Abstract
The desire to understand the home advantage has spurred a large body of research describing the contribution of differences in family background and parenting styles to school readiness and the achievement gap. Using the National Household Education Survey at two time points provides a fuller picture of the trends in parenting and cognitive activities with young children before the onset of formal schooling. The results presented here show that although more educated parents participate more in cognitive activities with their young children in 1991 and 2001, participation is increasing for all parents. This is true for a widely expected parental activity, reading to your child, and for a direct instruction activity, teaching letters, words, and numbers. It is also true for the more creative activities of music, arts and crafts, and telling stories. Contrary to images of some families as deficient and others as overly exuberant, all families in this sample are increasing engagement with their young children in a range of cognitive activities aimed at school readiness because parents as well as children get socialized to the culture of schooling. Modern American schooling requires parents to be engaged in the education process of their children and attempt to create advantage for them.
Research on differences in school outcomes often investigates what has come to be called the home advantage (Lareau, 1989) or the differences in family resources and childhood experiences that give some children an advantage in schooling and also “school readiness.” Here, I look at parent cognitive engagement with young children before the onset of formal schooling as one way in which modern American parents try to create advantage for their children. Although this research investigates the cognitive component of school readiness, school readiness is a broad and sometimes controversial term.
School readiness for young children
Horace Mann (1796–1859) once said, “Education then, beyond all other devices of human origin, is the great equalizer of the conditions of men, the balance-wheel of the social machinery.” He believed that the common school movement would create equality of opportunity for all American citizens. However, by the mid-20th century, it was obvious that poor children needed a “head start.” Access to free K to 12 education alone was not enough to remedy the broad inequalities in society. Some children have an advantage in schooling because their early experiences more successfully prepare them for the challenges of formal schooling; regardless of the actual cognitive benefits, schools reward particular skills often categorized as “school readiness.” This is of particular concern because we know that young children spend their time in a variety of activities, and these activities vary both within and between national cultures (Tudge, 2008). Furthermore, only some of these activities build what are currently considered school readiness skills. Research has also found that a mismatch between the learning cultures of home and school results in the exclusion of some children in early childhood education classrooms (Brooker, 2002) because the curriculum is better aligned with some types of early childhood experiences. This has led to broad concerns about schooling and its alignment with dominant middle-class culture.
Even the term “school readiness” is somewhat controversial. It implies that there is an agreed upon set of behaviors and skills needed for school success (see Lewit and Baker, 1995, for a fuller discussion). However, surveys have shown that teachers and parents differ on their opinions about school readiness. Sixty two percent of parents report that teaching children to share is essential to school readiness, but the majority also report that it is essential to teach children the alphabet and numbers, and more than 40 percent also feel that it is essential to teach children reading and how to hold a pencil before formal school entry (O’Donnell, 2008). 1 In comparison, teachers give higher importance to behavioral skills over cognitive skills. For example, teachers overwhelmingly want students to be healthy; well rested and nourished; and able to communicate their needs, wants, and thoughts, as well as show enthusiasm and curiosity (Heaviside and Farris, 1993). They also place significant importance on sharing, attentiveness, self-control, and sensitivity to others. Of least importance are specific cognitive skills such as knowing the alphabet and numbers.
However, some children have an advantage in schooling and that advantage is associated with specific types of knowledge and skills that are often part of the construct “school readiness.” For example, literature from the Early Childhood Longitudinal Study—Kindergarten shows that children who start kindergarten with greater cognitive knowledge and skills, who are read to frequently, and who have more positive approaches to learning have an academic advantage over children who do not start kindergarten with these resources (US Department of Education, 2002). Furthermore, the evidence suggests that certain types of school readiness activities are associated with family background. For example, household income is positively related to the quality of cognitive stimulation in the home environment of young children (Votruba-Drzal, 2003) resulting in economically disadvantaged children being less ready for school than their middle-income peers (Stipek and Ryan, 1997) and black and Hispanic children being less ready for kindergarten than white children (Rock and Stenner, 2005).
