarly, Goldenberg et al. (2001) found that, among Latine kindergartners, parents who valued accurate decoding more than story comprehension focused their involvement more on phonics-based instruction than on understanding. Extending to child outcomes, Hindman and Morrison (2012) suggested that some parents emphasized letters and literacy while others emphasized talking and comprehension. Critically, children of letter-focused families grew more on alphabet knowledge over time, whereas children of meaning-focused families grew more on meaning-focused skills. Still other work has supported specific parts of this equation. For example, DeBaryshe et al. (2000) showed that, compared with mothers whose beliefs were more skills-oriented, mothers holding more holistic beliefs tended to model and strengthen their children’s whole-language literacy behaviors and exploration. Thus, we follow the division of more skills- and code-based vs. more holistic and language-based beliefs to investigate relations among parents’ beliefs, literacy involvement, and children’s early literacy.

Importance of Investigating ECLS-K 2011

The ECLS-K 2011 dataset’s unique value for the exploration of parents’ beliefs and practices lies in its extensive and diverse sample, complemented by high-quality measures, that together provide rich insights into the early literacy of kindergarten-aged children from various backgrounds. Such a dataset enables the study of nuanced factors in home literacy environments and their impact on later academic achievement. It addresses gaps in earlier research that may have relied on smaller, less diverse samples. Furthermore, the continued application of the national dataset in contemporary research, such as those studies by Reynolds et al. (2022) and Rhinehart et al. (2021), illustrates its enduring relevance in yielding significant insights into child development during their kindergarten years.

Aim of the Current Study: Teasing Apart Beliefs and Practices

Early childhood literacy descriptive studies and interventions frequently observe parents’ practices during book reading, but relatively few also collect information about families’ beliefs (Akemoglu et al., 2021; Hindin et al., 2017; Steiner et al., 2021; Sy & Schulenberg, 2005; Timperley et al., 2022; Troseth et al., 2019). Consequently, it is unclear how beliefs and practices together support child outcomes. In theory, beliefs may contribute to child outcomes in ways distinct from practices, such as shaping children’s values or nudging parents to provide specific materials in the household. However, we have little research to explicitly map out these relations, particularly among socio-demographically diverse families. Thus, we aim to test an adequately powered model that teases apart exactly how literacy beliefs predict literacy (book reading) practices, and at the same time how beliefs and practices matter, independently and together, for children’s literacy outcomes. Moreover, because demographic characteristics also matter, we include a variety of factors in a large and diverse sample in order to parse the variability attributed to each construct appropriately. In this study, to learn more about the interconnections between parent beliefs and involvement through book reading in children’s literacy development, we ask the following questions:

  1. (1)

    How are demographic characteristics (ethnicity, SES, child age) predictive of parents’ literacy beliefs and their home reading practices during children’s kindergarten year?

  2. (2)

    How do parents’ literacy beliefs contribute to parents’ home reading practices during children’s kindergarten year?

  3. (3)

    How do parents’ literacy beliefs and home reading practices relate to children’s language and literacy development during their kindergarten year? Specifically, to what extent to they predict incremental, unique variance in kindergarten reading growth, net of the effects of demographic background factors?

In light of existing research (as above), we posit several hypotheses: First, we predict that parents’ literacy beliefs and practices will vary with families’ demographic characteristics, but that differences between ethnic/racial groups and along socioeconomic lines and child age will be modest, contradicting classic deficit-model perspectives that have framed minoritized families’ practices as less rich than those of White and/or affluent families. Second, we anticipate finding a significant, positive link between parents’ beliefs and their reading practices at home. Third, we expect a positive relation between children’s literacy development and both parents’ literacy beliefs and their practices. This integrated hypothesis framework aims to encapsulate the multifaceted nature of the home literacy environment and its relations with emergent literacy.

