Shared book reading: a Norwegian survey of reading practices in families

ABSTRACT

Access to books and a rich language environment at home are important for children’s language development. In this study we explored self-reported reading practices in families in Norway (N = 1001) to gain insight into the reading habits parents have with their young children, and the factors that best explain book reading in Norwegian homes. By investigating relationships between how many books parents reported reading with their child last week, demographic factors, and orientational variables, we found that book reading is better explained by orientational factors – such as children’s interest in books and having reading routines in the home – than by demographic factors (such as parents’ educational level). The proportion of parents reporting reading no books, or only one book with their child in the last week was unexpectedly large (33%). The implication of these findings for the culture of reading in Norwegian homes is discussed.

KEYWORDS:

• Reading practices
• 3- to 6-year-olds
• home reading
• reading routines
• shared book reading

Introduction

It is well established that the literacy environment at home is the most important arena for children’s early language learning, and research shows that the literacy environment at home has a direct effect on children’s vocabulary and pre-reading skills (Sénéchal & LeFevre, 2014; Sylva et al., 2004). Children’s language skills develop interdependently with their literacy experiences (Saracho, 2017) and a rich literacy environment is characterized by shared book reading (SBR), cognitively challenging talk, and access to books (Mol & Neuman, 2014; Worthington & van Oers, 2017). Despite the likelihood that the home is the most important context for children’s language and literacy development, literacy activities in homes are an underexplored field, especially within the Norwegian context. Previous studies in this area have provided information about the impact of SBR at home on children’s second language skills (Grøver et al., 2020), and which socio-material factors that play a role in parents’ attitudes to SBR at home (Kucirkova & Grøver, 2022). Still, to our knowledge, there are no studies aiming to gain knowledge on the extent parents read with their children under the age of six in Norwegian families, and which factors best predict the frequency of reported book reading. In this study, we explore parents’ reports of their reading practices, based on parental responses to a representative, national survey of families in Norway. We thereby contribute knowledge to the field of SBR at home, that can form a benchmark for future studies in the Norwegian context and be a guideline for policy makers’ efforts to improve literacy experiences with young children.

Why read with children: the Norwegian context

There are good reasons to assume a strong reading culture in Norway and that SBR is perceived by parents as important. Children’s literature has a long tradition in Norway, with its own support programs and institutions. The role of literature and shared reading in children’s upbringing has found legitimacy as an educational function, and the principle that literature should be both useful and entertaining has been the central focus (Mjør et al., 2000). The first Norwegian picture book, where rhymes, lullabies, and songs were collected entirely with children in mind, was published in 1888 (by Holst & Nielsen, 1888), and children’s literature has been available and accessible to families in Norway for almost 100 years. The government procurement schemes for literature, which include children’s literature, have made an important contribution to children and families having free access to literature through public and school libraries (Arts Council Norway, 2023). By Norwegian law (Public library act, 1986), all municipalities in Norway must have a public library, and in areas with scatted population a book buss or a book boat provide opportunities to loan books for free, which has an equalizing effect between families with home ownership of books and those who don’t.

Children’s books are culturally nuanced and closely linked to educational guidelines, and contemporary Norwegian children’s literature production is characterized by aesthetics, language, and form that are as important as themes. In addition to the historical emphasis on reading at home, SBR is emphasized in the curriculum for ECEC institutions (Norwegian Directorate for Education and Training, 2017) and over 90% of all children aged 1–5 attend an ECEC centre (Statistics Norway, 2022a).

The linguistic stimulation provided through children’s books is considered a unique factor in parent–child interactions as it expands adult speech with a variety of words and grammatical complexity (Logan et al., 2019). In comparison with other popular activities, such as reminiscing, play (Crain-Thoreson et al., 2001), and TV co-viewing, parents produce more and richer language during SBR (Hanson et al., 2021). Beyond language and literacy stimulation, book reading fosters social and emotional development (Cutler & Palkovitz, 2020), and bonding between caregivers and children (Audet et al., 2008). Furthermore, children who grow up in homes with frequent SBR develop positive reading habits over time, a higher predisposition to read frequently and broadly (Baker et al., 1997), and higher academic achievement in later years (Barnes & Puccioni, 2017).