In addition to family background, parenting style is also associated with school readiness. Research shows that cognitively stimulating and patient/nurturant parenting styles are associated with cognitive and social school readiness (McGroder, 2000), as are maternal achievement expectations, quality of mother–child interaction during early childhood (Hess et al., 1984), maternal sensitivity (Brooks-Gunn et al., 2002), the availability of play materials in the home (Bradley and Caldwell, 1984), and early language development (Forget-Dubois et al., 2009). Other research has found that parenting styles mediate the relationship between family background and school readiness. Parents who speak to their children frequently; use a variety of words; and a depth of nouns, modifiers, and tenses are more likely to have children with large vocabularies, and these children perform better on tests of achievement/intelligence at age 5 and 9. Parent linguistic style is intimately connected to family socioeconomic status and has long-term implications for school success (Hart and Risley, 1995).
The last several decades have brought a heightened public awareness that early childhood experiences influence school performance. School readiness is not determined solely by individual abilities; rather, it is enhanced by people and environments. The knowledge that early childhood experiences, largely in the form of parenting, contribute to school readiness has resulted in public policy that encourages parent behaviors aimed at school readiness. For example, school readiness is the first goal in Goals 2000. 2 Objective 2 of Goal 1 reads “Every parent in the U.S. will be a child’s first teacher and devote time each day to helping such parents’ preschool child learn, and parents will have access to the training and support parents need” (103 Congress, 1994). But initiatives have gone further than posing school readiness as a goal; they have provided parents with guidelines for the preparation of their children for entering school through the pursuit of cognitive development. The America Reads Challenge, for instance, promulgated by President Clinton made “Ready, Set, Read” activity kits available to all families and caregivers of children aged 5 and below (Koralek et al., 1997). And, the second Bush administration made guidebooks available for parents and families such as Helping Your Preschool Child that include recommendations for encouraging cognitive development in young children (US Department of Education, 2002 [1993]). But this type of public policy has had limited impact. Research consistently shows that school readiness is intimately tied to family background and parenting style thus leading to broad concerns about school readiness as a construct.
Theoretical underpinnings and research questions
Because we know that some children have an advantage in schooling and that the advantage is associated with family background and parenting style, school readiness is sometimes dismissed as merely a component of middle-class culture. Therefore, research that looks at families at one time point often portrays some families as doing “this,” others “that,” each equally legitimate or perhaps some as overly exuberant or deficient, but this portrayal neglects the possibility that parents might be similar in their pursuits with children. Education, as a powerful institution, encourages the patterns of parenting to move in the same direction. I argue that the cause of this is the centrality of education as an institution in modern society. This article uses a neo-institutional framework in which to explore trends in parenting of young children in the context of schooling (e.g. Meyer, 1977).
Institutions and individuals create social meaning. Education, as a powerful institution in modern society, shapes society by classifying people into socially constructed and institutionally embedded categories with distinct social status. This includes not only the legitimated process of social stratification but also a host of other constructions about children, schooling, and parenting. In other words, schooling creates new roles for modern society in a number of expanded ways over time. As adult status became increasingly tied to educational attainment over the 20th century (Hout, 1988), families increasingly engaged in the schooling process and early childhood was increasingly incorporated into the purview of schooling. For example, family media such as parenting magazines increasingly included cognitive development as a central topic (Wrigley, 1989), and school reform efforts increasingly included attempts to draw parents into the schooling process (Domina, 2005). This involvement typically begins before formal schooling (Schaub, 2010). In addition to parenting, schooling also began encroaching on family time as demonstrated by average homework time (Bennett and Kalish, 2006; Gill and Schlossman, 2003), especially for our youngest pupils or those 6–8 years old (Hofferth and Sandberg, 2001a; Timmer et al., 1985). As larger proportions of the population are incorporated into schooling, the role in schooling for the typical American family increasingly stresses creating individual advantage and highlighting individual successes.