Method

Participants

We used the ECLS-K: 2011 Kindergarten- Fifth Grade public-use data file, containing information about 18,174 children and their parents, teachers, and schools from kindergarten to fifth grade. When weighted, data represent American kindergarten entrants in 2011, offering a unique window into a diverse group of American families. After normalizing the weight (M = 1) focused on children with outcome data in the fall and spring of kindergarten (W9C19P_20), a total of 9,227 students and parents were included in the dataset. Kindergarten entrants were, on average, 67.5 months of age (SD = 4.4, range: 44.8–90.8). Most students were white (n = 4,700), 2,440 were of Latine heritage, 899 were African American, 637 were Asian, 39 were Native Hawaiian or Pacific Islander, 77 were American Indian or Alaska Native, and 398 were of two or more races. Because of sample size issues, we coded ethnicities of Native Hawaiian or Pacific Islander, American Indian or Alaska Native, and two or more races into a combined category of Other Diverse Participants (N = 514). For most families (78.7%), the responding parent was the biological mother. These parents were, on average, 34.5 years old (SD = 6.59, range: 19–74). On average, children had 1.51 siblings (SD = 1.11, range: 0–2). Parents’ SES, calculated based on parents’ education, occupation, and household income, ranged from − 2.33 to 2.60 (M=-0.02, SD = 0.82), where a value of 0 represents the average SES across the nation at the time of data collection.

Procedure

A total of 1,319 schools were selected for the national survey based on the sample frame for 2006-07 Common Core of Data and the 2007–2008 Private School Survey according to census region, locale, kindergarten enrollment, religious affiliation, percentage of students eligible for the free lunch program, and other school type. Children in these schools were recruited by selected characteristics, including census region, locale, religious affiliation, and ethnicity, to obtain a generally representative sample of American kindergarteners, and weights were developed for participants to ensure representation.

Parental involvement and beliefs data were collected via parent interviews when their children entered kindergarten in Fall 2011. Nearly all (95%) of the interviews were conducted by telephone with the parent or guardian who identified themselves as the person who knew the most about the child (and, as above, yielded 78.7% of biological mothers). The 9,227 students in our study were from 831 schools; on average, each school had around 11 students participating in the data collection. A trained staff member directly assessed the children’s reading scores using a project-designed measure emphasizing basic reading skills such as print familiarity, letter recognition, beginning and ending sounds, recognition of common words, and decoding multisyllabic words in the fall and spring of the kindergarten year. A language screener was used to determine whether children whose home language was not English could particulate in the English assessment.

Measures

Background Variables

Parents’ SES and Ethnicity were Measured through a Survey

Parent Book Reading Practices

Parents’ book reading was measured through a survey. The analysis focuses on parents’ involvement in home reading when the children entered kindergarten in Fall 2011 (our three specific variables were frequency of reading books to children, length of each reading time, and number of children’s books). The specific phone interview question related to parental involvement was How often do you read books to your child? with possible responses including not at all (1), once or twice a week (2), 3–6 times a week (3), and every day (4). Other questions included: How long do you read to your child each time? and How many books does your child have? Responses featured parent-reported minutes and numbers (respectively).

Very few (1.2%) parents reported not reading to their children, while 13.4% read once or twice a week, 35.1% read 3 to 6 times a week, and 50.3% read every day. Reported length per reading averaged 20.27 min (SD = 10.12), ranging widely from 1 to 60 min. Children’s reported number of books at home averaged 88.03 (SD = 133.72), ranging widely from 0 to 4000 (Table 1). Correlations between these variables were weak: those who read more often read for slightly longer (r = .09, p < .001) and often had slightly more books (r = .19, p < .001). In addition, those with more books read for slightly longer (r = .07, p < .001).

Table 1 Descriptive data of independent variables
Full size table

Exploring differences across ethnicities, there were only nuanced differences in the reading practice variables across ethnic groups, as ANOVA showed that there was a significant difference between groups in reading frequency F(4, 8408) = 120.32, p < .001, length of reading F(4, 8301) = 23.95, p < .001, and number of books F(4, 8384) = 130.50, p < .001. As for reading frequency (see Table 2), the White group read books most frequently, followed by Other Diverse families, African American families, Asian families, and Hispanic families; differences between each group were generally statistically significant but of a small effect size (see Table 3). As for the length of reading, African Americans reported the longest reading duration, followed by Other Diverse families, Hispanic families, Asian families, and White families. As for the number of children’s books at home, the White group had more books than all the other groups, followed by Other Diverse families, Asian, Hispanic, and African American families, although effect sizes were very small to small in magnitude.