Due to a national focus on the production of quality children’s books in Norwegian, and wide publication of the positive effects of SBR, most Norwegian parents are familiar with the concept and benefits of reading with their children. These benefits are regularly communicated during mandatory health care visits as a part of the national health program (Norway Health, 2020) and later through childcare centres. Literacy activities, both those that parents do together with their children and those that parents and children do on their own, are seen as having cultural value for children’s development and well-being. Therefore, surveys targeting parental involvement in literacy activities and the facilitation of early reading endeavours, may be vulnerable to conformity and social desirability bias, leading to a possible overreporting of positive behaviour (Krumpal, 2011). It is then surprising that when parents of 5th grade children were asked to report retrospectively on early reading activities in the Progress in International Reading Literacy Study (PIRLS) 2016 survey (IEA, 2016), only 38% of Norwegian parents reported that they often engaged their children in literacy activities during early childhood. Danish and Swedish parents reported similarly low rates of early reading activities, placing the Nordic countries far behind the top-ranking countries such as Russia (65%), Kazakhstan (65%), and Georgia (56%).

Similarly, the Norwegian Book Sellers Association has reported a decline in parental reading compared to earlier surveys (Bokhandlerforeningen, 2022). More parents than before report that they don’t read for their child or do so rarely. However, the proportion of parents reporting that they read often for their children has increased a little over the years. According to their survey 64% of parents with children under the age of 10 in their household report reading for their children daily or two to three times a week, and women tended to read more than men. Reading habits were also closely related to parent education level. However, as this was a commercial, not scientific survey, the results should be interpreted with caution.

Shared book reading (SBR) and the factors that influence its frequency

Home literacy activities encompass a broad range of endeavours, from library visits and daily reading routines to activities more directly related to children’s language and literacy learning, such as shared, conversational, and repeated reading (Sénéchal et al., 2017). Several interview studies of parents’ orientations towards SBR have shown that parents have positive attitudes towards the practice (e.g., Farrant & Zubrick, 2013), and the frequency of parent–child dyadic book reading has been shown to be a significant predictor for later internal motivation for reading (Demir-Lira et al., 2018). Motivation for reading is commonly measured by surveying parents and asking how many books they read with their child in the last week.

Sylva and colleagues (2004) have shown that SBR predicts children’s literacy skills independent of family SES, however, other studies have found clear evidence of the influence of gender and socioeconomic status (SES) on the frequency of parent–child reading (see, for example, Compton-Lilly et al., 2012; Guiberson & Ferris, 2019). Research has also shown how positive parental attitudes and beliefs about reading influence a child’s interest in reading (DeBaryshe, 1995), and may predict the number of books that parents read with their child. However, it has been suggested that a type of “Matthew effect” may exist, where children in families that enjoy better living conditions, may benefit more from parent educational and interventions promoting at home literacy activities (Bakermans-Kranenburg et al., 2005). A variety of factors should therefore be considered in SBR research, and in the design of effective interventions for supporting SBR (Xu & Gao, 2021).

Demographic factors

Demographic factors, such as parents’ gender and levels of education and income, have been found to be related to the frequency of home literacy activities (see, for example, Bokhandlerforeningen, 2022; Hofslundsengen et al., 2019; Napoli et al., 2021).

In other studies, families that report lower socio-economic status (SES) also report reading less frequently than families with higher SES (Bus et al., 1995), and across ten European countries, it was found that lower SES families engaged in fewer literacy activities with their young children (Hemmerechts et al., 2017). Children from higher SES Dutch families have been observed to experience a more stimulating home literacy environment, and this, in turn, has resulted in better vocabulary and reading comprehension skills at school (Van Steensel, 2006). The level of income and parental education in families, commonly used as indicators of SES, have been shown to be reliable predictors of SBR in families (Levine et al., 2020), and this can be explained by the Family Invest Model (Conger & Donnellan, 2007; Vasilyeva et al., 2018). According to this theory, families characterized by higher incomes have the freedom to invest their resources in a way that promote their children’s development, while lower income families will prioritize urgent needs. This also applies to other types of social capital, such as educational level, and implies that parents with higher educational backgrounds will prioritize activities that stimulate their children’s academic and social development. They will also possess knowledge, skills, and values that will have an impact on, for example, how much time they invest in SBR and the number of books they read. In this study, which is part of a wider study, we did not gather information on family income but hypothesize that parents’ educational level will be associated with their reading practices.