However, we have only a sketchy picture of what kinds of cognitive and school readiness activities parents do with young children or what proportion of parents is engaging in these activities at the turn of the 21st century. School readiness is one little explored means by which parents attempt to create advantage for their children. Here, I explore parental engagement in cognitive activities with young children before the onset of formal schooling as one means by which parents encourage school readiness. Do family background characteristics dictate the types of activities parents choose to engage in with their young children? Over the second half of the 20th century, parents increasingly spent more time engaged in cognitive activities with their preschool and kindergarten age children (Schaub, 2010), but which types of activities do parents pursue?
Methods
Data and sample
The National Household Education Surveys (NCES) provide data that are well suited to addressing parent engagement in cognitive activity with their young children at the end of the 20th century. Collected by the National Center for Education Statistics (NCES), the NHES data sets look at a wide range of education issues, and they are the first large-scale attempt by NCES to develop a household survey instead of school-based data collection. And importantly, these surveys span a decade of increasing concern and programs about school readiness on the part of local and national policymakers. NCES is mandated to collect and report information on the condition of education in the United States, but it has primarily collected data from teachers, students, schools, school districts, and state education agencies; household data collection has been limited.
The data are drawn from nationally representative samples of noninstitutionalized civilian households. 3 These complex data sets screen a large number of households to gather background information and determine eligibility for participation in one or more of the components. 4 Each survey has at least two components, but I use only the parent interview component that contains information on early childhood education. I look specifically at the parents of children who are 3–5 years of age. Interviews were carried out with the parent or guardian most knowledgeable about the child’s education.
The parent interview of young children in the NHES is designed to collect information on children’s experiences in a wide range of settings, including their homes, childcare arrangements, early childhood education programs, and schools. In addition, the range of experiences covered in the survey includes cognitive activities, developmental activities, and entertainment activities. I am particularly interested in cognitive activities parents engage in with young children. My analysis sample includes only 3- to 5-year-olds but before kindergarten enrollment.
Data used here from the NHES are weighted using the final child weight. A two-stage weighting procedure was used with the first stage of weighting associated with the probability of sampling telephone numbers or households; adjustments are associated with particular years, but one common to all waves of data is the adjustment for oversampling of high minority concentration areas. The second stage of weighting, the person weight, is associated with each extended interview. Creating these weights involved several steps and results in a representative sample of all children eligible for each survey (Nolin et al., 1997).
Measures
Outcomes
The data contain consistent dependent variables across the years. This analysis uses five measures of parent engagement in cognitive activities with young children. Four of these variables ask a yes/no question that, if answered yes, is followed up by an additional opportunity to respond 1–2 times or 3 or more times per week. They include “In the past week has anyone in your family done the following things with (child)?”: (1) told a story; (2) taught letters, words or numbers (LWN); (3) taught songs or music; and (4) worked on arts and crafts. In addition, parents were asked “How many times have you or someone in your family read to (child) in the past week? Would you say ... not at all, once or twice, 3 or more times or every day?”
Predictors
I investigate changes in parent engagement in various types of cognitive activities with young children through mothers’ education because past research has documented that mothers spend more time caring for children (Ahmeduzzaman and Roopnarine, 1992; Aldous et al., 1998; Bryant and Zick, 1996; Hofferth and Sandberg, 2001; Pleck, 1997) and that mothers’ educational attainment is positively associated with the frequency of mother–child interaction (Hill and Stafford, 1980; Zick and Bryant, 1996), the quality of mother–child interaction (Clarke-Stewart et al., 1994) and enriching activities (Hill and Stafford, 1974, 1985; Leibowitz, 1974, 1977).
In addition to mothers’ education, the NHES contains several other consistent independent variables across years. They include household income, number of siblings in the household, school enrollment, child age, child gender, child race/ethnicity, mothers’ employment status, and number of parents in the household (see appendices 1 & 2 for means and standard deviations).