Table 2 Descriptive analysis of beliefs and practices by ethnicities
Full size table
Table 3 Results of ANOVA examining the effect sizes of group differences
Full size table

Parents Literacy Beliefs

Parents’ literacy beliefs were assessed using two questions administered during the phone survey. The questions related to parents’ literacy beliefs were: How important is it for your child to know letters? How important is it for your children to express their needs through language? The responses to the questions ranged from essential (1) to not important (5). We reverse-coded the response for ease of interpretation.

Regarding the importance of knowing letters (See Table 2), 0.4% thought it was not important, 1.7% not very important, and 16.5% somewhat important; most rated letters as very important (50.0%), while 31.4% rated letters as essential. As for the importance of expressing needs through language, 0.05% of parents considered it unimportant, 0.1% not very important, and 3.8% somewhat important; most found it very important (48.8%) or essential (38.6%). The zero-order correlation between these variables (letter and language beliefs) was 0.41 (p < .001), suggesting that families who rated one item highly may rate the other highly similarly. However, many families rated one more highly than the other (t =-0.34.43, p < .001).

As for differences across ethnicities, 93.6% of African American parents, 88.4% of Hispanic parents, 80.6% of parents from Other Diverse groups, 79.2% of Asian parents, and 76.1% of White parents considered knowing letters very important or essential. ANOVA showed a significant difference between the groups in parental beliefs about letter importance, F(4, 8420) = 21.35, p < .001. Tukey’s HSD Test for multiple comparisons found that there were significant differences between nearly every group, such that the African American group endorsed a higher level of importance for language skills (93.6%) than any other groups (p < .001). However, analysis of effect sizes revealed small differences (see Table 3).

Compared with this, 96.8% of African American parents, 96.4% of Hispanic parents, 96.2% of White parents, 94% of Other Diverse parents, and 89.0% of Asian parents considered expressing needs via language as very important or essential. ANOVA found a significant difference among groups in holistic language belief, F(4, 8418) = 53.57, p < .001, with significant differences between nearly every group. However, effect sizes were small (see Table 3).

Child Reading Skills

The child reading IRT score in the national dataset in Fall 2010 and Spring 2011 captured children’s literacy development. The reading assessment focused on word recognition, vocabulary knowledge, and reading comprehension. The assessment was administered to the children individually by trained and certified child assessors. The reliability of the reading assessment with 205 items was 0.92 for Fall 2010 kindergarten and 0.94 for Spring 2011 kindergarten. Children’s reading scores in fall averaged 54.78 (SD = 1.74), ranging from 33.14 to 164.33. Their reading scores in spring averaged 69.61 (SD = 14.68), ranging from 32.52 to 133.55.

Missing Data

Within the weighted sample (n = 9,227), missingness ranged from 0.39% for reading scores in Fall 2010 and Spring 2011 to 8.5% for parent literacy beliefs and home reading practices. We used ANOVA and chi-square analyses to determine whether missingness was related to ethnicity and socioeconomic status. Analyses showed that there were significant differences in missingness based on ethnicity, F(1, 9225) = 144.84, p < .001, and socioeconomic status, F(9225) = 144.44, p < .001. We thus assumed data to be missing at random (Widaman, 2006), meaning in conjunction with variables in the dataset, and we used multiple imputation (50 times) to deal with the data missingness in R by using the MICE function.

Statistical Analysis Plan

R was used for descriptive analysis and to test regression models, accounting for the clustering of families/children within schools. As can also be seen from Table 1, distributions of data, including fall reading score, spring reading score, length per reading, and book numbers, are skewed. Thus, in R, the bias-corrected bootstrapping method was used to handle non-normality in estimating effects.

Results

Q1: How are Demographic Characteristics Predictive of Family Literacy Beliefs and Home Reading Practices?