Parental orientations towards reading

Positive parental attitudes and beliefs about SBR have a direct influence on children’s reading interest, as reported by parents or observed in situ (DeBaryshe, 1995). In this study, we have considered parental orientations towards SBR in terms of their reported own interest in reading and the reading routines they have established with their children.

Parental interest in reading. Results from the latest PIRLS survey (IEA, 2016) have shown that children from a low socioeconomic background that have parents that enjoy reading, perform just as well in reading as their peers from a high socioeconomic background with parents that don’t enjoy reading (Støle et al., 2021). This suggest that parents’ enjoyment of reading encourages their children’s interest for reading. Parental attitudes towards reading have been found to relate to the reading opportunities that they provide their children (DeBaryshe & Binder, 1994). Parents own reading habits, such as the degree to which they enjoy reading for their own pleasure, are also likely to correspond with how much they read with their own children (Hartas, 2012) and we expect parental interest for their own reading to be related to both their children’s interest in book reading and how many books they read each week to their children.

Reading routines. Routines can be understood as daily or repetitive behaviours that directly involve the child and at least one adult, who acts in an interactive or indicative way (Sytsma et al., 2001). Regular, and positive routines have been shown to have positive effects on important areas in family life, such as parent–child closeness and stability (Ren & Fan, 2019).

The content of family routines changes naturally with the child’s age. For example, due to age related variation in attention, reading activities and routines with toddlers are likely to be shorter than with 4- to 5-year-olds. The degree to which the child’s age influences family SBR has not been widely studied. However, research from the ECEC and school contexts shows that reading and language activities decrease with the child’s age (Lunde, 2012; Skaftun et al., 2021).

Routines form structural frames for a family’s everyday life, and if reading activities are a part of these routines, it is likely that parents with such routines will read more to their children than those who do not. We therefore hypothesize that the existence of routines will be an explanatory factor for the number of books that parents report reading with their children.

Child’s interest in reading

An indicator of children’s interest in book reading, and other reading activities, is how often they ask to be read to or independently interact with books. Previous studies have shown that participation in reading activities in ECEC is predicted by children’s language skills, and that children with weak language skills, in particular boys, have less interest in and often avoid literacy activities (Stangeland et al., 2018).

In a socio-culturally diverse sample, almost all interviewed parents said that their child enjoyed being read to, regardless of gender, income level, or ethnicity (Baker & Scher, 2002). The literature is, however, less clear on what children’s interest in reading, as reported by their parents, predicts (see Whitehurst & Lonigan, 1998).

While parents’ interest in reading is relatively stable, children’s interest in reading evolves over time (Hume et al., 2015). Farver and colleagues (2006) found a predictive value of parents’ reports of children’s self-initiated book reading for the children’s reading achievement in second grade. Furthermore, Bracken and Fischel’s (2008) regression analysis of parent reports of reading behaviours of 233 low-income pre-schoolers found that children’s reading interest is a small but significant predictor of children’s letter knowledge. Therefore, if the child’s interest in book reading is indicative for how often their parents read with them, the child’s interest, or lack thereof, will also be an explanatory factor for the number of books that are read.

In the current study, we therefore expected there to be a relationship between parents’ reports of their own orientations towards reading, and their child’s interest in book reading, with the reported number of books that were read in the previous week.

The current study

In this study we ask

1. How are demographic factors related to how many books parents report reading with their children aged 3–6 years?

2. How are parental orientations related to how many books parents report reading with their children aged 3–6 years?

3. How is the child’s interest related to how many books parents report reading with their children aged 3–6 years?

Methods

Procedure

This study is based on self-reported survey data. The questions in the survey were based on areas that were identified as important in a preliminary literature review, as well as questions used in an earlier small-scale study on parents’ reading habits. The draft questions were reviewed by early childhood experts from the National Centre for Reading Education and Research, Norway, and modified where recommended.

The fieldwork for the survey was outsourced to an online survey company (Norstat Ltd.), specialists in conducting online surveys in Norway since 1997, with access to a nationally representative database of parents from diverse backgrounds and geographical locations across Norway. Norstat recruited respondents for the survey, distributed the survey to the participants, and ensured a 100% response rate.

