The role of mothers' beliefs in students' self-concept of ability development

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ The role of mothers' beliefs in students' self-concept of ability development © 2018 Elsevier Inc Accepted version (Final draft) Pesu, Laura; Aunola, Kaisa; Viljaranta, Jaana; Hirvonen, Riikka; Kiuru, Noona Pesu, L., Aunola, K., Viljaranta, J., Hirvonen, R., & Kiuru, N. (2018). The role of mothers' beliefs in students' self-concept of ability development. Learning and Individual Differences, 65(2018), 230-240. https://doi.org/10.1016/j.lindif.2018.05.013 2018 MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 1 Abstract The aim of this study was to investigate whether the child-related competence beliefs of mothers are associated with the development of Finnish adolescents’ self-concept of mathematics and literacy ability during their transition from primary to lower secondary school and whether these associations depend on adolescents’ level of performance. The results showed that, first, adolescents’ self-concept of mathematics and literacy ability decreased over time. The impact of maternal beliefs on the linear trend of the self-concept of mathematics and literacy ability was dependent on the level of students’ performance. Mothers’ high beliefs buffered against the decrease in adolescents’ self-concept of ability in mathematics, but only among high-performing adolescents. In turn, mothers’ high beliefs in adolescents’ literacy ability were detrimental to the development of low-performing adolescents’ self-concept of ability in literacy, whereas mothers’ beliefs had no effect on the change in the self-concept of students with average or high literacy performance. Keywords: self-concept of ability, mothers’ beliefs, transition to lower secondary school, mathematics, literacy MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 2 The Role of Mothers’ Beliefs in Students’ Self-Concept of Ability Development Previous studies have revealed that students’ self-concept of ability is an important predictor of their behavior (Durik, Vida, & Eccles, 2006). The self-concept of students’ ability has been defined as students’ perception of their current competence in a particular subject area (Eccles, 2005), and it has been found to be associated with, for example, their academic performance (Valentine, DuBois, & Cooper, 2004) and choice of college major (Musu-Gillette, Wigfield, Harring, & Eccles, 2015). One of the most important factors affecting the development of selfconcept of ability is parental beliefs about their children’s abilities (Frome & Eccles, 1998; Gniewosz, Eccles, & Noack, 2012), and this effect has been found among primary (Frome & Eccles, 1998; Spinath & Spinath, 2005; Simpkins, Fredricks, & Eccles, 2012) and secondary school children (Fredricks & Eccles, 2002; Simpkins et al., 2012), as well as during the transition to lower secondary school (Gniewosz et al., 2012). Recent evidence, however, suggests that the impact of parental beliefs may vary depending on students’ level of performance. There is some indication that high-performing students may be more prone to being affected by mothers’ beliefs than low-performing students (Pesu, Viljaranta, & Aunola, 2016a; Pesu, Aunola, Viljaranta, & Nurmi, 2016b). However, more evidence is needed on the differential impact of parental support on lowand high-performing students. As far as we know, only two prior studies have investigated the moderating role of students’ performance level in the associations between mothers’ beliefs and students’ selfconcepts of ability. The first of these studies focused on first grade children (Pesu et al., 2016a), while the second focused on lower secondary school students (Pesu et al., 2016b). Besides the obvious need to replicate these studies’ findings with other samples, neither of them focused on critical educational transitions. It has been argued that during school transitions, students are MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 3 particularly sensitive to external feedback, such as parental beliefs regarding their competencies (Ruble, 1994). Consequently, there is a need for a novel understanding of students’ self-concept of ability development and the role of mothers’ beliefs on this development during the different transition phases. The present study aims to examine the extent to which mothers’ beliefs about their children’s abilities predict their children’s self-concept of ability development in mathematics and literature during the critical transition from primary to lower secondary school and the possible moderating role of students’ level of performance in these associations. 