We first explored differences in variables of interest across background factors. Correlations between SES, child age, and the other variables can be found in Table 4. Overall, most variables showed modest correlations with SES, except for skill-based literacy belief (r = -.03, p < .001) and reading time length (r = -.06, p < .001). However, child age was only significantly correlated with parents’ holistic language belief (r = .04, p < .01). We next created one regression model for each of the two beliefs and one for each of the three types of home literacy practices, both of which included SES, ethnicity, and child age as predictors. Complete results are presented in Tables 5 and 6.

Table 4 Correlation between SES, child age, and parent literacy beliefs and practices
Full size table
Table 5 Multilevel regression model results for child reading outcome
Full size table
Table 6 Semi-partial r squared of the multilevel regression model
Full size table

Beliefs – Skills-Based Belief

Regarding beliefs about letters, families of African American (β = 0.25, p < .001) and Hispanic (β = 0.09, p < .001) backgrounds reported stronger beliefs than white families (reference group) about the importance of letter knowledge. SES was not a predictor of families’ endorsement of the importance of letter knowledge (p = .994).

Beliefs – Holistic Language Belief

Concerning beliefs about holistic language, families of Hispanic (β = − 0.0.11, p < .001) and Asian (β = -0.27, p < .001) backgrounds reported less emphasis on language than White and African American families. In addition, higher SES (β = 0.08, p < .001) was a predictor of stronger beliefs about the importance of language. Across all analyses, effect sizes were small.

Practices - Frequency

 Considering parents’ home reading practices, parents’ frequency of reading to their children was significantly predicted by parents’ demographic characteristics, in that families of color – including Asian (β = -0.33, p < .001), Hispanic (β = -0.28, p < .001), and African American (β = -0.11, p < .001) families reported less frequent reading than White families. Higher SES (β = 0.19, p < .001) and child’s younger age (β = − 0.007, p < .001) was also linked to more frequent reading.

Practices – Time Length

 Regarding reading time length, families of African American (β = 2.99, p < .001), Hispanic (β = 1.40, p < .001), and Other Diverse ethnicities (β = 2.04, p < .001) read for longer than White families.

Practices – Books

 In addition, number of children’s books was significantly predicted by minority ethnicity, with fewer books in the home reported by Asian (β = -57.55, p < .001), African American (β = -50.97, p < .001), Hispanic (β = -41.09, p < .001), and other ethnicities (β = -11.84, p = .048). More affluent families had more books (β = 34.46, p < .001).

Q2: How Do Parents’ Literacy Beliefs Contribute to Parents’ Home Reading Practices?

To test the links between beliefs and practices, we conducted three additional regressions for each belief, with practices as outcomes and controlling for the covariates including SES, ethnicity, and child age. Parents’ beliefs about expressing needs with holistic language predicted the frequency of reading books (β = 0.05, p = .005) and number of children’s books in the home (β = 9.06, p < .001), but it did not predict the length of reading (p = .507). Parents’ beliefs about letter importance significantly predicted the frequency of book reading (β = 0.04, p = .022) and length of reading (β = 0.64, p < .001) and negatively predicted the number of books (β=-4.97, p = .012).

Q3: How Do Parents’ Literacy Beliefs and Home Reading Practices Relate to Children’s Language and Literacy Development?

In this multilevel analysis, we modeled the two measurement time points (Fall 2010 and Spring 2011) at level 1, time-invariant child characteristics at level 2, and school at level 3; we first fit a null model to understand the extent to which the outcome was nested within children and schools, after which we fit a random intercept model to explore whether change over time among children varied at either level. The results showed that 67% of the variance in reading scores was explained by differences between children, while 8% was explained by schools, necessitating a three-level model to answer this research question.

On average, reading score increased 14.6 points from Fall 2010 to Spring 2011 (p < .001). Most relevant for the current study, accounting for these covariates, reading growth was stronger in families with more skills-oriented belief (β = 1.99, p < .001), and more frequent book reading with children (β = 1.22, p < .001). In addition, Change was more pronounced among higher SES children (β = 5.04, p < .001), those of Asian ethnicity (β = 4.52, p < .001) and Other Diverse ethnicity (β = 0.98, p < .01), and older children (β = 0.41, p < .001). Learning was slower among African American (β = -1.21, p < .001) and Hispanic (β =-1.23, p < .001) negatively predict reading scores.