For this study, researchers selected and analysed the questions covering the previously documented predictors of SBR (demographic variables) as well as theoretically driven predictors (variables measuring parental orientations and child’s interest).

Measure of shared book reading (SBR)

Parents were asked to respond to the question “During the last week, how many books have you read with your child?” by selecting an option for 1 book through to 6 books, “more than 6 books”, or “no books”. Responses to this question were operationalized as the measure of parent reported SBR and were analysed as both a continuous and categorical outcome at different stages of the study, as described in the following sections.

Respondents’ demographic characteristics

The survey was completed by 1001 parents living in Norway. From that set, 961 parents reported complete data on all demographic variables, and constitute the reduced sample used in this study. This missingness can be considered random, as can be observed in the insignificant variation in the descriptive statistics presented in Table 1.

Table 1. Descriptive demographic statistics for full survey and reduced sample.

	Full survey	Reduced sample
Number of cases	1001	961
Responding parent gender (female): number (percentage)	502 (50%)	481 (50%)
Age responding parent: mean (SD)	37.9 (6.58)	37.9 (6.43)
University education responding parent: number (percentage)	645 (65%)	636 (66%)
Child gender (female): number (percentage)	515 (51%)	495 (52%)
Age child: mean (SD)	4.3 (0.96)	4.3 (0.96)
Number of children in family: mean (SD)	2.1 (0.85)	2.1 (0.86)

Parents had a mean age of 37.9 years, with 55% of the sample aged between 30 and 39 years old. Given the targeted recruitment strategy, the parents’ gender was evenly distributed across male and female, with less than 1% responding that they preferred not to answer or identified as non-binary. Over 96% of respondents had completed high school, with 66% reporting having also completed university education, both indicators in line with the level of education for the population between 30 and 40 years of age in Norway (Statistics Norway, 2022b). All respondents had a child aged between 3 and 6 years at home (which was a condition for answering the survey), and if parents stated that they had several children between the ages of 3–6, they were instructed to respond considering the child who most recently had a birthday. The children’s ages were evenly distributed across 3, 4, 5, and 6 years old (mean 4.3), and 97% of the children were attending an ECEC centre. The children’s gender was evenly distributed across girls and boys, with less than 1% of respondents stating that they preferred not to provide the gender of the child or that the child identified as non-binary. On average there were 2.1 children per household, with 24% of households being classified as ‘large families’ with 3 or more children. Parents were also asked about the language spoken at home and responses indicated that 96% of families spoke Norwegian at home, and 15% spoke another language as well as, or instead of, Norwegian. We explored this variable, but found it had no analytical impact.

Variables on parental orientations towards SBR, and children’s interest in SBR

In addition to exploring the relationship between demographic variables and parent reported SBR (operationalized as the number of books read with the child in the last week), we explored the relationships between parent reported SBR, parental orientations towards reading, and the child’s interest in reading. Table A1 (see appendix) displays the questions used for these analyses, the response categories for each, and their summary statistics.

Analytic procedure

For each category of responses where a logical grouping of data occurred, a Pearson Chi-square test was performed to investigate differences in distributions across groups, beyond what could reasonably be expected to randomly occur. Pearson correlations explored individual linear relationships between continuous and ordinal variables (with a consistent and meaningful order) and the outcome. Finally, multiple linear regression was used to assess the proportion of the variance in the outcome that could be explained by these variables.

Analyses were conducted independently by two researchers, one using R version 4.2.0 (R Core Team, 2021) and Rstudio version 2022.02.3 (Rstudio Team, 2021) and the other using SPSS Statistics 28 (IBM Corp., 2021) and the results were compared. No significant nor substantive discrepancies occurred, and the results presented here are those of the analyses conducted in R. Alpha was set at 0.05, and p-values are reported with the standard notation of *** p < 0.001, ** p < 0.01, and * p < 0.05. No assumptions of causality or directionality were made due to the non-experimental nature of the data.

Results

Comparisons across demographic groups

For the purpose of analysing whether significant differences occurred across groups in our data, we first categorized our outcome, number of books read in the last week, as displayed in Table 2.

Table 2. Summary of number of books of read (parent reported SBR).