Analytic approach
The logistic regression model predicts the probability of each dependent dichotomous variable (read, LWN, music, story, and craft) occurring separately over time. For each of the five activities, less and more educated parents are compared at two time points. This allows the patterns of parent behaviors to emerge and permits the similarities as well as differences between parents to be explored. Logistic Regression requires me to dichotomize the dependent variables. I dichotomized story, music, LWN, and craft variables at “not at all” and “1 to 2 times last week” or “three or more times last week.” Read is dichotomized at “not at all,” “1 to 2 times last week,” and “three or more times last week” or “every day.” The dichotomizing cutoff decisions are based on the distributions of the dependent variables. They provide reasonable distributions, such that in some instances, approximately 50 percent of the population falls above and below the cutoff of three or more times per week. However, it also gives instances where higher percentages fall either above or below the cutoff mark. This in effect provides normal distribution of the frequency distributions. The Logistic Regression model allows me to examine each of the parent behaviors separately (see Appendices 3 to 7). Probabilities are then displayed in graphs over time, so that the change from 1991 to 2001 is presented visually.
Results
Table 1 shows the percentages and frequencies for each dependent variable. Each dependent variable increases from 1991 to 2001. In 1991, 60 percent the of parents were teaching their young child letters, words, and numbers three or more times per week; this increased by 14 percent by 2001. Each of the other variables starts off lower 5 and increases between 11 and 16 percent between 1991 and 2001. These frequencies imply that a significant percentage of parents see frequent, active engagement with their young children in cognitive activities as part of the routine parenting role.
Frequencies and percentages of parental engagement in cognitive activities with young children.
Table 2 shows the change in probability for each dependent variable by mother’s education. Interestingly, the patterns are not identical across all five dependent variables, rather, three distinct patterns emerge. I consider each dependent variable separately.
Change in probabilities in each of the dependent variables according to the level of education of the mother, 1991–2001.
HS: high school.
Reading to your child
It is widely agreed upon in the early childhood literature that reading to your child is an important activity that contributes to the cognitive development and reading readiness of young children. Research shows that reading to your child builds reading skills (Saracho, 1997; Snow et al., 1998) as well as stimulates imagination, builds vocabulary, and introduces the parts of story (Moss and Fawcett, 1995). Frequencies and percentages in Table 1 indicate that the percentage of parents responding that they read to their child every day last week increased from 42 to 58 percent from 1991 to 2001. Table 2 shows that within each category of mothers’ education, the probability of reading to your child every day increased from 1991 to 2001. Figure 1 visually displays the change. Controlling for a number of family background characteristics, the probability of reading to your child every day increased at a similar rate for mothers with a high school diploma or less and a bachelor’s degree or more. Families with more educated mothers remained more likely to read to their child every day than families with less educated mothers, but families within all categories of mothers’ education increased from 1991 to 2001 and the rate of change was remarkably similar.

Probability estimated to read every day to her child by level of education in 1991 and 2001.
Letters, words, and numbers (LWN)
Teaching letters, words, and numbers is direct cognitive instruction by parents to children. In teaching letters, words, and numbers, parents impart specific skills needed for literacy and numeracy. Table 1 shows that the percentage of parents responding they taught letters, words, and numbers three or more times last week increased from 60 to 74 percent from 1991 to 2001. Table 2 shows that within each category of mothers’ education, the probability of teaching letters, words, and numbers to your child three or more times last week increased from 1991 to 2001. Figure 2 visually displays the change. Controlling for a number of family background characteristics, the probability of teaching letters, words, and numbers to your child three or more times last week increased at a slightly different rate for families with mothers with a high school diploma or less and a bachelor’s degree or more. In 1991, the probabilities across families with varying levels of mothers’ education were similar, but in 2001, families with more educated mothers emerged more likely to do direct teaching than families with less educated mothers. Families in all categories of mothers’ education increased from 1991 to 2001 although the rate of change was different across levels of education.

Probability estimated to teach letters, words, and numbers to her child by level of education in 1991 and 2001.