As for the incremental validity for the variables, drawing on the semi-partial R squared of the model, the results indicate that parents’ skills-oriented literacy belief explains 0.81% of the variance int eh outcome (CI: 1.67–2.32), while parents’ frequency of reading 0.42% (CI: 0.90–1.53) also contributed (0.42%) to the overall model. The effects of parents’ holistic language belief (R2 = 0.0062%, CI: − 0.16-0.71) and the time length per reading (R2 = 0.0046%, CI: − 0.0069-0.038) were both neglectable and not significant (p > .01). The overall model is quite effective, explaining 47.04% of the variance in the outcome variable.

Discussion

The study aimed to examine the relationship between parents’ literacy beliefs, parental involvement, and children’s literacy outcome in kindergarten, while also looking into SES and ethnicity. First, the results show that most parents, regardless of ethnicity, highly endorsed the importance of letter knowledge and language competence for their young children; interestingly, families from African American and Hispanic backgrounds more strongly emphasized letter knowledge, while White and African American families, as well as more affluent families, more strongly endorsed language development. Second, SES was not a predictor of families’ endorsement of the importance of letter knowledge. However, higher SES was a predictor of stronger beliefs about the importance of language. Third, on average, American parents read to their children nearly every day for approximately 20 min each time, but while White families reported reading more often, those of African American, Hispanic, and other non-White ethnicities read for more time. Thus, some subtle demographic differences emerged.

Accounting for various background variables, we pinpointed complex ways in which parents’ beliefs mattered for their practices. Specifically, parents who more heavily endorsed language skills read more often and had more books, but did not read for longer periods. Conversely, parents who valued letter knowledge read more frequently and for longer sessions, but had fewer books in the home. At kindergarten entry (Fall 2010), reading scores were higher among children whose parents emphasized letter knowledge, read more often, and had more books. At the end of the year (Spring 2011), reading achievement was still higher among families who read more often and emphasized letter and/or language knowledge.

Overall, by exploring several facets of parents’ beliefs and practices around reading in a large, diverse, nationally representative sample, we were able to map out in a nuanced way the importance of various types of family inputs and child outcomes, as well as to challenge deficit-model perspectives about the habits and contributions of families from minoritized and/or economically disadvantaged backgrounds.

Variability of Literacy Beliefs and Reading Involvement

An intriguing finding of the current paper is that there is diversity in parents’ beliefs about the importance of knowing letters and using language across and within ethnic groups, in parents’ literacy involvement at home through three distinct facets of book reading-related practice (frequency, length, and number of books). The present data reveals high average levels of value for early learning and use of practices around book reading across all sociodemographic groups. Indeed, in the data, families of color endorse both letters and language as important, and some groups exceed White families’ ratings on the former. This finding is particularly significant in light of earlier research, which suggested that families in minoritized communities have less investment in children’s learning (Auerbach, 1989; Sosa, 1996; Raikes et al., 2006), an idea that neglects the biased structural, racial, and economic systems in which these families are situated. Our study challenges these deficit-model perspectives, in line with recent evidence which recognizes the active involvement of families of color in literacy development (Brinkley et al., 2022; Burris et al., 2019; Davis et al., 2015).

Across belief and practice, this work also revealed differences within ethnic groups, providing additional evidence that families from minoritized populations cannot be thought of as part of a monolithic group and that ethnic background does not determine families’ involvement behavior (see Hindman et al., 2012 for similar conclusions with younger children). Together, results remind us that all families are best understood in nuanced ways. In the future, more interviews can be conducted among the ethnic groups to investigate their reading quality and content, such as conversations the parents have with children, the affective aspects of the interaction, and the literacy skills-related or inquiry-oriented discussion.