Number of books read with child in the last week
0	1	2	3	4	5	6	More than 6
n = 117	n = 204	n = 177	n = 136	n = 79	n = 62	n = 34	n = 152
Read very few		Read some			Read a lot
33%		41%			26%

The subsequent analysis showed several significant differences in the reported number of books read across groups, beyond what could reasonably be expected to randomly occur (as evidenced by a statistically significant Chi-square test statistic). Mothers reported reading “some” or “a lot” of books with their child more frequently than fathers (χ² = 33.31, p = 0.03), as did parents aged under 38 compared to older parents (χ² = 33.31, p < 0.001) and parents with university level education compared to those with less education (χ² = 39.61, p < 0.001). No statistically significant difference in the number of books reported to be read was observed when grouped by the gender of the child (χ² = 1.706, p = 0.43), but 76% of parents of 3- to 4-year-old children reported reading “some” or “a lot” of books with their children, while only 55% of parents of 5- to 6-year-olds reported the same (χ² = 48.57, p < 0.001). Finally, whereas 69% of families with just 1 or 2 children reported reading “some” or “a lot” of books with their children, that figure dropped to 59% in larger families (χ² = 14.42, p < 0.001).

To identify the reasons why parents read with their child, they were asked to select three reasons, including the option “other”, from a list of nine. As shown in Table A2 (see appendix), the top reasons selected by the survey respondents were: (1) “Because it’s cosy” n = 789, (2) “To provide language stimulation” n = 489, and (3) “To give the child experiences” n = 473, closely followed by “Because it promotes reading and writing skills” n = 448. Overall, 73% of respondents selected at least 1 academic reason for reading with their child (either “To provide language stimulation” or “Because it promotes reading and writing skills”, or both), but no significant difference in the frequency of these responses across demographic groups was observed.

Variance explained by demographics, parental orientations, and child’s interest

In the second stage of this analysis, we explored the linear relationship between our variables of interest and our outcome – the number of books parents reported to have read with the child in the last week. Firstly, we examined the Pearson correlations between each variable and the outcome, as displayed in the correlation matrix in Table 3.

Table 3. Correlation matrix for all variables.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1 Parent ed. Level	1.00
2 Parent age	0.08 **	1.00
3 Parent (female)	0.05	−0.29***	1.00
4 Child (female)	−0.07 *	−0.09 **	0.06	1.00
5 Child age	−0.02	0.09 **	−0.09 **	0.00	1.00
6 Children in family	−0.02	0.10 **	−0.04	0.01	0.18***	1.00
7 Regular reading routines	0.16***	−0.01	0.07 *	−0.02	−0.03	−0.09 **	1.00
8 Usually read with child	−0.05	−0.08 *	0.22***	0.04	−0.01	−0.04	0.08 *	1.00
9 Enjoy read with child	0.06	−0.01	0.12***	0.00	−0.11***	−0.05	0.27***	0.21***	1.00
10 Enjoy read alone	0.17***	0.10 **	0.25***	−0.03	−0.10 **	−0.09 **	0.24***	0.23***	0.42***	1.00
11 Child alone takes book (reverse coded)	0.01	−0.06 *	0.08 *	0.08 *	−0.08 **	0.05	0.13***	0.01	0.12***	0.07 *	1.00
12 Child doesn’t like being read to	−0.10 **	0.00	−0 05	0.01	0.05	−0.02	−0.24***	0.00	−0.34***	−0.19***	−0.14***	1.00
13 Number of times child requested to be read to	0.16***	−0.12***	0.11***	0.04	−0.12***	−0.06	0.48***	0.03	0.26***	0.17***	0.31***	−0.29***	1.00
14 Number of books read	0.15***	−0.12***	0.16***	0.05	−0.20***	−0.14***	0.46***	0.02	0.27***	0.22***	0.29***	−0.26***	0.66***	1.00

Note: ***p < 0.001, **p < 0.01, and *p < 0.05.

An exploration of the correlation matrix indicated that a number of variables were correlated with each other. For example, having reading routines, the reported number of times the child requested to be read to, and the number of books reported read with the child, are all moderately and significantly correlated. That said, tests to evaluate whether the data met the assumption of collinearity indicated that multicollinearity was not a concern for the final model described in the following paragraphs. Tolerance for each variable was between 0.69 and 0.95 (well above 0.10, generally considered an indicator of multicollinearity) and VIF was between 1.02 and 1.44 (well below 2.5, conservatively considered a reason for further investigation of multicollinearity).