1.1. Self-concept of Ability among Adolescents The self-concept of one’s ability refers to an individual’s perception of his or her competence in a certain domain (Eccles, 2005; Wigfield & Eccles, 2000). Previous research has shown that it is highly domain-specific and that children have distinct mathematics and verbal domains in the academic self-concept of abilities (Arens et al., 2011). Self-concept of ability has been shown to impact school achievement and skill development in both the domains of mathematics and literacy: the higher the self-concept of ability in a certain domain (e.g., in mathematics), the higher the subsequent performance in this domain (Marsh & Martin, 2011; Marsh, Trautwein, Lüdtke, Köller, & Baumert, 2005; Valentine et al., 2004). Children tend to have overly positive perceptions of their abilities when they enter school and during the first school year (Aunola, Leskinen, Onatsu-Arvilommi, & Nurmi, 2002; Wigfield & Eccles, 2000), but these perceptions tend to become more realistic and more negative during later school years (Dweck, 2002; Jacobs, Lanza, Osgood, Eccles, & Wigfield, 2002). The stage–environment fit model (see Eccles et al., 1993) focuses on the impact of school transitions on adolescent development and states that a poor fit between changes in individual (e.g., an intensified need for autonomy) and contextual (e.g., stricter grading practices) MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 4 levels may lead to a decrease in ability beliefs during a school transition. Accordingly, previous research has documented that students’ self-concepts of abilities decrease, on average, during the transition to lower secondary school (Gniewosz et al., 2012). It has been suggested that this declining trend is due to the fact that the transition to lower secondary school comes with changes in adolescents’ social relationships as well as in the ways in which adolescents are given feedback in school (see Wigfield, Eccles, Schiefele, Roeser, & Davis-Kean, 2006). During school transitions, adolescents are also faced with many other changes, including new academic challenges: courses become increasingly difficult, students move from a classroom teacher system to a subject teacher system, and there is a change in the reference group to which a student compares his/her skills. Moreover, it has been argued that during school transitions, students are especially susceptible to external feedback, such as parental beliefs concerning their competencies (Ruble, 1994). From this point of view, the transition from primary to lower secondary school can be assumed to be a particularly important period in the development of the self-concept of one’s ability and the factors associated with this development. 1.2. Associations between Parental Beliefs and Students’ Self-concept of Abilities Parental beliefs have been suggested to play an important role in the formation of students’ ability beliefs (Eccles, 1993). Eccles et al.’s expectancy-value model proposes some mechanisms through which parental beliefs are associated with students’ self-concepts of abilities (Eccles, 1993; Eccles et al., 1983; Simpkins et al., 2012). Parents may directly communicate their beliefs to their offspring by encouraging their children to do better in school or by giving them positive feedback (Eccles, 1993). Parents may also indirectly convey their beliefs to their children through the ways in which they behave with their children (Eccles, 1993). There is also empirical evidence showing that parents’ beliefs about their children’s abilities have an impact on these MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 5 children’s ability perceptions (Eccles, 1993). For example, in studies by Gniewosz et al. (2012, 2014), parents’ perceptions of their children’s abilities were found to have a positive impact on the domain-specific self-concept of ability among sixth and seventh grade students (Gniewosz et al., 2012) and among fifth to seventh graders (Gniewosz et al., 2014). Although not yet widely examined, it has recently been suggested that the impact of adults’ beliefs on students’ selfconcept of ability development may be dependent on students’ level of performance. Pesu et al. (2016a) found that the role of teachers’ beliefs in the development of first grade students’ selfconcept of ability was particularly evident among high-performing students in both the domains of mathematics and literacy. Similarly, mothers’ beliefs were positively related to the development of their children’s self-concept of mathematics ability among high-performing students in grades seven to nine, but less so among low-performing students (Pesu et al., 2016b). One theoretical proposal that lends support to the assumption that lowand highperforming students may be differently impacted by adults’ beliefs is self-verification theory (Kwang & Swann, 2010). This theory proposes that people prefer others to see them in the same way in which they see themselves, even when a person’s self-perception is negative. People seek self-verification because it makes the world seem more coherent and predictable (Kwang & Swann, 2010). Thus, according to self-verification theory, people who see their abilities as low, for example, in mathematics, have a preference for others to see their abilities in mathematics as low. By contrast, people who consider their abilities in mathematics as high prefer others to perceive these abilities as high. It is thus possible that high-performing students are more likely to benefit from positive