Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Aug 17.
Published in final edited form as: J Res Adolesc. 2019 Feb 16;30(Suppl 2):298–314. doi: 10.1111/jora.12477

Child, Mother, Father, and Teacher Beliefs about Child Academic Competence: Predicting Math and Reading Performance in European American Adolescents

Diane L Putnick 1, Chun-Shin Hahn 1, Charlene Hendricks 1, Joan T D Suwalsky 1, Marc H Bornstein 1
PMCID: PMC6697633  NIHMSID: NIHMS1042529  PMID: 30771240

Abstract

Beliefs about child competence in math and reading have important implications for academic performance in adolescence. However, it is unclear whether children’s own beliefs are the most important predictor of their academic performance or whether parents’ and teachers’ beliefs about child competence influence child academic performance. We assessed mothers’, fathers’, teachers’, and children’s beliefs about European American children’s (N = 189) competence in math and reading at age 10 and children’s math and language performance at ages 10, 13, and 18 years. Confirmatory factor models demonstrated that children’s and teachers’ beliefs had lower loadings on a latent variable of child competence in math and reading than mothers’ beliefs. Children’s self-competence beliefs in math and reading were not significantly correlated, suggesting children may use dimensional comparisons when assessing their own competence. Mothers’, fathers’ and teachers’ assessments of child competence in math were strongly correlated with their assessments of child competence in reading. Controlling for stability in academic performance, family SES, and other reporters, mothers and fathers who rated their children’s math competence higher had adolescents who performed better in math, and fathers who rated their children’s reading competence higher had adolescents who performed better in language tasks. However, children who rated their own competence higher in math and reading had lower math and language (for girls only) performance in adolescence. European American children may use dimensional comparisons that render them poorer judges of their math and reading competence than parents.

Keywords: competence beliefs, academic performance, reporter

Competence beliefs are perceptions of ability in a particular domain. Children hold such beliefs about their own self-competence (e.g., I am good at math), and parents and teachers also hold beliefs about child competence (e.g., this child is a strong reader). The expectancy-value model of achievement motivation (Wigfield & Eccles, 2000, 2002) considers multiple distal (e.g., cultural milieu, affective memories) and proximal (e.g., expectancies for success, values) influences on children’s achievement-related choices. In this model, competence beliefs (or self-concepts of ability) are theorized to influence achievement behavior through children’s expectations for success. However, several previous studies have found that competence beliefs and expectancies for success form a single construct (Eccles at al., 1993; Wigfield et al., 1997). Following previous work, here we define competence beliefs broadly to encompass beliefs about child ability, expectancies for success, and persistence.

Children’s beliefs about their own efficacy provide a foundation for motivation and learning across domains (Zimmerman, 2000). Children’s beliefs about their competence in academic subjects have important implications for the effort they put into mastering those subjects and for their later performance (Dweck, 2002; Hansford & Hattie, 1982; Marsh, 1990). However, children’s beliefs are not constructed alone. Children receive input from parents and teachers as well as feedback from their own academic experiences (Altermatt, Pomerantz, Ruble, Frey, & Greulich, 2002; Parsons, Adler, & Kaczala, 1982; Parsons, Kaczala, & Meece, 1982). Children may also hold systematic biases about their own competence (e.g., Entwisle & Baker, 1983; Möller & Marsh, 2013). Different reporters of children’s competence have medium to large agreement with one another (Authors, under review; Tiedemann, 2000; Wigfield et al., 1997), but it is unclear which reporters (child, mother, father, or teacher) have the greatest predictive validity for children’s later academic performance and whether reporters’ beliefs simply reflect concurrent child performance, or if competence beliefs predict child academic performance in adolescence controlling for concurrent performance.

Child Beliefs About Academic Competence

According to the expectancy-value model of achievement motivation (Wigfield & Eccles, 2000, 2002), children’s self-competence beliefs relate to their achievement-oriented choices and performance (e.g., Musu-Gillette, Wigfield, Harring, & Eccles, 2015; Simpkins, Davis-Kean, & Eccles, 2006; Trautwein, Lüdtke, Roberts, Schnyder, & Niggli, 2009). Children choose to engage in activities they feel competent doing. Specifically, European American children’s beliefs about their competence in academic subjects relate to their performance in those subjects (Marsh, 1990; Obach, 2003; Richardson, Abraham, & Bond, 2012; Wigfield & Eccles, 2000, 2002). For example, children’s competence beliefs in math and reading are strong predictors of their later performance, even when controlling for their earlier performance (Wigfield & Eccles, 2000, 2002). Meece, Wigfield, and Eccles (1990) found that middle-class European American children’s competence beliefs in math related to better school grades through children’s expectancies for success, but these findings do not always hold for other ethnic groups (Graham, 1994; Okagaki, 2006; Stevenson, Chen, & Uttal, 1990). In longitudinal, nationally representative samples of middle- and high-school students, Marsh (1990) and his colleagues (2005) found that having a higher academic self-concept predicted better school grades and test scores, but grades and test scores did not predict corresponding increases in self-concept (or did so to a very small degree). Still, children’s beliefs about their own academic competence may not be the only (or most influential) set of beliefs. Parents and teachers also hold beliefs about child competence that may inflect children’s development.

Parent and Teacher Beliefs About Child Academic Competence

Beyond children’s beliefs about their own competence, the competence beliefs of “socializers” (e.g., parents and teachers) relate to children’s competence beliefs (e.g., Frome & Eccles, 1998; Phillips, 1987; Tiedemann, 2000; Upadyaya & Eccles, 2015). Children’s own competence beliefs relate moderately to highly with parents’ and teachers’ reports of their competence (Authors, under review; Tiedemann, 2000; Wigfield et al., 1997), but it is unclear whether one reporter’s beliefs about child competence is more important for children’s future academic performance than another’s. The expectancy-value model of achievement suggests that a child’s own competence beliefs would be the best predictor of academic performance. In the expectancy-value model, socializers’ beliefs predict children’s beliefs, which in turn predict children’s performance; socializers’ beliefs do not directly relate to child performance. However, children may not be the best judges of their own competence. Newman (1984) and Nicholls (1979) suggested that, around 10 years of age (the youngest age studied here), correlations between self-perceptions and ability fall into the .40 range and the correlations are stable as children age. Davis-Kean (2005) learned that parents’ educational expectations for their child are directly related to child performance (as well as indirectly through parental behavior) in European American families but not African American families. Perhaps parents’ beliefs or teachers’ beliefs are more predictive of child performance than children’s own beliefs.

Parents may be good judges of their child’s academic competence because they have watched the child’s development from the beginning, they see the child in various settings, and they receive regular reports from school. Maternal competence beliefs about children’s math and reading are significant independent predictors of children’s math and reading performance, respectively, even after controlling for children’s own competence beliefs (Entwisel & Baker, 1983; Halle, Kurtz-Costes, & Mahoney, 1997; Herbert & Stipek, 2005; Jacobs, Davis-Kean, Bleeker, Eccles, & Malanchuk, 2005). We found no comparable studies of paternal competence beliefs as independent predictors of child performance. We study both here.

Teachers observe children’s performance in the classroom and on tests. Teachers may be particularly good judges of child academic competence because they are highly knowledgeable and experienced about child abilities at a particular age/stage of development, and they have a ready pool of same-aged peers with whom to compare the child. Jussim and Harber (2005) noted that teachers’ expectations for performance relate to children’s academic performance, but the relations are generally small (see also Boerma, Mol, & Jolles, 2015). Herbert and Stipek (2005) found that teachers’ reports of child math and literacy competence correlated with low-income children’s achievement, but they did not control for child or parent reports to determine whether teachers have a unique perspective on child competence.

Do Competence Beliefs Only Reflect Concurrent Performance or Are They Predictive?

It is also unclear whether different reporters’ beliefs about child competence predict child performance later in development in equal or different degrees. For example, if children, parents, and teachers have strong beliefs about a child’s math competence, perhaps these beliefs encourage the child to work harder and increase performance over time. The link between children’s own competence beliefs and change in academic performance has been well documented (Marsh, 1990; Marsh et al, 2005; Wigfield & Eccles, 2000, 2002). Pomerantz and Dong (2006) found that European American mothers’ beliefs in their children’s academic competence predicted an increase in children’s school grades over time, suggesting that mothers’ competence beliefs were not simply a reflection of concurrent performance, but anticipated later change in child academic performance. No studies of how fathers’ competence beliefs predicted later child performance were found, but we would expect similar relations to those of mothers. There is also evidence that teachers’ beliefs about child competence in math and reading relate to changes in children’s own competence beliefs (Upadyaya & Eccles, 2015), but it is unclear whether those changes translate into better child math and reading performance. Tiedemann (2000) studied the math competence beliefs and performance of 3rd and 4th grade German students. Teachers’ beliefs about child competence predicted an increase in children’s math grades, but child, mother, and father competence beliefs did not. It should be noted, however, that teachers were the ones assigning the grades, so shared source variance may have explained this relation. Rubie-Davies and colleagues (2014) found that teachers’ expectations for child performance had enduring indirect effects on later child performance through concurrent performance, but not independent of concurrent performance. In the present study, we explore how different reporters’ unique perceptions of child competence relate to concurrent performance in middle childhood, and how unique perceptions of child competence predict later adolescent academic performance (measured by objective tests) controlling for concurrent academic performance.