After examining the correlations, we then grouped the variables into the classifications of demographics, parental orientations, and child’s interest. To explore how much of the variance in the outcome could be explained by each classification, a series of multiple linear regressions were run incorporating the variables that had previously been identified as having significant correlations, and the R² statistic was compared. Finally, a combined model including all variables, and retaining only those that were statistically significant, measured how much variance in the number of books parents reported having read with their child could be explained in total by demographics, parental orientations, and child’s interest. These models are displayed in Table 4.

Table 4. Summary of regression models explaining variance in parent reported number of books read.

	Model 1: Demographics	Model 2: Parental orientations	Model 3: Child’s interest	Model 4: Combined
Intercept	4.29 ***	−0.55	1.88 ***	2.76 ***
Parent education level	0.64 ***
Parent age	−0.03 *
Parent gender (female)	0.50 ***			0.30 **
Child gender (female)
Child age	−0.40 ***			−0.25 ***
Children in family	−0.27 **			−0.22 ***
Regular reading routines		1.99 ***		0.86 ***
Usually read with child
Enjoy read with child		0.46 ***
Enjoy read alone		0.12
Child alone takes book (reverse coded)			0.18 ***	0.20 ***
Child doesn’t like being read to			−0.16 **
Number of times child requested to be read to			0.55 ***	0.47 ***
F	20.20	94.81	253.90	156.29
p	<.001	<.001	<.001	<.001
R^2	0.10	0.23	0.45	0.50

Note: ***p < 0.001, **p < 0.01, and *p < 0.05.

Almost all variables were statistically significantly correlated with the parent reported number of books read with the child in the last week (with the exception of the child’s gender and parents’ response to whether they usually read with their child), and many of them retained their significance when regressed on the outcome along with the other variables from the same classification. However, when all variables from the three classifications were regressed on the parent reported number of books read, the only variables to retain significance were parents’ gender (mothers read more to their child), child’s age (parents read more to younger children), number of children in the family (parents of smaller families read more to their child), having regular reading routines established (which has the largest positive impact on the number of books read), how regularly children look in a book their own (frequency is positively related to number of books read), and the number of times that the child requested to be read to (where more requests resulted in more books read with the child). In combination, these variables explained 50% of the variance in the parent reported number of books read with the child in the last week (R²= 0.50). Models 1–3 indicate that it is the child’s interest that has the greater explanatory power for the number of books reported read (R²= 0.45), followed by parents’ orientation (R²= 0.23), and lastly by demographic variables (R²= 0.10). These findings are summarized in Figure 1.

Figure 1. Relationship between demographic factors, parental orientations, child’s interest, and SBR.

Discussion

The variation in SBR observed in this study is large, with up to 33% of parents reporting reading no book or only one book with their child in the last week, and only 26% of parents reporting reading more than five books (Table 2). This could be understood as a low degree of conformity and social desirability bias in the survey data, but it can also be an expression of the value that book reading with children is given in Norwegian homes in 2022. Held together with 2016 PIRLS data (IEA, 2016) showing that parents do not retrospectively report a high frequency of book reading with their children, and studies showing a decline in parental book reading over the years (Bokhandlerforeningen, 2022), the variation in parent reported SBR in our study may imply that the reading culture is not as strong in Norwegian homes as we have believed it to be. More in-depth research should be conducted to test the validity of this conclusion.

The relationship between demographic factors and SBR

Our results (Table 4) show that there is an independent relationship between several demographic factors and the number of books parents report reading with their child. Parents’ gender, education level and age, the child’s age, and the number of children in the home are all significantly related to parent reported SBR. This suggests that younger mothers with higher education and less than three children in the home, are the typical profile of parents reporting that they read a lot (five or more books per week) with their child, a finding that corresponds with previous studies in Norway (Bokhandlerforeningen, 2022).