teacher and parental beliefs than low-performing students because positive beliefs fit better with their perceptions of themselves. Low-performing students might have strong negative self-perceptions, which would contradict positive beliefs expressed by MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 6 teachers and parents; this lack of fit between self-perceptions and external evaluations may explain why teachers’ and parents’ positive beliefs have a less positive impact on lowperforming students’ self-concept of ability (see Pesu et al., 2016b). Moreover, it is possible that for low-performing students whose self-concepts of abilities are low, the negative beliefs of adults would be in line with these students’ self-concepts and, thus, may have less detrimental impacts on their self-concepts than they might have on those of high-performing students. Another relevant theory is self-enhancement theory, which proposes that people have a need to see themselves positively (for a review, see Leary, 2007). It has been argued that the selfenhancement motive is a “cornerstone” of psychological activity (Sedikides & Gregg, 2008). According to self-enhancement theory, people tend to engage in self-serving biases that presumably enable them to maintain positive conceptions of themselves. An example of this kind of bias is the tendency to attribute positive outcomes to the self and negative outcomes to external factors (e.g., Blaine & Crocker, 1993). In light of self-enhancement theory, students might have a need to obtain a positive view of themselves and, in this case, would use the positive feedback from adults to serve this need. Based on this theory, mothers’ beliefs should be positively connected to students’ self-concept of ability, regardless of students’ performance level. Previous research has shown that, at the mean level, parental beliefs have a positive effect on students’ self-concept of ability (e.g., Eccles, 1993; Gniewosz et al., 2012). Thus, an important next step in the research on self-concept of ability would be to study whether the predictive effects of parental beliefs are actually similar for students at all ability levels. The present study focuses on mothers’ beliefs as a predictor. It has previously been argued that the role of teachers’ and parents’ beliefs in students’ self-concept of ability may be particularly MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 7 evident among high-performing students (Pesu et al., 2016a, 2016b). However, this issue represents an under-studied area, and there is a need for further investigation. 1.3. The Finnish School System The present study was conducted in Finland. In the Finnish educational system, children start school by attending pre-school in the year they turn six. In the year children turn seven, they go to compulsory comprehensive school, which is divided into primary school (grades 1–6) and lower secondary school (grades 7–9). Thus, the transition from primary school to lower secondary school takes place between grades six and seven. In the Finnish school system, the transition from primary to lower secondary school is the first remarkable transition for students. It includes some important changes in the school environment, such as changing school and moving from a classroom teacher system to a subject teacher system. In lower secondary school, students also start to study new school subjects. Moreover, a change of school also means changes in peer relations. In Finnish primary and lower secondary school, students do not need to decide whether to take higheror lower-level courses. Instead, they all follow the same curriculum and are taught at the same academic level. This characteristic of the Finnish school system is different from, for example, the system in Germany, where students must decide which achievement-based secondary school track they will take as early as 10 years of age (Gniewosz & Noack, 2012). In Finland, as compulsory courses in lower secondary school are at the same level for everyone, both highand low-performing students study in the same classrooms. Moreover, students need to decide whether to go to vocational school or upper secondary school after grade 9, when they are 15. 1.4. Research Questions and Hypotheses MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 8 As described above, there has been a dearth of research on the connections between mothers’ beliefs and children’s self-concept of ability development in mathematics and literacy during the critical transition from primary to lower secondary school and the possible role of students’ level of performance in these associations. Consequently, the present study focuses on the following research questions: 1) How do students’ self-concepts of their mathematics and literacy abilities change during the transition to lower secondary school? We hypothesized that students’ self-concept of ability declines during grades six and seven in both mathematics and literacy (Cole et al., 2001). 