Sources of Error in Competence Beliefs

Although competence beliefs may be predictive, they may also be incorrect and/or systematically biased. Dimensional comparison theory (Möller & Marsh, 2013) and the internal/external frame of reference model (Marsh, 1986) posit that one method children use to construct their self-beliefs is (internally) comparing their own performance in contrasting domains (see also Pinxten et al., 2015). For example, children who perform well in math may perceive themselves to be less competent in reading. Consequently, children’s competence ratings in math and reading are often unrelated or negatively related, whereas their achievement scores in math and reading are positively correlated (Marsh et al., 2015; Möller, Pohlmann, Köller, & Marsh, 2009). As Möller and Marsh (2013) pointed out, dimensional comparisons “lower the self-concept in the worse off domain while raising it in the better off domain” (p. 546). Hence, dimensional comparisons may lead relatively higher-performing children to overestimate their competence and relatively lower-performing children to underestimate their competence. When objective measures of competence and performance are covaried, there could be no relation or inverse relations between competence beliefs and later performance. Dimensional comparison theory has been shown to apply to children’s internal self-beliefs, but it is unclear whether other reporters use dimensional comparisons to inform their ratings of child competence. Preliminary evidence suggests that parents and teachers may not use dimensional comparisons in their evaluations of child math and language ability (Helm, Müller-Kalthoff, Mukowski, & Möller, 2018; Van Zanden et al., 2017). If other reporters do not use dimensional comparisons, their ratings of child competence may be more accurate. This study investigates whether beliefs about child competence in math and reading are correlated for children, parents, and teachers.

Another potential source of error in children’s competence beliefs is gender bias. It is commonly reported that boys have higher competence beliefs than girls in math and girls have higher competence beliefs than boys in reading (Eccles, Wigfield, Harold, & Blumenfeld, 1993; Wigfield & Eccles, 1994; Wigfield et al., 1997), which may or may not reflect actual performance differences. Parents have also been shown to hold gender stereotypes about children’s math ability that affect parents’ beliefs about their children’s competence (Tiedemann, 2000). However, it is less clear whether girls’ and boys’ competence beliefs (or parents’ and teachers’ beliefs about girls’ and boys’ competence) are equally predictive of academic performance. Two studies of relations between children’s math and literacy self-competence beliefs and performance reported no evidence for gender moderation (Logan & Johnson, 2009; Meece, Wigfield, & Eccles, 1990). We were unable to find any tests of child gender as a moderator of relations between parents’ or teachers’ beliefs about child competence and child academic performance. The degree to which children, parents, and teachers hold gender biases could be reflected in their ratings of child competence, which could in turn differentially affect relations with later academic performance. Hence, we included child gender as a moderator of the effects of different reporters’ beliefs about child competence on adolescent academic performance.

This Study

Differences in the predictiveness of reporters’ beliefs about child competence have implications for theory and research. If European American children use dimensional comparison for self-beliefs and adults do not, for example, children’s self-competence beliefs may be less accurate and hence poorer predictors of later academic performance than parents’ or teachers’ beliefs. However, this systematic error may be an important source of information about why children perform well or poorly in academic subjects (e.g., a unique effect of children’s beliefs). Holding inaccurate beliefs about child competence may lead to real changes in academic performance over time. For example, if children view themselves as gifted in math, they may not extend the effort needed to perform well as math becomes more challenging. Furthermore, knowing that one reporter’s beliefs are more predictive of later performance than another’s has implications for choosing reporters in research settings.

The current study had two related goals. The first goal was to assess reporter relations to determine whether all reporters’ beliefs were related. As an indicator of the relative agreement of reporters on “true” competence, we assessed how different reporters’ beliefs loaded on a latent variable of academic competence. A relatively higher loading on a latent variable for one reporter than another would indicate that the first reporter was better at rating the latent competence than the other.

The second goal was to assess whether each reporter’s beliefs about child competence uniquely predicted the child’s academic outcomes. First, zero-order correlations between reporters’ beliefs about child competence and child academic performance were assessed and compared. Then, unique relations between each reporter’s beliefs about child competence and academic performance were assessed, controlling for stability in child academic performance, other reporters, and family SES. It is possible that child, parent, and teacher competence beliefs are simply a reflection of the child’s current performance and have no relation to later performance once current performance is controlled (e.g., reporters’ competence beliefs simply reflect the current level of skill which remains stable over time). Alternatively, reporters’ beliefs about child competence may relate to later child performance controlling for current performance. For example, the expectancy-value model of achievement motivation suggests that competence beliefs may motivate a child to make different achievement-related choices (e.g., studying harder, choosing more challenging classes), which lead to increases in performance.

We expected children to be “poorer” raters of their own competence than parents and teachers. First, parents and teachers have unique insights into the child’s developmental course and overall capabilities relative to other children. Parents and teachers likely use social comparisons (e.g., comparing the child’s performance with that of siblings and peers), but it is still unclear whether parents and teachers use dimensional comparisons to inform their ratings of child competence. If parents and teachers use dimensional comparisons, we would expect their ratings of the child’s math and reading competence to be largely uncorrelated or even negatively correlated.

This longitudinal study was designed to assess children at three important developmental waves. Child competence beliefs in reading and math decline throughout elementary and middle school (Jacobs, Lanza, Osgood, Eccles, & Wigfield, 2002). The first wave was when children were in elementary school (age 10), and parents and teachers maintained a large degree of control over academic choices (e.g., study habits, course selection). In elementary school, most children have a consistent teacher for all school subjects, allowing teachers to evaluate child competence across domains and use dimensional comparisons. The second and third waves were at the end of middle school (age 13) and end of high school (age 18), when parents and teachers exercised somewhat less control over academic choices, and academic performance is crucial for future prospects (e.g., high school graduation, college admittance). We also chose outcome measures that reflected these goals – objective tests of academic ability and College Board scores.

Method

Participants

European American families with healthy firstborn children were recruited through mass mailings and newspaper advertisements from various neighborhoods and schools in the Washington, DC, metropolitan area, including suburbs in Maryland and Virginia and rural areas in West Virginia. Data were collected prospectively and longitudinally in multiple waves when children were 10, 13, and 18 years old. Altogether, 230 families provided some data on competence beliefs at the 10-year data collection wave. We excluded 10 families from which only 1 reporter provided competence belief data. Of the remaining 220 families, 31 had no academic performance data at ages 13 or 18 years and were excluded. Hence, the final study sample consisted of 189 families (81 girls and 108 boys) from which at least two reporters provided competence beliefs data at 10 years and had available data on children’s reading and/or math performance at ages 13 and/or 18 years. We compared all competence belief variables collected when the children were 10 years old between the 189 families who were in the final study sample versus the 41 who were not. No significant differences emerged between these two groups of families, ps ranged from .16 to .92.

Of these 189 families, 188 children, 166 mothers, 156 fathers, and 153 teachers provided competence beliefs data at age 10 years; 178 children provided data at 13 years; and 114 children provided data at 18 years. Missing data were missing completely at random (MCAR; see Preliminary Analyses). On average, children were 10.26 years (SD = 0.17, n = 189), 13.85 years (SD = 0.27, n = 178), and 18.21 years (SD = 0.36, n = 114) of age, respectively, at the three data collection waves. Most 10-year-olds were assessed in or just after 4th (57%) or 5th grade (42%), and 13-year-olds were assessed in or just after 7th (20%), 8th (71%), or 9th (9%) grades. The 18-year-olds were assessed in the spring of, or the summer following, their senior year of high school. Most (94%) of the 18-year-olds planned to attend college or other schooling directly after high school. At the 10-year assessment, the majority of the mothers and fathers were married and living together (n = 163; 86.24%) and had obtained at least a bachelor’s degree (mothers: 77.09%; fathers: 74.71%). Mothers were on average 40.91 years (SD = 5.27, range = 26.94 – 55.87), and fathers were 43.33 years (SD = 6.43, range = 27.78 – 63.67), of age. Family SES (Hollingshead, 1975), measured at age 10, ranged from 25 to 66 and averaged 54.93 (SD = 9.67). Of the 153 teachers who provided data at the 10-year assessment, the majority were the children’s 4th (54.90%, n = 84) or 5th (35.95%, n = 55) grade teachers (only 1 (0.65%) was a 3rd grade teacher, and 13 (8.50%) did not report their grade).