A body of research has found that socioeconomic status (SES) is a robust predictor of children’s achievement and skills. In the present study, it is therefore not surprising that parents’ education level is a significant contributor to the explained variance in parent reported SBR in the demographic model (Model 1). This finding is also aligned with the Family Investment Model (Conger & Donnellan, 2007; Vasilyeva et al., 2018), indicating that parents with high educational background will tend to prioritize activities that stimulate the academic and social development of their children. The effect of SES on children’s outcomes is a key educational and political concern, as this variation is difficult to mitigate. A very high proportion of children attend Norwegian ECEC institutions where SBR is emphasized in the Norwegian curriculum, but this does not replace SBR in the children's homes. Simply stated, children who participate in SBR at home spend more time on language and literacy activities than the children who do not participate in SBR at home, and when these children also come from higher SES families, the gap between those who have, and those who do not, increases. Interestingly, in our study we find that when motivational factors were included in the analysis (Model 4), parents’ education level (our proxy for SES) was no longer a significant predictor of parent reported SBR.

Child’s age was a significant contributor to the explained variance of parent reported SBR in both the demographic model (Model 1) and the combined model (Model 4). Therefore, parents reading more with younger children seems to be a robust finding. A possible explanation for this is that as children grow older their activity level increases, they become more independent, and they may exchange reading activities with books for other artifacts, such as digital devices. In addition, SBR is often performed as a dyadic activity, with an adult reading with one child. Held together with the impact of the number of children in the household (significant in both Models 1 and 4), a hypothesis could be that reading activities are more integrated in the family’s daily life when the parents have just one child. As children get older, they may be more likely to have siblings, and their parents may not have the capacity for the same amount of reading as before. Considering that reading and verbal interaction with teachers also seem to decrease with age in ECEC and in school (Lunde, 2012; Skaftun et al., 2021), the development of activities and programs to promote reading with four- and five-year-olds may be indicated by the findings in this study.

There was no consistent pattern across demographic groups regarding the reasons that parents gave for reading with their child. The majority of parents (N = 789) selected that they read “because it is cosy” – offered as the first alternative in the list of possible responses – which indicates that parents in general have a strong emotional and social motivation to read with their child. Although 73% of the sample selected at least one academic reason for reading with their child (promoting language stimulation and/or writing skills), there were no significant differences between the responses of parents with higher or lower education.

The answer to our first research question, therefore, is that demographic factors such as parent’s gender, age, and education level, the child’s age, and the number of children in the home, can explain some of the variation in how many books parents report reading with their children. However, when motivational variables are included in the analysis (Model 4) the impact of parent’s gender, child’s age, and the number of children in the home is reduced, and parents’ education level and age are no longer significant. The demographic variable model (Model 1) has less explanatory power than both the parental orientation model (Model 2) and child interest model (Model 3), implying that demographics may have less explanatory power for the parent reported SBR in Norway than first assumed.

The relationship between parental orientations, child’s interest, and SBR

In our second and third research questions we explored the degree to which parental orientations towards reading, and the child’s interest in reading, were related to the number of books parents reported reading with their child. Although our analysis found that parents’ interest for reading for themselves was not significantly associated with how many books they reported reading with their child, having regular routines for reading (which would naturally be established by parents) was found to be strongly associated with parent reported SBR. Parents that reported having regular routines for SBR, such as before sleep, after mealtimes, or during library visits, irrespective of whether they themselves were avid readers, generally reported reading more books with their children. This is to be expected, as reading routines provide regularity and predictability for both parents and children, eventually resulting in habits that become second nature within families.

Child’s interest, as captured by questions about when the child last looked in a book alone and the number of times the child requested to be read to in the last week, was also strongly associated to the number of books that parents reported reading with them. Whereas previous studies have found that participation in reading activities in ECEC have been dependent on the child’s language skills, and that children with low language skills tend to avoid such activities (Stangeland et al., 2018), the current study adds child’s initiative as an important factor for how many books they are being read at home. This suggests that children who are already interested in books and ask parents to read to them, will be read to more than children who don’t have the same motivation. Motivation for reading starts with curiosity and the enjoyment of books, and a collaboration between ECEC and parents to awaken this in children who don’t ask to be read to, might be a key to motivating children’s interest in reading, and may result in increased SBR at home. In addition, the findings of this study suggest that policymakers’ emphasizing the importance of established reading routines may be a practical approach to encouraging more reading at home and in ECEC.