Songs and music
Songs and music are widely incorporated into early childhood education as an important component of building early cognitive skills. Singing encourages phonological awareness (Bryant et al., 1990; MacLean et al., 1987) and encourages language and rhythm awareness (Moss and Fawcett, 1995). Table 1 shows that the percentage of parents responding they engaged in music activities with their young child three or more times last week rose from 39 to 54 percent from 1991 to 2001. Table 2 shows that within each category of mothers’ education, the probability of engaging in music activities with your child three or more times last week increased from 1991 to 2001. Figure 3 visually displays the change. Controlling for a number of family background characteristics, the probability of engaging in music activities with your child three or more times last week increased at a slightly different rate for mothers with a high school diploma or less and a bachelor’s degree or more. In 1991, families with more educated mothers were more likely to engage in music activities, but by 2001, the gap had closed. Families in all categories of mothers’ education increased from 1991 to 2001 although the rate of change was different across levels of education.

Probability estimated to engage their children in music activities by level of education in 1991 and 2001.
Telling stories
Telling stories is a worldwide cultural tradition. It has been linked to better reading skills (Glazer, 1989; Sonnenschein et al., 1996) because it stimulates the imagination, builds vocabulary, and introduces the parts of story (Moss and Fawcett, 1995). Table 1 shows that parent reports of engaging in storytelling with their young children rose from 39 to 54 percent from 1991 to 2001. Table 2 shows that within each category of mothers’ education, the probability of engaging in storytelling with your child three or more times last week increased from 1991 to 2001. Figure 4 visually displays the change. Controlling for a number of family background characteristics, the probability of telling stories to your child three or more times last week increased at a slightly different rate for mothers with a high school diploma or less and a bachelor’s degree or more. In 1991, families with more educated mothers were slightly more likely to engage in storytelling with their young child three or more times last week, but by 2001, the gap had grown. Families in all categories of mothers’ education increased from 1991 to 2001 although the rate of change was different across levels of education.

Probability estimated to tell stories to her child by level of education in 1991 and 2001.
Arts and crafts
Arts and crafts are creative activities crucial to the development of young children that build fine motor skills, improve math learning, and support discovery (Armistead, 1996; Baker, 1992). Table 1 shows that the percentage of parents reporting they engaged in arts and crafts activities with their young child three or more times last week rose from 35 to 46 percent from 1991 to 2001. Table 2 shows that within each category of mothers’ education, the probability of engaging in arts and crafts activities with your young child three or more times last week increased from 1991 to 2001. Figure 5 visually displays the change. Controlling for a number of family background characteristics, the probability of engaging in arts and crafts activities with your young child three or more times last week increased at a similar rate for mothers with a high school diploma or less and a bachelor’s degree or more. Families with more educated mothers remained more likely to engage in arts and crafts activities with their young child three or more times last week than less educated mothers, but families in all categories of mothers’ education increased from 1991 to 2001, and the rate of change was remarkably similar.

Probability estimated to engage their children in arts and craft activities by level of education in 1991 and 2001.
Summary of findings
These five separate activities that parents might engage in with their young child show some interesting trends. First, American parents appear to be highly engaged in cognitive activities with their young children. In 1991, a large proportion of all parents were engaging in the cognitive activities explored here. And from 1991 to 2001, all parents increasingly responded yes to engaging their young child in each of the five activities. In addition, within all levels of education, the probabilities increased from 1991 to 2001. Second, the rates of change from 1991 to 2001 vary, such that the gap remains the same in two of the activities, increases in two of the activities, and decreases in one of the activities. 6 But finally, in general, all families, regardless of the education level of the mother, are increasing engagement in cognitive activities with young children at roughly similar rates. This includes the widely expected parent activity, reading to your child, as well as a direct instruction activity, teaching letters, words, and numbers. But it also includes creative activities such as music, arts and crafts, and telling stories.
Discussion
The initial idealistic intentions of public education as the “great equalizer” have slowly been replaced as a slow wave of research consistently shows that the achievement gap appears before the onset of formal schooling and persists or even grows during the schooling years. We now know that children come to school with varying experiences that give some an advantage in formal schooling. Research has shown that this advantage in schooling is linked to family background and parenting style. But it is less clear how that advantage is created. As a result, popular images often depict less educated and low-income parents as deficient or more educated and middle/high-income parents as overly exuberant.