Because parenting, child development, and educational processes are known to vary with ethnicity (Bornstein, 2015; Davis-Kean, 2005; Murry, Hill, Witherspoon, Berkel, & Bartz, 2015; Yamamoto & Holloway, 2010), we recruited an ethnically homogenous sample to understand the effects of competence beliefs on the academic performance of European American families. By including only families of a single ethnicity, we intentionally avoided an ethnicity-socioeconomic status confound that has vexed the academic achievement motivation literature and would also cloud our findings (Bornstein, Jager, & Putnick, 2013; Jager, Putnick, & Bornstein, 2017).

Procedure

At 10 and 13 years, separate packets of questionnaires were mailed to the family prior to in-person visits and were completed by mothers and fathers. At age 10 years children completed measures at a home visit, and at age 13 children completed measures at a laboratory visit. At age 10, mothers were also asked to provide their child’s teacher with a packet of questionnaires to complete. The packet contained a letter explaining the study, a consent form, and an addressed, stamped envelope to return the completed questionnaires directly to the research team. At 18 years, data were collected via a password-protected internet-based survey. Informed consent was obtained from mothers, fathers, and teachers, and assent or consent (depending on child age) was obtained from children. Study procedures were approved and monitored by [blind] Institutional Review Board (protocol #88-CH-32, title [blind]).

Measures

Children’s competencies at 10 years.

Children, mothers, fathers, and teachers each completed the Children’s Competence Beliefs and Subjective Task Values (CBTV; Eccles & Wigfield, 1995; Eccles, Wigfield, Harold, & Blumenfeld, 1993; Wigfield et al., 1997) measure to assess their perceptions of children’s competencies in math and reading. Reporters were asked to rate children’s competencies on a 7-point scale, with higher scores indicating greater competence. Children answered separate questionnaires about math and reading, but parents and teachers answered similar questions about math and reading on the same form. Previous studies with this measure have demonstrated good reliability and validity (Eccles, O’Neill, & Wigfield, 2005; Fredricks & Eccles, 2002; Jacobs et al., 2002).

Subscale scores used in the current study were based on a factor analysis conducted by Eccles and colleagues (1993). Items used to assess child, mother, father, and teacher perceptions of children’s competencies in math and reading are presented in the Appendix. The items used to create the subscale scores largely tap similar aspects of children’s competencies, even though there are some differences in the number and content of the items across domains and raters. Children’s beliefs about their competence in math and reading were assessed by 6 items in each domain. Mothers and fathers were asked 4 parallel questions in each domain to assess their beliefs about children’s competence in math and reading. Teachers’ beliefs about children’s competence in math and reading were assessed by 5 items in each domain. Subscale scores were calculated by taking the mean of the items within each domain according to each reporter. The subscales used in the current study demonstrated good reliability according to child (αMath = .86, αReading = .88), mother (αMath = .90, αReading = .92), father (αMath = .92, αReading = .93), and teacher (αMath = .88, αReading = .88) report.

Academic performance at 10, 13, and 18 years.

At ages 10 and 13 years, the Woodcock–Johnson Psycho-Educational Battery–Revised (WJ–R; Woodcock & Mather, 1989) Test of Achievement was administered by trained researchers to assess mathematical competence and language skills. At 10 years Calculation, Letter-Word Identification, and Passage Comprehension subscales were administered, and at 13 years Applied Problems, Letter-Word Identification, Passage Comprehension, and Dictation subscales were administered. Calculation assessed children’s ability to perform paper-and-pencil math computations. Items ranged from writing numbers through numerical operations (addition, subtraction, multiplication, division) as well as more advanced math, if appropriate (geometric, trigonometric, logarithmic, and calculus operations). Applied Problems measured children’s ability to analyze and solve math problems. Initial items required application of simple number concepts; the majority of items required children to listen to the problem, recognize the mathematical procedure that must be followed, and perform the appropriate calculations. Letter-Word Identification required children to orally read a list of single words of increasing difficulty. Passage Comprehension measured reading comprehension by presenting sentences or passages with missing words. Children were asked to provide a word that made sense in the context of the passage. Dictation measured children’s writing skills with items such as copying letters, spelling, punctuation, capitalization, and word usage. Items were presented in order of increasing difficulty and scored 0 (incorrect or no response) or 1 (correct response). The overall WJ-R was standardized on a nationally representative sample from 2 to 95 years of age. Raw scores were converted to standard scores with M = 100 and SD = 15. Standard scores were used.

At 18 years, adolescents self-reported their College Board SAT scaled scores in Math and Verbal/Critical Reading sections. We did not collect information about when the SAT was taken or whether it was taken more than once. The Math section assessed students’ understanding and application of mathematical principles as well as numerical reasoning ability. Three main categories of math content were tested: Heart of Algebra included linear equations, systems of linear equations, and linear functions; Problem Solving and Data Analysis included statistics, modeling, and problem-solving skills; and Passport to Advanced Math included non-linear expressions, radicals, exponentials, and other topics that form the basis of more advanced math. Critical Reading consisted of three sections with varying items on sentence completion, reading short and long passages, and answering questions. The Math and Verbal/Critical Reading scaled scores had a possible range of 200–800.

Covariate.

Because academic performance varies with socioeconomic status (SES), we controlled for family SES when predicting academic outcomes. The Hollingshead (1975) Index, a measure of family SES, was scored at 10 years. The Hollingshead Index was computed as the weighted sum of the education and occupation scores for the household (possible score range = 8–66).

Results

Preliminary Analyses

Prior to analysis, univariate and multivariate distributions of all variables were examined for normality, outliers, and influential cases, and standard transformations were applied to resolve problems of non-normality (Tabachnick & Fidell, 2012). To fix skewed distributions, children’s and teachers’ competence beliefs in math and reading were raised to the second power and mothers’ and fathers’ competence beliefs in math and reading were raised to the third power. Preliminary analyses of all competence beliefs and academic achievement data indicated that 12.64% of the total data points were missing completely at random, Little’s MCAR test χ2(365, N = 189) = 394.17, p = .14. However, due to attrition over time and our use of SAT scores as the outcome at 18 years, more data was missing at 18 years than at earlier time points. Of the 75 adolescents that were missing at 18 years, 45 did not contribute any data and 30 contributed some data, but did not report SAT scores. Compared to adolescents who were not missing at 18 years (and had available data at 10 or 13 years; ns = 102–113), those that were missing at 18 years (ns = 72–75) scored lower on all academic performance indicators, ts(174–184) = −2.50 to −6.09, ps = .013 to < .001. Multiple Imputation (MI) was conducted in SPSS version 21, which uses a fully conditional Markov Chain Monte Carlo algorithm. Despite the mean differences reported above, including information about adolescents’ 10- and 13-year academic scores in the imputation model should produce unbiased estimates of the missing SAT scores at 18-years (Rubin, 1987).

To reduce bias and maximize efficiency, the imputation procedure was conducted in the following steps, as recommended by White, Royston, and Wood (2011). First, we imputed 70 datasets using a simple imputation model that included only child gender, all competence beliefs and academic performance variables, and Hollingshead Index (as an auxiliary variable), but omitted all Gender x Competence belief interaction terms. Next, to identify candidate interactions to include in a final imputation model, multiple regressions with academic performance as dependent variables and gender, competence belief, and Gender x Competence belief as predictors in the models were then conducted separately for each rater using the preliminary 70 imputed datasets. Because all interaction terms were omitted from the imputation model, test statistics of the interaction effects would be too conservative, hence p < .10 was used as the criterion to select which Gender x Competence belief interaction terms to include in the final imputation model (White et al., 2011). Doing so avoided including all 8 Gender x Competence belief interaction terms in the final imputation model which might not be relevant to the academic performance variables and could attenuate relations among study variables in the imputed datasets. Pooled results suggested that child gender might serve as a moderator between children’s self-reported (p = .06) and fathers’ ratings (p = .09) of child competence in reading and WJ-R Letter-Word Identification, between children’s self-reported competence in reading and WJ-R Dictation (p = .09), and between fathers’ (p = .08) and teachers’ (p = .08) ratings of child competence in math and WJ-R Applied Problems.

The final imputation model included child gender, all competence beliefs and academic performance variables, the Hollingshead Index, and 4 interaction terms (two-way interaction terms of child gender with children’s and fathers’ reading competence beliefs, and with fathers’ and teachers’ math competence beliefs). Only data for continuous variables were imputed as child gender had complete data. Seventy data sets were successfully generated.

Because we were interested in general language performance rather than domain-specific performance, we computed a 10-year language summary score as the mean standard score of the WJ-R Letter-Word Identification and Passage Comprehension subscales (rpooled = .57), and a 13-year language summary score as the mean standard score of the WJ-R Letter-Word Identification, Passage Comprehension, and Dictation subscales (rspooled ranged from .53 to .67 between subscales). Finally, because the 13- and 18-year academic outcomes were highly correlated in both math (rpooled = .75 between 13-year WJ-R Applied Problems and 18-year SAT Math) and language skills (rpooled = .81 between 13-year language summary score and 18-year SAT Verbal/Critical Reading), we computed a summary score each for math and language as the mean standard scores of their respective 13- and 18-year outcome variables to assess math and reading outcomes.