Literacy Beliefs Predicting Home Reading Involvement

Another significant contribution of this work is the finding that parents’ beliefs can be enacted in home reading practices such as providing and reading books with children. Over time, research about parents’ beliefs has often involved smaller and less diverse samples, making it difficult to understand how beliefs and practices might be intertwined (Elliott & Bachman, 2018; Skibbe et al., 2008). The current study uses a large national dataset to tease apart these links. Interestingly, we found that parents’ beliefs about the importance of letters were positively related to frequency and length of reading, and parents’ holistic language beliefs were linked to the frequency of reading and number of books. This discrepant pattern of findings echoes some earlier work (Hindman & Morrison, 2012; Lonigan et al., 2013; Sénéchal et al., 2017) that highlights differences among parents in their emphasis on code- vs. meaning-related literacy activities. Thus, this study lends some additional support to the idea that some parents may focus on more highly teachable, constrained skills (sometimes colloquially referred to as ABCs and 123s). In contrast, others may target more diffuse, global outcomes such as vocabulary and general knowledge.

Literacy Beliefs and Reading Outcomes

We focus on investigating the effect of literacy beliefs and practices on children’s reading outcome changes through multilevel modeling. Parents’ skills-oriented beliefs predicted their children’s reading changes through the kindergarten year with even larger effect size than their frequency of reading books. This predictive relationship likely emerges because alphabetic knowledge is a foundational skill for reading that factors heavily into emergent reading skills at the start of kindergarten (NELP, 2008). Moreover, as children are just entering formal schooling, much of their prior exposure to literacy still depends on parents’ beliefs and practices (Tsirmpa et al., 2021), which emphasizes the unique contribution of the parents’ practices to variations in children’s learning in addition to classroom and peer sources. However, it is also possible that parents’ early belief in the importance of alphabet skills (akin to children’s development of these skills) often coincides with other beliefs and experiences that foster later reading, making these beliefs a marker for interest in other aspects of literacy development as well.

Parents’ holistic language-related beliefs did not predict reading outcome changes. One possible explanation may lie in the relatively small focus in kindergarten classrooms on oral language and vocabulary development, including oral assessment in reading (Cunningham et al., 2009; Morrison et al., 2019; Wright & Neuman, 2014). We found this finding interesting because, in theory, as classrooms tend dedicate more time to code-focused instruction, home exposure to new words and encouragement for self-expression may play a relatively greater role in children’s development over the year. Perhaps this association does not emerge in the current data because parents’ beliefs about expressing needs with holistic language are themselves predictors of the frequency of reading books and the number of children’s books in the home; thus, parents’ holistic beliefs still make contributions to children’s reading, but future research could address the possible mediation role of parents’ beliefs on reading outcomes through parents’ practices.

Limitations and Future Research

This research employs a quantitative dataset analysis approach with some limitations. First of all, due to the smaller populations of certain ethnic groups, we created a single Other Diverse Group category, which limits our capacity to understand the unique experiences of individuals within this group. Secondly, parents’ literacy beliefs were measured with two items targeting (respectively) holistic language-focused beliefs and skills-based, letter-knowledge beliefs. However, future researchers should specify more detailed beliefs related to individual elements of literacy, such as letter recognition, semantics, print knowledge, and reading comprehension. Third, the current study explores only English reading skills; similar work exploring reading development in languages other than English would be of value. Fourth, our study’s scope of assessing parents’ literacy beliefs was limited to inquiring about the value they place on their children developing pre-literacy skills, which does not fully encompass the broader range of parental perceptions about their role in literacy development as outlined in the theoretical framework. Last, more qualitative work is needed to present the complexity and variability of the parents’ literacy practices and possible reasons underlying these phenomena.

Conclusion

Parents’ beliefs are not a heavily researched topic in early childhood. However, our research suggests that parents’ beliefs uniquely explain variability in their home practices related to children’s literacy outcomes in kindergarten. Parents from different ethnic backgrounds vary in their literacy beliefs, with families of color more positively endorsing the importance of early literacy skills relative to white families. These finding challenges deficit-model conclusions about families of color. Moreover, different types of beliefs (i.e., literacy-focused versus language-focused) are linked to different forms of parental literacy involvement at home, and both beliefs and practices impact child outcomes.