The comprehensive nature of the American education system encourages parents to be highly engaged and attempt to create advantage in their children’s education. Most depictions of parental involvement in schooling illustrate the differences between middle-class and working-class parents of school-aged children, but the addition of time adds a new dimension to our knowledge of families. Here, I argue that parent engagement in cognitive activities with young children before the onset of formal schooling is a form of parental involvement aimed at enhancing school readiness and therefore creating advantage. Parents increasingly engaged in cognitive activities with their young children over the second half of the 20th century (Schaub, 2010); these data show that the trends by mothers’ education vary with activity, but engagement in all the activities investigated here increases from 1991 to 2001 regardless of mothers’ education level.
Families with mothers in both education categories engage in the cognitive activities analyzed here. In addition, engagement increases for both education categories for all activities. 7 Three patterns emerge from the five activities analyzed. Families with the most educated mothers are always the most likely to engage in each activity in 1991, and they maintain that advantage in 2001 in all activities except music. But all families, regardless of mothers’ education category, are increasing their engagement in all of the activities investigated here.
The evidence suggests individuals get socialized to the culture of schooling. Modern American schooling requires parents to be engaged in the education process of their children and attempt to create advantage for them. Education is a powerful institution. As behaviors become part of the expected parenting role (i.e. institutionalized), parents attempt to create advantage for their children through new behaviors, it is generally more educated mothers’ who acquire these behaviors first. But all parents are engaging in cognitive activities with their young children, and the proportion of parents engaging is increasing in all education categories. Contrary to images of some families as deficient or overly exuberant, all families are increasing engagement of their young children in a range of cognitive activities aimed at school readiness.
One new component of the ‘good parent’ role is cognitive developer, and this new component will continue to impinge upon the time of parents. As school readiness expectations increasingly shift from socioemotional to skill based, it is likely that parents will add new activities that will build skills under the school readiness construct. More educated mothers will gain these activities first, but they will eventually become normative behavior.
Footnotes
Appendix
Logistic regression of arts and crafts activities three or more times last week on mothers’ education.
| 1991
a
|
2001
b
|
|||||
|---|---|---|---|---|---|---|
| B | SE | Exp(B) | B | SE | Exp(B) | |
| Mothers’ education | .208*** | .045 | 1.231 | .208*** | .052 | 1.232 |
| Total household income | .045* | .021 | 1.046 | .031 | .023 | 1.032 |
| Number of siblings in the household | .071** | .026 | 1.074 | .022 | .035 | 1.022 |
| Mothers’ employment | ||||||
| Full-time | −.137 | .076 | .872 | −.211* | .090 | .809 |
| Part-time | .009 | .081 | 1.009 | −.065 | .104 | .937 |
| Looking for work | −.238 | .140 | .788 | −.160 | .188 | .852 |
| Child’s age | −.042 | .046 | .959 | .090 | .059 | 1.095 |
| Child’s gender | −.107 | .061 | .898 | −.340*** | .074 | .712 |
| Race/ethnicity | ||||||
| Black | −.317** | .104 | .729 | −.416*** | .123 | .659 |
| Hispanic | .077 | .098 | 1.080 | −.135 | .105 | .874 |
| Other | −.161 | .164 | .851 | −.123 | .165 | .884 |
| Enrollment in school | .010 | .066 | 1.010 | .017 | .081 | 1.017 |
| Family composition | .040 | .089 | 1.041 | .069 | .093 | 1.072 |
Source: National Household Education Survey, National Center for Education Statistics, 1991 and National Household Education Survey, National Center for Education Statistics, 2001.
SE: standard error.
Constant: −0.772; Nagelkerke R2: .026.
Constant: −0.583; Nagelkerke R2: .042.
P < .05, **P < .01, ***P < .001.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