Analyses were run on all 70 imputed datasets. Test statistics from each imputed data set were automatically pooled in SPSS using Rubin’s (1987) rule to obtain a single set of pooled parameter estimates, which took into consideration the variance both within and between imputations (Schafer, 1997). When discussing effect sizes, Cohen’s (1988) criteria for small r = .10, medium r = .30, and large r = .50 are used.

Descriptive Statistics

Table 1 displays the means and standard deviations of competence beliefs in math and reading from children, mothers, fathers, and teachers, and the child’s academic performance. Transformed variables and imputed datasets were used in analysis; for ease of interpretation, descriptive statistics are presented in the variables’ original metrics using available Ns. The SD and ranges indicate considerable variation on all measures, as is commonly found in the literature. Mean differences in competence beliefs among raters as well as patterns of inter-rater agreement and inter-domain agreement across all four reporters at age 10 have been reported separately (Authors, under review).

Table 1.

Descriptive Statistics of Competencies Beliefs and Children’s Academic Achievement

Sample Size Mean SD Range
10-year Math Competence
 Child 187 5.35 1.10 2–7
 Mother 166 5.73 1.11 2–7
 Father 155 5.72 1.06 2–7
 Teacher 148 5.48 1.18 2–7
10-year Reading Competence
 Child 184 5.57 1.08 2–7
 Mother 166 5.75 1.26 2–7
 Father 156 5.76 1.26 2–7
 Teacher 152 5.41 1.20 2–7
Woodcock-Johnson Revised
 10-year
  Calculation 186 119.43 16.59 82–191
  Letter-Word Identification 185 116.26 13.99 89–150
  Passage Comprehension 184 117.03 10.86 86–150
 13-year
  Applied Problems 178 117.96 17.77 85–169
  Letter-Word Identification 178 114.56 13.79 91–157
  Passage Comprehension 178 119.45 17.18 91–176
  Dictation 176 99.30 15.82 71–153
18-Year SAT Math 114 651.58 94.82 430–800
18-Year SAT Critical Reading/Verbal 114 637.19 101.25 400–800
Open in a new tab

Reporter Relations

Table 2 presents zero-order correlations that were pooled across imputations. Correlations were medium to large between reporters for math (rs = .36 to .66) and reading (rs = .41 to .77) competence. As expected from dimensional comparison theory, children’s math and reading competence beliefs were not correlated (r = .02, ns), but mothers’, fathers’, and teachers’ math and reading competence beliefs were (rs = .44 to .69, ps < .001), suggesting that adults may not use dimensional comparisons to inform their ratings of child competencies.

Table 2.

Pooled Zero-order Correlations among Study Variables

1 2 3 4 5 6 7 8 9 10 11 12
10-year Math Competence
 1. Child 1
 2. Mother .57*** 1
 3. Father .43*** .66*** 1
 4. Teacher .36*** .53*** .49*** 1
10-year Reading Competence
 5. Child .02 .16* .13 .16* 1
 6. Mother .07 .44*** .36*** .27*** .64*** 1
 7. Father .01 .27*** .47*** .20* .61*** .77*** 1
 8. Teacher .06 .25*** .29*** .69*** .41*** .47*** .47*** 1
Math Performance
 9. 10-year WJ-R Calculation .48*** .54*** .57*** .49*** .18* .35*** .34*** .32*** 1
 10. 13/18-year Math Summary Score .30*** .59*** .59*** .50*** .20* .38*** .45*** .42*** .62*** 1
Language Performance
 11. 10-year Language Summary Score .14 .43*** .48*** .38*** .49*** .73*** .68*** .48*** .48*** .56*** 1
 12. 13/18-year Language Summary Score .09 .43*** 50*** .39*** .46*** .67*** .69*** .50*** .52*** .74*** .83*** 1
Covariate
 13. Hollingshead Index .06 .06 .12 .15 .06 .06 .05 .12 .22** .38*** .22** .33***
Open in a new tab

Note. N = 189, 70 imputed datasets.

*

p < .05,

**

p < .01,

***

p ≤ .001.

Two separate confirmatory factor analysis (CFA) models for competence beliefs in math and reading were evaluated using EQS 6.1 (Bentler, 2006; Bentler & Weeks, 1980). In both models, factors were standardized to have unit variances which allowed us to test the significance of all 4 factor loadings in each model. After CFA models were confirmed, we constrained factor loadings to be equal between pairs of reporters to assess differences in reporter loadings on latent academic competence. To avoid the problem of capitalization on chance due to the number of the equality constraints tested, a change in χ2 of 5 or more in the univariate Lagrange multiplier test was required to set the constraint free (Scott-Lennox & Lennox, 1995).

Excellent model fits were achieved for the CFA models: χ2(df = 2, N = 189) = 1.72, p = .42, Comparative Fit Index (CFI) = 1.00, Standardized Root Mean Squared Residual (SRMR) = .02 for the math competence belief model; χ2(df = 2, N = 189) = 0.54, p = .76, CFI = 1.00, SRMR = .01 for the reading competence belief model. Factor loadings were all significant at the .001 level. For competence beliefs in math, mothers’ factor loading (standardized regression coefficient ẞ = .92) was significantly higher than children’s (ẞ = .62; χ2(1) = 11.04, p = .001) and teachers’ (ẞ = .60; χ2(1) = 11.04, p < .01). Fathers’ factor loading was .73. All other pairs of comparisons did not differ. For competence beliefs in reading, mothers’ factor loading (ẞ = .90) was significantly higher than children’s (ẞ = .72; χ2(1) = 6.69, p = .01) and teachers’ (ẞ = .57; χ2(1) = 12.59, p < .001), and fathers’ (ẞ = .86) was significantly higher than teachers’ (χ2(1) = 9.93, p < .01). All other pairs of comparisons did not differ.

Predictiveness of Competence Beliefs

Competence beliefs from all 4 reporters at age 10 related significantly to children’s 13/18-year academic performance, and the correlations were all large with the exception of the medium size correlation between children’s self-competence beliefs in math and their later math academic outcome (Table 2). We applied Fisher’s z-transformation for correlation coefficients to assess the statistical significances of differences in the correlations between pairs of raters. In the math domain, the correlation between children’s self-competence beliefs and 13/18-year math performance was smaller than those of mothers’ and fathers’ (both zs = −3.55, p < .001) and of teachers’ (z = −2.31, p < .05) competence beliefs with adolescent performance. Mothers’, fathers’, and teachers’ correlations did not differ from one another. In the language domain, the correlation between children’s self-competence beliefs and 13/18-year language performance was smaller than those of mothers’ (z = −3.02, p < .01) and fathers’ (z = −3.38, p < .001) competence beliefs with adolescent performance, and the correlation between teachers’ competence beliefs and 13/18-year language performance was smaller than those of mothers’ (z = −2.52, p < .05) and fathers’ (z = −2.88, p < .01) competence beliefs with adolescent performance. Mothers’ and fathers’ correlations did not differ from each other, and children’s and teachers’ correlations did not differ from each other.

Children’s competence beliefs in one domain were largely unrelated to their concurrent or later performance in the other domain (rs = .09 - .20; Table 2). However, mothers’ (rs = .35 - .43), fathers’ (rs = .34 - .50), and teachers’ (rs = .32 - .42) beliefs about child competence were at least moderately correlated with both concurrent and later performance in the other domain.

Tables 3a and 3b present summaries of hierarchical regression analyses for competence beliefs predicting academic performance in math and reading, respectively. At the first step we assessed the unique predictive ability of beliefs about child competence from each of the 4 reporters on 13/18-year academic outcomes controlling for 10-year child academic performance, family SES, child gender, and other reporters in the models. In the second step, we tested if child gender moderated relations between raters’ competence beliefs and adolescent academic outcomes. To control for stability in academic performance, the 13/18-year math summary score was controlled for 10-year WJ-R Calculation; and the 13/18-year language summary score was controlled for the 10-year language summary score. Multicollinearity was monitored in all models, and variance inflation factors (VIFs) were always smaller than 4, indicating that multicollinearity was not an issue.

Table 3a.

Hierarchical Regression Analyses: Competence Beliefs in Math Predicting Math Performance Controlling for Stability in Academic Achievement, Family SES, Child Gender, and Other Reporters

14- and 18-year Math Summary Score
B SE Part Correlation t FMIa
Step 1
10-year WJ-R Calculation 0.018 0.004 .24 4.63*** .14
Family SES 0.024 0.005 .24 4.59*** .21
Child Gender 0.084 0.106 .04 0.79 .19
Competence Belief Child −0.015 0.006 −.14 −2.58** .22
Competence Belief Mother 0.003 0.001 .18 3.17** .34
Competence Belief Father 0.002 0.001 .14 2.39* .36
Competence Belief Teacher 0.008 0.005 .09 1.52 .31
Step 2
Gender x Competence Belief Child 0.001 0.011 .00 0.07 .17
Gender x Competence Belief Mother 0.001 0.002 .02 0.32 .24
Gender x Competence Belief Father 0.001 0.002 .02 0.33 .32
Gender x Competence Belief Teacher −0.002 0.011 −.01 −0.15 .29
Open in a new tab

Note. N = 189, 70 imputed datasets. Pooled R2 was not available; R2 ranged from .53-.62 in Step 1 and .52-.62 in Step 2 in the 70 imputed datasets.

a

Fraction Missing Information (FMI) represents the amount of information that is missing from a parameter estimate because of the missing data.