Whereas we find a significant association between demographic factors and parent reported SBR in this study, the greatest proportion of explained variance in parent reported SBR is associated with the reported level of the child’s interest in reading, followed by parental orientations towards reading. Therefore, despite earlier research suggesting the presence of a “Matthew Effect” in home reading activities (Bakermans-Kranenburg et al., 2005), where higher SES families tended to benefit more from home language interventions, our findings suggest that an increased interest from children in reading, which could be stimulated through established reading routines, may partially compensate for the effect of socioeconomic factors in the early acquisition of language skills.

Study limitations

This study is based on self-reported survey data, and as such has several limitations. Despite the size of the sample, and our confidence that the study design ensured a representative sample of respondents, the survey was conducted within the Norwegian context and is therefore only moderately generalizable without validation in other Nordic countries. In addition, due to the nature of survey data, the findings are correlational and should not, therefore, be interpreted in a causal manner.

It is also important to emphasize that SBR is about quality as well as quantity. Quality is linked to the interactive nature of shared reading experiences, and is, for example, about the child being an active participant and not a passive listener during reading (Whitehurst et al., 1988). Some of these qualitative aspects were explored in this survey (such as the interaction of parents and children with the multisensorial aspects of reading) the results of which have already been published (Kucirkova et al., 2023). However, because the present study is based on characteristics typically measured with numeric variables, we focus here exclusively on quantity.

It is also possible that the differences in reading practices vary more than we have been able to identify between the youngest (3-year-olds) and oldest (6-year-olds) in this study. As children’s reading interests vary between 3–6 years, with younger children possibly to read one book several times, whereas older children may want to read multiple different books, we intentionally did not ask how many different books the parents read with their child last week. There is however a chance that parents who read one book multiple times, did not report this, and this is a possible limitation of this study.

Finally, we expected that the socio-cultural significance of SBR in the Norwegian context, and the framing of the survey (which may have primed respondents towards positive responses), may have generated socially desirable answers in response to the survey questions. Therefore, despite a lower degree of parent reported book reading than anticipated, it is still possible that these responses are more elevated than reality.

Future studies could validate our findings and deepen knowledge about SBR in Nordic homes through replication of this study in other countries, the inclusion of questions regarding different types of home-based reading activities (such as listening to audiobooks and digital books), and the exploration of home-based SBR through different research designs (such as observational or experimental studies). In addition, a future study that explored latent variables and factors that may be causally linked to the frequency of SBR could make an interesting contribution to the literature in the field. However, the objective of the study was to describe the relationship that different variables have with parent reported SBR within a national sample and identify which individual variables together explain a significant portion of the variance in our outcome. Due to the self-reported nature of the data, we believe this approach is sufficiently methodologically ambitious, but more ambitious designs could be considered with more reliable data (for example observational or experimental studies).

Conclusion and implications

The findings from this study contribute new knowledge on reading practices in families living in Norway, and these have implications for research, ECEC practice, and for policy makers within the field of education. From an historical point of view, where the value of children’s literature and shared book reading has been central in the Norwegian culture, findings from this study suggest that the culture for reading with young children in Norway may be weaker than we have believed it to be. In future reading and language interventions involving parents, researchers should consider that there is great variation in how much parents report reading with their child, and that SBR may decrease with the child’s age. From a language learning perspective, where a rich literacy environment, including access to books, is associated with language growth and later advantage in reading (Lervåg & Aukrust, 2010), our results suggest a learning gap between children who are interested in books and ask their parents to read with them, and children who don’t. This is an alarming finding, as low interest in literacy activities is associated with poor language skills (Stangeland et al., 2018). However, child interest is something that can be encouraged and awakened through the pedagogical practice in ECEC, especially when advising parents. Our findings suggest that older children in ECEC should be targeted for reading activities, and that a renewed focus on reading in ECEC could positively impact on children’s interest in reading and subsequently on the number of books that their parents read with them.

Having reading routines is strongly related to the number of books that parents report reading with their children, regardless of parental educational level, and this is also a promising finding. There is the potential for narrowing the gap in language learning opportunities by establishing routines that result in frequent book reading. Held together with previous research showing that parental enjoyment for reading can compensate for lack of higher education (Støle et al., 2021), our results suggest that encouraging children’s interest in book reading, and establishing fixed reading routines at home, might be keys for increasing shared book reading with young children, within families in Norway.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Research Council of Norway [grant numbers 319918, 314117, 275576].