*

p < .05.

**

p < .01.

***

p < .001.

Table 3b.

Hierarchical Regression Analyses: Competence Beliefs in Reading Predicting Language Performance Controlling for Stability in Academic Achievement, Family SES, Child Gender, and Other Reporters

13- and 18-year Language Summary Score
B SE Part Correlation t FMIa
Step 1
10-year WJ-R Language Summary score 0.556 0.064 .35 8.72*** .15
Family SES 0.017 0.004 .18 3.92*** .34
Child Gender −0.009 0.078 −.01 −0.11 .24
Competence Belief Child −0.003 0.004 −.03 −0.70 .23
Competence Belief Mother 0.000 0.001 .02 0.48 .29
Competence Belief Father 0.002 0.001 .15 3.51*** .26
Competence Belief Teacher 0.006 0.004 .07 1.45 .40
Step 2
Gender x Competence Belief Child 0.023 0.010 .10 2.41* .31
Gender x Competence Belief Mother −0.001 0.001 −.03 −0.69 .18
Gender x Competence Belief Father −0.001 0.001 −.02 −0.53 .33
Gender x Competence Belief Teacher 0.004 0.008 .02 0.54 .31
Open in a new tab

Note. N = 189, 70 imputed datasets. Pooled R2 was not available; R2 ranged from .71-.79 in Step 1 and .71-.80 in Step 2 in the 70 imputed datasets.

a

Fraction Missing Information.

*

p < .05.

**

p < .01.

***

p < .001.

Controlling for 10-year academic performance, family SES, child gender, and other reporters in the model, children’s, mothers’, and fathers’ competence beliefs at 10 years each uniquely predicted the 13/18-year math summary score. Removing their shared variance with other predictors in the model, mothers’ (rpart, pooled = .18) and fathers’ (rpart, pooled = .14) perceived higher beliefs in their children’s math competence each uniquely predicted higher levels of math performance later (and lower competence beliefs uniquely predicted lower performance); whereas children who reported higher unique competence beliefs in math at age 10 scored lower on the 13/18-year math summary score (rpart, pooled = −.14). Teachers’ competence beliefs at 10 years did not predict later math performance over and above other predictors in the model. Child gender did not moderate the predictiveness of raters’ competence beliefs on later math performance.

Controlling for 10-year language performance, family SES, child gender, and other reporters in the model, fathers’ perceived higher beliefs in their children’s reading competence at age 10 uniquely predicted higher 13/18-year language summary scores (rpart, pooled = .15). Child gender moderated the predictiveness of children’s competence beliefs at 10 years on 13/18-year language outcomes. Girls who reported higher self-competence beliefs in reading at age 10 scored lower on the 13/18-year language summary score (rpart, pooled = −.15, B = −0.015, SE = 0.007, t = −2.29, p < .05) controlling for 10-year academic performance, family SES, and other reporters in the model. Boys’ self-competence beliefs in reading at age 10 did not predict 13/18-year language performance (rpart, pooled = .05, B = 0.005, SE = 0.006, t = 0.81, ns), controlling for 10-year academic performance, family SES, and other reporters in the model. Mothers’ and teachers’ competence beliefs at age 10 did not independently predict adolescent language performance over and above other predictors in the model.

Discussion

This study investigated how reports of European American children’s math and reading competence from four different reporters related to children’s actual math and language performance in adolescence. Taken together, these findings indicate that European American children engage in dimensional comparisons for math and reading, confirming those of Möller and colleagues (2009, 2013), but our findings suggest that parents and teachers may not. Although we did not assess competence beliefs past the age of 10, evidence for dimensional comparison has been found in children from elementary through high school across many countries (Möller et al., 2009), as well as in German college students (Müller-Kalthoff, Jansen, Schiefer, Helm, Nagy, & Möller, 2017). Dimensional comparisons are not a transient source of bias in children’s reports. Wolff, Helm, and Möller (2018) suggested that dimensional comparisons are used by adolescents to differentiate the self, a motivation that may only be held by children and not parents or teachers.

European American mothers and fathers tended to be the best reporters in terms of their loadings on a latent variable of child competence, and in terms of their unique predictiveness to adolescent academic performance. One advantage of this study design was that we included fathers and explored their competence beliefs separately from (and in comparison to those of) mothers. Despite sharing a home environment and many of the same experiences with the child, mothers and fathers in this study had unique perspectives on children’s academic competence (Jager, Bornstein, Putnick, & Hendricks, 2012; Jager et al., 2016). Mothers’ competence beliefs had the highest loadings on the latent competence belief factors, and fathers’ loadings were similar to mothers’. Zero-order correlations of mothers’ and fathers’ competence beliefs with child performance outcomes were also similar. Finally, fathers’ unique competence beliefs were related to adolescents’ math and reading performance, whereas mothers’ unique competence beliefs were only related to adolescents’ math outcomes. Phillips (1987) found that fathers’ (but not mothers’) estimates of their child’s ability related to their child’s self-competence beliefs. Similarly, Wagner and Phillips (1992) analyzed the interactions of children with mothers and fathers on the same sample, finding that fathers’ (but not mothers’) warmth was associated with child perceived math competence. More research on fathers’ influences on children’s academic self-concepts, study habits, and expectations is warranted.

Overall, this study suggests that European American children’s self-competence beliefs may not be the best predictor their future academic performance. Children’s self-competence beliefs had lower loadings on a latent factor of child competence than mothers’, their competence beliefs were less predictive of outcomes than parents’, and, when controlling concurrent performance and competence beliefs of other socializers (mother, father, and teacher), children’s own competence beliefs were inversely associated with their later academic performance. Similarly and surprisingly, teachers’ competence beliefs may not be the best predictors of later child performance. Although we found no evidence of dimensional comparisons or gender-moderated relations for teachers, teachers’ competence beliefs had lower loadings on the latent factor than mothers’ (and fathers’ for reading), predictive relations between teachers’ competence beliefs and adolescent academic performance were smaller than mothers’ and fathers’ in reading, and teachers’ competence beliefs were not uniquely predictive of adolescent academic performance in either domain when controlling for concurrent performance, family SES, and other reporters. Perhaps teachers are poorer judges of actual ability (given the limited scope of their experience with the child), or teachers’ beliefs are less predictive than parents’ because teachers change year-to-year whereas parents are more knowledgeable about their child and are a consistent influence on their children’s academics throughout their lives. Another possible reason for teachers’ poorer predictive relations is that the teacher scale included more differentiated items than parent and child scales (see Appendix).

The expectancy-value model of achievement motivation suggests that parent competence beliefs predict child competence beliefs, but no direct unique relations with child performance are theorized independent of the child’s own beliefs, as we found here. It should be noted that this study was not a full test of the expectancy-value model. We did not include a separate measure of expectancies for success, nor did we include measures of task values or other related constructs. Still, this study suggests that parents’ unique beliefs may lead to better European American adolescent academic performance. At age 10, parents exert a great deal of control over children’s homework, school choices, and extracurricular activities. Children make few independent decisions about their academics at this age. However, Dweck (2002) suggested that age 10 is the beginning of a new kind of thought process. Children in the 10- to 12-year range (relative to 7- to 8-year-olds) view ability as more stable and predictable, use social comparison more to inform their self-competence beliefs and motivation, and have largely accurate competence beliefs. As children enter adolescence, these higher-level thought processes coupled with greater independence in decision-making may lead to a shift in predictive relations between child self-competence beliefs and their own performance. High school students’ own competence beliefs, for example, may be more important to their academic performance than parents’ and teachers’ beliefs about child competence. It may be that dimensional comparisons decline as adolescents develop higher-level cognitive skills, but Marsh and colleagues (2015) found evidence for dimensional comparisons in high school (see also Möller et al., 2009). Although this study found that parents’ beliefs about child competence predicted European American child performance in adolescence, we cannot rule out the possibility that parents’ and teachers’ competence beliefs relate to later academic performance through changes in the child’s own competence beliefs as children age into adolescence. Future research should assess children’s, parents’, and teachers’ beliefs about child competence and academic performance longitudinally from childhood through adolescence so that stability and change in competence beliefs can be assessed and accounted for in prediction. Specifically, with repeated assessments of children’s, parents’, and teachers’ beliefs about child competence, future studies could test competing models of development regarding whether parents’ and teachers’ beliefs influence academic performance directly or only through children’s self-competence beliefs.

This study found that in European American families, when a parent believes in the child’s competence (independent of the child’s own competence beliefs and actual performance), the child performs better later, even controlling for earlier performance. Statistically speaking, by including all reporters in the same model the agreement (common variance) among reporters is removed, leaving only reporter unique variance. If the portion of the variance that each reporter shares is viewed as an accurate portrayal of the child’s competence, then the unshared variance is at best an idiosyncratic view of the child, and at worst an inaccurate view of the child. Hence, when a reporter’s unique variance is predictive of later performance, it suggests two possible explanations. The first explanation is that the reporter’s idiosyncratic view is based on some information that is not shared by the other reporters (e.g., setting-sensitive information; De Los Reyes et al., 2013). The mother’s unique beliefs could easily be interpreted this way. Mothers are often the primary conduit of information to and from school (Tan & Goldberg, 2009), and mothers have the benefit of knowing intimately what happens at home. Hence mothers may hold information that fathers and teachers do not have.

The second explanation is that the reporter’s unique beliefs about child competence are inaccurate. For example, children who use dimensional comparisons tend to overestimate their competence in the favored domain and underestimate their competence in the unfavored domain. If children believe they are less competent than they actually are, they may focus effort on that domain, thereby improving their skills, whereas if children are overconfident (believing they are more competent than they actually are), they may not work as hard and perform worse later. Conversely, having a parent who views the child as competent in reading, even if the child struggles, might lead the parent to provide the child with more opportunities to maximize reading skills (e.g., more advanced books), which improve verbal skills later. A reporter’s unique variance likely represents a combination of the reporter’s idiosyncratic view, some systematic error or bias, and some random error. Regardless, children’s and parents’ unique variance is predictive of later child performance. Perhaps it is beneficial for European American parents to have slightly inflated beliefs about their child’s competencies because having a parent who believes the child is inherently competent spurs children to work harder in those areas to live up to expectations. Future research should explore what parents, teachers, and children do (e.g., study habits, course choices, tutoring) that may mediate the link between unique competence beliefs and later child performance.

Only 1 (of 8) effect in this study was moderated by child gender. European American girls who had stronger unique competence beliefs in reading at age 10 performed worse on language outcomes later; this effect was not significant for boys. European American girls tend to have slightly higher reading/English competence beliefs than boys in 4th through 6th grade (Andre, Whigham, Hendrickson, & Chamber, 1999; Wigfield et al., 1997; Wigfield & Eccles, 1994). Perhaps European American girls are more likely to overestimate their reading competence at age 10 and hence perform worse later, whereas European American boys have more realistic reading competence beliefs. As this is the first study (to our knowledge) to show a stronger negative unique relation between reading competence beliefs and performance for girls, future studies should be undertaken to corroborate this finding.

Study limitations include the slightly varying questions across reporters (except mothers and fathers, who received identical questions), the use of an older version of the Woodcock Johnson (although it was the current version when data collection began), the exclusion of task value items that may relate to competence beliefs, the absence of repeated assessments of competence beliefs in adolescence, reliance on SAT scores at 18 years, and the relatively high average socioeconomic status of the participants. Given the sample in this study, the findings can only reasonably be generalized to middle class European American adolescents. Future research is needed to understand how competence beliefs relate to performance in other ethnic and socioeconomic groups. Future research should also include measures of the values placed on math and reading as well as academic engagement. Values and engagement may moderate or mediate relations between competence and academic achievement (Chouinard, Karsenti, & Roy, 2007; Raftery, Grolnick, & Flamm, 2012).

Taken together, this study’s findings have implications for improving European American adolescent academic performance. Dimensional comparisons increase systematic error in children’s reports that may render their self-reports of competence less reliable or predictive than others’ reports about them. Accounting for concurrent performance, children’s more positive beliefs about their own competence lead to lower performance later. Evidence suggests that the amount of positive bias (overestimation of ability) in one domain essentially cancels out the negative bias (underestimation of ability) in another such that there is no overall benefit to children’s self-concepts (Müller-Kalthoff et al., 2017). Hence, it may be beneficial to intervene with children to help them view their own competencies more accurately, and to understand that competence is plastic to experience (Mangels, Butterfield, Lamb, Good, & Dweck, 2006). For example, a school-based intervention designed to improve children’s competence beliefs and attitudes toward math demonstrated efficacy in improving girls’ math competence beliefs (Falco, Summers, & Bauman, 2010). Furthermore, the child’s own competence beliefs may not be the most important target for intervention, at least in elementary school. Our findings suggest that targeting parent competence beliefs could also help to improve children’s academic outcomes, although this finding awaits replication in more diverse samples. Raftery and colleagues (2012) highlighted the importance of parental involvement in children’s school success, but more research is needed to understand what parents who hold high competence beliefs do to help children improve their academic performance.

Acknowledgments

This research was supported by the Intramural Research Program of the NIH, NICHD.

Appendix

Items to Assess Child, Mother, Father, and Teacher Perceptions of Children’s Competencies in Math and Reading

Child Mother/Father Teacher
Math
1. How good in math are you? 1. How good is your child in math? 1. Compared to other children, how hard does this child try in math?
2. If you were to list all the students in your class from the worst to the best in math, where would you put yourself? 2. In comparison to other children, how would you evaluate your child’s performance in math? 2. How well is this child performing in math compared to how well you believe s/he could?
3. Compared to most of your other school subjects, how good are you in math? 3. Compared to other children, how much innate ability or talent does this child have in math? 3. Compared to other children, how much innate ability or talent does this child have in math?
4. How well do you expect to do in math this year? 4. How well do you think your child will do in math next year? 4. How well do you expect this child to do next year in math?
5. How good would you be at learning something new in math? 5. Compared to other children, to what extent does this child give up when faced with a difficult problem in math?
6. In general, how hard is math for you? (reversed)
Reading
1. How good in reading are you? 1. How good is your child in reading? 1. Compared to other children, how hard does this child try in reading?
2. If you were to list all the students in your class from worst to best in reading, where would you put yourself? 2. In comparison to other children, how would you evaluate your child’s performance in reading? 2. How well is this child performing in reading compared to how well you believe s/he could?
3. Compared to most of your other school subjects, how good are you in reading? 3. Compared to other children, how much innate ability or talent does this child have in reading? 3. Compared to other children, how much innate ability or talent does this child have in reading?
4. How well do you expect to do in reading this year? 4. How confident is your child in his/her ability to do well in reading? 4. How well do you expect this child to do next year in reading?
5. How good would you be at learning something new in reading? 5. Compared to other children, to what extent does this child give up when faced with a difficult problem in reading?
6. In general, how hard is reading for you? (reversed)
Open in a new tab

References

  1. Altermatt ER, Pomerantz EM, Ruble DN, Frey KS, & Greulich FK (2002). Predicting changes in children’s self-perceptions of academic competence: A naturalistic examination of evaluative discourse among classmates. Developmental Psychology, 38, 903–917. doi: 10.1037/0012-1649.38.6.903 [DOI] [PubMed] [Google Scholar]
  2. Andre T, Whigham M, Hendrickson A, & Chambers S (1999). Competency beliefs, positive affect, and gender stereotypes of elementary students and their parents about science versus other school subjects. Journal of Research in Science Teaching, 36, 719–747. doi: [DOI] [Google Scholar]
  3. Bentler PM (2006). EQS 6 structural equations program manual. Encino, CA: Multivariate Software, Inc. [Google Scholar]
  4. Bentler PM, & Weeks DG (1980). Linear structural equations with latent variables. Psychometrika, 45, 289–309. doi: 10.1007/BF02293905 [DOI] [Google Scholar]
  5. Boerma IE, Mol SE, & Jolles J (2016). Teacher perceptions affect boys’ and girls’ reading motivation differently. Reading Psychology, 37, 547–569. doi: 10.1080/02702711.2015.1072608 [DOI] [Google Scholar]
  6. Bornstein MH (2015). Children’s parents. In Lerner RM (Ed. in chief) & Bornstein MH & Leventhal T (Eds.), Ecological settings and processes in developmental systems. Volume 4 of the Handbook of child psychology and developmental science (7th ed., pp. 55–132). Hoboken, NJ: Wiley. [Google Scholar]
  7. Bornstein MH, Jager J, & Putnick DL (2013). Sampling in developmental science: Situations, shortcomings, solutions, and standards. Developmental Review, 33, 357–370. doi: 10.1016/j.dr.2013.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chouinard R, Karsenti T, & Roy N (2007). Relations among competence beliefs, utility value, achievement goals, and effort in mathematics. British Journal of Educational Psychology, 77, 501–517. doi: 10.1348/000709906X133589 [DOI] [PubMed] [Google Scholar]
  9. Cohen J (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. [Google Scholar]
  10. Davis-Kean PE (2005). The influence of parent education and family income on child achievement: The indirect role of parental expectations and the home environment. Journal of Family Psychology, 19, 294–304. doi: 10.1037/0893-3200.19.2.294 [DOI] [PubMed] [Google Scholar]
  11. De Los Reyes A, Ehrlich KB, Swan AJ, Luo TJ, Van Wie M, & Pabón SC (2013). An experimental test of whether informants can report about child and family behavior based on settings of behavioral expression. Journal of Child and Family Studies, 22, 177–191. doi: 10.1007/s10826-012-9567-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Dweck CS (2002). The development of ability conceptions. In Wigfield A & Eccles JS (Eds.), Development of achievement motivation (pp. 57–88). San Diego, CA: Academic Press. doi: 10.1016/B978-012750053-9/50005-X [DOI] [Google Scholar]
  13. Eccles JS, O’Neill SA, & Wigfield A (2005). Ability self-perceptions and subjective task values in adolescents and children. In Moore KA & Lippman LH (Eds.), What do children need to flourish? Conceptualizing and measuring indicators of positive development (pp. 237–249). New York, NY: Springer. [Google Scholar]
  14. Eccles J, & Wigfield A (1995). In the mind of the actor: The structure of adolescents’ achievement task values and expectancy-related beliefs. Personality and Social Psychology Bulletin, 21, 215–225. doi: 10.1177/0146167295213003 [DOI] [Google Scholar]
  15. Eccles J, Wigfield A, Harold RD, & Blumenfeld P (1993). Age and gender differences in children’s self‐and task perceptions during elementary school. Child Development, 64, 830–847. doi: 10.1111/j.1467-8624.1993.tb02946.x [DOI] [PubMed] [Google Scholar]
  16. Entwisle DR, & Baker DP (1983). Gender and young children’s expectations for performance in arithmetic. Developmental Psychology, 19, 200–209. doi: 10.1037/0012-1649.19.2.200 [DOI] [Google Scholar]
  17. Falco LD, Summers JJ, & Bauman S (2010). Encouraging mathematics participation through improved self-efficacy: a school counseling outcomes study. Educational Research and Evaluation, 16, 529–549. doi: 10.1080/13803611.2011.555101 [DOI] [Google Scholar]
  18. Fredricks JA, & Eccles JS (2002). Children’s competence and value beliefs from childhood through adolescence: Growth trajectories in two male-sex-typed domains. Developmental Psychology, 38, 519–533. doi: 10.1037/0012-1649.38.4.519 [DOI] [PubMed] [Google Scholar]
  19. Frome PM, & Eccles JS (1998). Parents’ influence on children’s achievement-related perceptions. Journal of Personality and Social Psychology, 74, 435–452. doi: 10.1037/0022-3514.74.2.435 [DOI] [PubMed] [Google Scholar]
  20. Graham S (1994). Motivation in African Americans. Review of Educational Research, 64, 55–117. doi: 10.3102/00346543064001055 [DOI] [Google Scholar]
  21. Halle TG, Kurtz-Costes B, & Mahoney JL (1997). Family influences on school achievement in low-income, African American children. Journal of Educational Psychology, 89, 527–537. doi: 10.1037/0022-0663.89.3.527 [DOI] [Google Scholar]
  22. Hansford BC, & Hattie JA (1982). The relationship between self and achievement/performance measures. Review of Educational Research, 52, 123–142. doi: 10.3102/00346543052001123 [DOI] [Google Scholar]
  23. Helm F, Müller-Kalthoff H, Mukowski R, & Möller J (2018). Teacher judgement accuracy regarding students’ self-concepts: Affected by social and dimensional comparisons? Learning and Instruction, 55, 1–12. doi: 10.1016/j.learninstruc.2018.02.002 [DOI] [Google Scholar]
  24. Herbert J, & Stipek D (2005). The emergence of gender differences in children’s perceptions of their academic competence. Applied Developmental Psychology, 26, 276–295. doi: 10.1016/j.appdev.2005.02.007 [DOI] [Google Scholar]
  25. Hollingshead AB (1975). The four-factor index of social status. Unpublished manuscript, Yale University, New Haven, Connecticut. [Google Scholar]
  26. Jacobs JE, Davis-Kean P, Bleeker M, Eccles JS, & Malanchuk O (2005). “I can, but I don’t want to”: The impact of parents, interests, and activities on gender differences in math. In Gallagher A & Kaufman J (Eds.), Gender difference in mathematics (pp. 246–263). New York, NY: Cambridge University Press. [Google Scholar]
  27. Jacobs JE, Lanza S, Osgood DW, Eccles JS, & Wigfield A (2002). Changes in children’s self-competence and values: Gender and domain differences across grades one through twelve. Child Development, 73, 509–527. doi: 10.1111/1467-8624.00421 [DOI] [PubMed] [Google Scholar]
  28. Jager J, Bornstein MH, Putnick DL, & Hendricks C (2012). Family members’ unique perspectives of the family: Examining their scope, size, and relations to individual adjustment. Journal of Family Psychology, 26, 400–410. doi: 10.1037/a0028330. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Jager J, Mahler A, An D, Putnick DL, Bornstein MH, Lansford JE, … & Deater-Deckard K (2016). Early adolescents’ unique perspectives of maternal and paternal rejection: Examining their across-dyad generalizability and relations with adjustment 1 year later. Journal of Youth and Adolescence, 45, 2108–2124. doi: 10.1007/s10964-016-0509-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Jager J, Putnick DL, & Bornstein MH (2017). More than just convenient: The scientific merits of homogeneous convenience samples. Monographs of the Society for Research in Child Development, 82, 13–30. doi: 10.1111/mono.12296 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Jussim L, & Harber KD (2005). Teacher expectations and self-fulfilling prophecies: Knowns and unknowns, resolved and unresolved controversies. Personality and Social Psychology Review, 9, 131–155. doi: 10.1207/s15327957pspr0902_3 [DOI] [PubMed] [Google Scholar]
  32. Logan S, & Johnson R (2009). Gender differences in reading ability and attitudes: Examining where these differences lie. Journal of Research in Reading, 32, 199–214. doi: 10.1111/j.1467-9817.2008.01389.x [DOI] [Google Scholar]
  33. Mangels JA, Butterfield B, Lamb J, Good C, & Dweck CS (2006). Why do beliefs about intelligence influence learning success? A social cognitive neuroscience model. Social Cognitive and Affective Neuroscience, 1, 75–86. doi: 10.1093/scan/nsl013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Marsh HW (1986). Verbal and math self-concepts: An internal/external frame of reference model. American Educational Research Journal, 23, 129–149. doi: 10.3102/00028312023001129 [DOI] [Google Scholar]
  35. Marsh HW (1990). Causal ordering of academic self-concept and academic achievement: A multiwave, longitudinal panel analysis. Journal of Educational Psychology, 82, 646–656. doi: 10.1037/0022-0663.82.4.646 [DOI] [Google Scholar]
  36. Marsh HW, Lüdtke O, Nagengast B, Trautwein U, Abduljabbar AS, Abdelfattah F, & Jansen M (2015). Dimensional comparison theory: Paradoxical relations between self-beliefs and achievements in multiple domains. Learning and Instruction, 35, 16–32. doi: 10.1016/j.learninstruc.2014.08.005 [DOI] [Google Scholar]
  37. Marsh HW, Trautwein U, Lüdtke O, Köller O, & Baumert J (2005). Academic self‐concept, interest, grades, and standardized test scores: Reciprocal effects models of causal ordering. Child Development, 76, 397–416. doi: 10.1111/j.1467-8624.2005.00853.x [DOI] [PubMed] [Google Scholar]
  38. Meece JL, Wigfield A, & Eccles JS (1990). Predictors of math anxiety and its influence on young adolescents’ course enrollment intentions and performance in mathematics. Journal of Educational Psychology, 82, 60–70. doi: 10.1037/0022-0663.82.1.60. [DOI] [Google Scholar]
  39. Möller J, & Marsh HW (2013). Dimensional comparison theory. Psychological Review, 120, 544–560. doi: 10.1037/a0032459 [DOI] [PubMed] [Google Scholar]
  40. Möller J, Pohlmann B, Köller O, & Marsh HW (2009). A meta-analytic path analysis of the internal/external frame of reference model of academic achievement and academic self-concept. Review of Educational Research, 79, 1129–1167. doi: 10.3102/0034654309337522 [DOI] [Google Scholar]
  41. Müller-Kalthoff H, Jansen M, Schiefer IM, Helm F, Nagy N, & Möller J (2017). A double-edged sword? On the benefit, detriment, and net effect of dimensional comparison on self-concept. Journal of Educational Psychology, 109, 1029–1047. doi: 10.1037/edu0000171 [DOI] [Google Scholar]
  42. Murry VM, Hill NE, Witherspoon D, Berkel C, & Bartz D (2015). Children in diverse social contexts. In Lerner RM (Ed. In chief) & Bornstein MH & Leventhal T (Eds.), Ecological settings and processes in developmental systems. Volume 4 of the Handbook of child psychology and developmental science (7th ed., pp. 416–454). Hoboken, NJ: Wiley. [Google Scholar]
  43. Musu-Gillette LE, Wigfield A, Harring JR, & Eccles JS (2015). Trajectories of change in students’ self-concepts of ability and values in math and college major choice. Educational Research and Evaluation, 21, 343–370. doi: 10.1080/13803611.2015.1057161 [DOI] [Google Scholar]
  44. Newman RS (1984). Children’s achievement and self-evaluations in mathematics: A longitudinal study. Journal of Educational Psychology, 76, 857–873. doi: 10.1037/0022-0663.76.5.857 [DOI] [Google Scholar]
  45. Nicholls JG (1979). Development of perception of own attainment and causal attributions for success and failure in reading. Journal of Educational Psychology, 71, 94–99. doi: 10.1037/0022-0663.71.1.94 [DOI] [PubMed] [Google Scholar]
  46. Obach MS (2003). A longitudinal-sequential study of perceived academic competence and motivational beliefs for learning among children in middle school. Educational Psychology, 23, 323–338. doi: 10.1080/0144341032000060156 [DOI] [Google Scholar]
  47. Okagaki L (2006). Ethnicity and Learning. In Alexander PA & Winne PH (Eds.), Handbook of educational psychology (pp. 615–634). Mahwah, NJ: Erlbaum. [Google Scholar]
  48. Parsons JE, Adler TF, & Kaczala CM (1982). Socialization of achievement attitudes and beliefs: Parental influences. Child Development, 53, 310–321. doi: 10.2307/1128973 [DOI] [Google Scholar]
  49. Parsons JE, Kaczala CM, & Meece JL (1982). Socialization of achievement attitudes and beliefs: Classroom influences. Child Development, 53, 322–339. doi: 10.2307/1128974 [DOI] [Google Scholar]
  50. Phillips DA (1987). Socialization of perceived academic competence among highly competent children. Child Development, 58, 1308–1320. doi: 10.2307/1130623 [DOI] [PubMed] [Google Scholar]
  51. Pinxten M, Wouters S, Preckel F, Niepel C, De Fraine B, & Verschueren K (2015). The formation of academic self-concept in elementary education: A unifying model for external and internal comparisons. Contemporary Educational Psychology, 41, 124–132. doi: 10.1016/j.cedpsych.2014.12.003 [DOI] [Google Scholar]
  52. Pomerantz EM, & Dong W (2006). Effects of mothers’ perceptions of children’s competence: The moderating role of mothers’ theories of competence. Developmental Psychology, 42, 950–961. doi: 10.1037/0012-1649.42.5.950 [DOI] [PubMed] [Google Scholar]
  53. Raftery JN, Grolnick WS, & Flamm ES (2012). Families as facilitators of student engagement: Toward a home-school partnership model. In Christenson SL, Reschly AL, & Wylie C (Eds.), Handbook of research on student engagement (pp. 343–364). New York, NY: Springer. [Google Scholar]
  54. Richardson M, Abraham C, & Bond R (2012). Psychological correlates of university students’ academic performance: A systematic review and meta-analysis. Psychological Bulletin, 138, 353–387. doi: 10.1037/a0026838 [DOI] [PubMed] [Google Scholar]
  55. Rubie-Davies CM, Weinstein RS, Huang FL, Gregory A, Cowan PA, & Cowan CP (2014). Successive teacher expectation effects across the early school years. Journal of Applied Developmental Psychology, 35, 181–191. doi: 10.1016/j.appdev.2014.03.006 [DOI] [Google Scholar]
  56. Rubin DB (1987). Multiple imputation for nonresponse surveys. New York, NY: Wiley. [Google Scholar]
  57. Schafer JL (1997). Analysis of incomplete multivariate data. London: Chapman & Hall. [Google Scholar]
  58. Scott-Lennox JA, & Lennox RD (1995). Sex-race differences in social support and depression in older low-income adults. In Hoyle RH (Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 199–216). Thousand Oaks, CA: Sage. [Google Scholar]
  59. Simpkins SD, Davis-Kean PE, & Eccles JS (2006). Math and science motivation: A longitudinal examination of the links between choices and beliefs. Developmental Psychology, 42, 70–83. doi: 10.1037/0012-1649.42.1.70 [DOI] [PubMed] [Google Scholar]
  60. Stevenson HW, Chen C, & Uttal DH (1990). Beliefs and achievement: A study of black, white, and Hispanic children. Child Development, 61, 508–523. doi: 10.1111/j.1467-8624.1990.tb02796.x [DOI] [PubMed] [Google Scholar]
  61. Tabachnick BG, & Fidell LS (2012). Using multivariate statistics (6th ed.). New York, NY: Pearson. [Google Scholar]
  62. Tan ET, & Goldberg WA (2009). Parental school involvement in relation to children’s grades and adaptation to school. Journal of Applied Developmental Psychology, 30, 442–453. doi: 10.1016/j.appdev.2008.12.023 [DOI] [Google Scholar]
  63. Tiedemann J (2000). Parents’ gender stereotypes and teachers’ beliefs as predictors of children’s concept of their mathematical ability in elementary school. Journal of Educational Psychology, 92, 144–151. doi: 10.1037/0022-0663.92.1.144 [DOI] [Google Scholar]
  64. Trautwein U, Lüdtke O, Roberts BW, Schnyder I, & Niggli A (2009). Different forces, same consequence: conscientiousness and competence beliefs are independent predictors of academic effort and achievement. Journal of Personality and Social Psychology, 97, 1115–1128. doi: 10.1037/a0017048 [DOI] [PubMed] [Google Scholar]
  65. Upadyaya K, & Eccles J (2015). Do teachers’ perceptions of children’s math and reading related ability and effort predict children’s self-concept of ability in math and reading? Educational Psychology, 35, 110–127. doi: 10.1080/01443410.2014.915927 [DOI] [Google Scholar]
  66. Van Zanden B, Marsh HW, Seaton M, Parker PD, Guo J, & Duineveld JJ (2017). How well do parents know their adolescent children? Parent inferences of student self-concepts reflect dimensional comparison processes. Learning and Instruction, 47, 25–32. doi: 10.1016/j.learninstruc.2016.09.002 [DOI] [Google Scholar]
  67. Wagner BM, & Phillips DA (1992). Beyond beliefs: Parent and child behaviors and children’s perceived academic competence. Child Development, 63, 1380–1391. doi: 10.1111/j.1467-8624.1992.tb01702.x [DOI] [PubMed] [Google Scholar]
  68. White IR, Royston P, & Wood AM (2011). Multiple imputation using chained equations: Issues and guidance for practice. Statistics in Medicine, 30, 377–399. doi: 10.1002/sim.4067 [DOI] [PubMed] [Google Scholar]
  69. Wigfield A, & Eccles JS (1994). Children’s competence beliefs, achievement values, and general self-esteem change across elementary and middle school. The Journal of Early Adolescence, 14, 107–138. doi: 10.1177/027243169401400203 [DOI] [Google Scholar]
  70. Wigfield A, & Eccles JS (2000). Expectancy–value theory of achievement motivation. Contemporary Educational Psychology, 25, 68–81. doi: 10.1006/ceps.1999.1015 [DOI] [PubMed] [Google Scholar]
  71. Wigfield A, & Eccles JS (2002). The development of competence beliefs, expectancies for success, and achievement values from childhood through adolescence. In Wigfield A & Eccles JS (Eds), Development of achievement motivation (pp. 91–120). San Diego, CA: Academic Press. doi: 10.1016/B978-012750053-9/50006-1 [DOI] [Google Scholar]
  72. Wigfield A, Eccles JS, Yoon KS, Harold RD, Arbreton AJ, Freedman-Doan C, & Blumenfeld PC (1997). Change in children’s competence beliefs and subjective task values across the elementary school years: A 3-year study. Journal of Educational Psychology, 89, 451–469. doi: 10.1037/0022-0663.89.3.451 [DOI] [Google Scholar]
  73. Wolff F, Helm F, & Möller J (2018). Testing the dimensional comparison theory: When do students prefer dimensional comparisons to social and temporal comparisons? Social Psychology of Education, 21, 875–895. doi: 10.1007/s11218-018-9441-2 [DOI] [Google Scholar]
  74. Woodcock R, & Mather N (1989). Woodcock-Johnson Psycho-Educational Battery- Revised. Chicago, IL: Riverside Publishing. [Google Scholar]
  75. Yamamoto Y, & Holloway SD (2010). Parental expectations and children’s academic performance in sociocultural context. Educational Psychology Review, 22(3), 189–214. doi: 10.1007/s10648-010-9121-z [DOI] [Google Scholar]
  76. Zimmerman BJ (2000). Self-efficacy: An essential motive to learn. Contemporary Educational Psychology, 25, 82–91. doi: 10.1006/ceps.1999.1016 [DOI] [PubMed] [Google Scholar]

RESOURCES