2) To what extent are mothers’ beliefs about their children’s abilities associated with the levels of students’ self-concept in mathematics and literacy ability, on one hand, and the development of their self-concept of ability during the transition from primary to lower secondary school, on the other? Based on earlier literature, we hypothesized a positive association between mothers’ beliefs and students’ level and developmental trend regarding selftheir concept of ability in both mathematics and literacy (e.g., Frome & Eccles, 1998; Jacobs et al., 2002; Pesu et al., 2016b). 3) Does the association between mothers’ beliefs and students’ self-concept of ability development differ on the basis of students’ level of performance? We hypothesized that the positive association between mothers’ beliefs about their children’s abilities and students’ selfconcept of ability development would be stronger among high-performing adolescents than among low-performing adolescents. This hypothesis was based on self-verification theory, which lends support to the assumption that high-performing students would be more likely to benefit from positive maternal beliefs than low-performing students, as these positive beliefs fit better with their self-perceptions (Kwang & Swann, 2010). Following the same line of reasoning, we MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 15 set to 0, 1, 2, and 3, respectively. Second, the adolescents’ initial level of performance in literacy and mathematics and their mothers’ beliefs concerning their literacy and mathematics abilities, along with the interaction term Performance × Mother’s belief (the interaction term was calculated as the multiplication of the standardized performance and belief variables), were added to the model to examine whether individual variation in the initial level and changes in the literacy/mathematics self-concept of ability would be predicted by these variables. At this stage, the impacts of the adolescents’ gender and their mothers’ level of education on the level and change in self-concept were controlled for. To interpret the nature of the interactions between performance and mothers’ beliefs on children’s self-concepts in greater detail, we used the Johnson-Neyman Technique (Preacher, Curran, & Bauer, 2006). This technique provides the regions of significance for the range of the moderator variable z (i.e., student performance), within which the simple slope of the dependent variable y (i.e., self-concept) on the independent variable x (i.e., mothers’ beliefs) is significantly different from zero at the chosen α-level. As a result, two values describing the upper and lower boundaries of the region of significance were calculated and plotted. The effect of mothers’ beliefs on the initial levels/changes in students’ self-concept was significant at the values of the students’ performance, which were less than the lower boundary and greater than the upper boundary, while the effect of the mothers’ beliefs was non-significant at the values of the students’ performance falling within the region. The analyses were carried out using the Mplus statistical package version 8.0 (Muthén & Muthén, 1998–2017). The model parameters were estimated using robust maximum likelihood (MLR estimator). Based on the missing at random (MAR) assumption, the full information maximum likelihood estimation was used, which enables the use of all the information in the data without imputation (Muthén & Muthén, 1998– MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 16 2015). Missingness at random (MAR) is a weaker condition for missing data than missingness completely at random (MCAR). In the MAR situation, missingness does not depend on the unmeasured variables, although it can depend on the values of the observed variables included in the analyses (Little, 1988). The goodness-of-fit of the estimated models was evaluated with five indicators: χ2-test, Bentler’s (1990) comparative fit index (CFI), the Tucker–Lewis Index (TLI), the root mean square error of approximation (RMSEA), and the standardized root mean square residual (SRMR). According to Hu and Bentler (1999), values above .95 for TLI and CFI, values below .06 for RMSEA, and values below .08 for SRMR can be considered as indicating a good fit between the hypothesized model and the observed data. 3. Results The means (Ms), standard deviations (SDs), and Pearson product-moment correlations of the study variables are shown in Table 1. ----------------Insert Table 1 about here------------------ 3.1. Self-Concept of Mathematics Ability First, the unconditional linear growth model comprising two growth components, i.e., the intercept growth factor (level) and the linear growth rate (linear trend), of the self-concept of mathematics ability was estimated. The intercept was defined as the initial status of the selfconcept of mathematics ability. The residual variances of the observed self-concept variables were allowed to be freely estimated. Moreover, the residuals of one of the self-concept items (i.e., difficulty of math) at different time points were allowed to correlate with each other. The model fitted the data well: χ2 (59, N = 814) = 253.66, p < .001; CFI = 0.96; TLI = 0.96; RMSEA = 0.06; SRMR = 0.03. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 17 The results showed that, at the mean level, the self-concept of mathematics ability decreased during the transition from primary school to lower secondary school (mean of the linear trend = -0.04, s.e. = 0.01, p < .001). The results further showed that there was statistically significant individual variation in both the level (Var = 0.54, s.e. = 0.03, p < .001) and the linear trend (Var = 0.03, s.e. = 0.00, p < .001) of the self-concept of mathematics ability. Second, the predictor variables (i.e., mathematics performance at T1, maternal beliefs at T1) and the interaction term between them were added to the previous model as predictors of the level and linear trend. To control for the impact of the initial level of the self-concept of mathematics ability on its trend, the path from the level component to the trend component was also estimated. Moreover, to control for the impact of gender and mothers’ level of education, paths from these variables to the level and trend components were estimated. The fit of the model was good: χ2 (109, N = 814) = 362.91, p < .001; CFI = 0.96; TLI = 0.96; RMSEA = 0.05; SRMR = 0.03. The results are shown in Figure 1. --------------Insert Figure 1 about here----------------- The results (Figure 1) showed that both the adolescents’ performance in mathematics and their mothers’ beliefs were associated with the level of the adolescents’ self-concept of mathematics ability: the higher the adolescents’ level of mathematics performance and the higher the mothers’ beliefs about their children’s mathematics performance, the higher the self-concept of mathematics ability the adolescents reported at the beginning of the sixth grade. However, the results further showed that the interaction term performance × maternal belief was also associated with the adolescents’ level of self-concept of mathematics ability. To interpret this interaction further, the Johnson-Neyman Technique was used. The results are shown in Figure 2. -------------------Figure 2 about here------------------- MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 18 The results (see Figure 2) demonstrate that mothers’ beliefs predicted statistically significantly both lowand high-performing students’ level of self-concept of mathematics ability: the higher the mothers’ beliefs, the higher the students’ initial level of self-concept of mathematics ability. However, mothers’ beliefs had a stronger effect on highthan lowperforming students’ level of self-concept of mathematics ability in Grade 6 (i.e., among highperforming students, both confidence intervals of the regression coefficient of mothers’ beliefs on students’ self-concept of ability in mathematics were further above zero than those of lowachieving students). The results further showed that the linear trend of self-concept of mathematics ability was predicted not only by the level of mathematics performance, but also by the interaction term performance × maternal belief. To interpret this interaction further, the Johnson-Neyman Technique was used. The results are shown in Figure 3. ----------- Insert Figure 3 ------------ As shown in Figure 3, the effect of mothers’ beliefs on the linear change in students’ selfconcept of ability in mathematics was statistically significant and positive only when student performance was above average (values above +0.2 SD): the higher the mothers’ beliefs, the less the decrease in self-concept of ability during the transition from primary to lower secondary school among adolescents whose math performance was above average or high. However, when the students’ mathematics performance was average or poor, their mothers’ beliefs showed no effect on the linear change in their self-concept of ability in mathematics. 3.2. Self-Concept of Literacy Ability MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 19 Next, an unconditional linear growth model comprising the two growth components (i.e., the intercept growth factor (level) and the linear growth rate (linear trend)) of the self-concept of literacy ability was estimated. The intercept was defined as the initial status of the self-concept of literacy ability. The residual variances of the observed self-concept variables were allowed to be freely estimated. Moreover, the residuals of one of the self-concept items (i.e., difficulty of literacy) at different time points were allowed to correlate with each other. The fit of the model was satisfactory: χ2 (115, N = 814) = 1013.62, p < .001; CFI = 0.82; TLI = 0.81; RMSEA = 0.10; SRMR = 0.06. The inspection of the modification indices suggested that estimating the residual correlations of the identical items of the self-concept between different time points would improve the fit of the model. Following these specifications, the fit of the model was good: χ2 (105, N = 814) = 385.65, p < .001; CFI = 0.94; TLI = 0.94; RMSEA = 0.06; SRMR = 0.05. The results showed that, at the mean level, students’ self-concept of literacy ability decreased during the transition from primary to lower secondary school (the mean of the trend = -0.04, s.e. = 0.01, p < .001). The results further showed that there was statistically significant individual variation in the level (Var = 0.32, s.e. = 0.02, p < .001), but not in the trend, of students’ self-concept of literacy ability (Var = 0.01, s.e. < 0.001, p = .15). Finally, predictors were added to the model to explain the level and trend of students’ self-concept of literacy ability. After fixing the non-significant path from the interaction term performance × maternal belief to the level of reading self-concept to zero, the fit of the model was: χ2 (178, N = 814) = 611.87, p < .001; CFI = 0.93; TLI = 0.92; RMSEA = 0.05; SRMR = 0.06. The results are shown in Figure 4. -----Insert Figure 4 about here---- MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 20 The results showed that both maternal beliefs and the level of literacy performance predicted the initial level of students’ self-concept of literacy ability: the higher the adolescents’ literacy performance and the higher the mothers’ beliefs concerning their children’s literacy ability, the higher the adolescents’ self-concept of literacy ability at the beginning of the sixth grade. The small variation in the linear trend of students’ self-concept of literacy ability, in turn, was only predicted by the interaction term performance × maternal belief. To interpret this interaction further, the Johnson-Neyman Technique was used. The results are shown in Figure 5. -----------Insert Figure 5 about here-------- As shown in Figure 5, the effect of mothers’ beliefs on the linear change in students’ selfconcept of ability in literacy was statistically significant and negative when student performance was below average (values below -1.4 SD): the higher the mothers’ beliefs, the more the selfconcept of ability in literacy decreased during the transition to lower secondary school among students whose literacy performance was poor. However, when the students’ literacy performance was average or high, their mothers’ beliefs showed no effect on the linear change in their self-concept of ability in literacy. 4. Discussion This study examined the development of self-concepts of mathematics and literacy abilities during the transition to lower secondary school in Finland—that is, from the beginning of grade six to the end of grade seven—and the role of mothers’ beliefs about their children’s abilities in this development. Our main interest was whether the associations between mothers’ beliefs and students’ self-concept of ability development differ on the basis of students’ level of performance. The results showed that, overall, students’ self-concepts of mathematics and literacy abilities declined during the transition to lower secondary school. As an important MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 21 contribution to the previous literature, the findings of the present study further revealed that the effect of mothers’ beliefs on changes in adolescents’ self-concepts varied with the level of student performance. Thus, although the adolescents’ self-concept of mathematics ability decreased from grades 6 to 7, their mothers’ high beliefs buffered against this decrease among above-average and high-performing adolescents. This kind of buffering effect was not evident among low-performing adolescents. In literacy, the results were somewhat different. Mothers’ beliefs had no significant effect on the change in the self-concept of literacy ability among adolescents with average or high literacy performance. Instead, mothers’ high beliefs in lowerperforming adolescents’ abilities were related, although weakly, to a steeper decrease in the selfconcept of literacy ability. 4.1. The Developmental Trend of Adolescents’ Self-concept of Ability during the Transition from Primary to Lower Secondary School The first aim of the study was to examine whether adolescents’ self-concepts of abilities in literacy and mathematics change, on average, during the transition from primary to lower secondary school. The results lend support to Hypothesis 1 by showing that the adolescents’ selfconcepts, both of their mathematics and literacy abilities, declined during the transition from primary to lower secondary school. In the Finnish context, our results are also in line with earlier findings reported in other educational systems and cultural settings, for example, in the U.S. (Fredricks & Eccles, 2002; Gniewosz et al., 2012; Jacobs et al., 2002). As suggested by the stage–environment fit model (Eccles et al., 1993), the multiple changes that adolescents face during school transitions may lead to a poor fit between changes in the individual (e.g., an intensified need for autonomy) and contextual (e.g., stricter grading practices) levels. This poor fit may explain the decrease in students’ ability beliefs during school transitions. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 22 4.2. The Role of Mothers’ Beliefs in the Development of Adolescents’ Self-concepts of Ability The results concerning the relationship between mothers’ beliefs and adolescents’ self-concepts of ability showed that, first, both in mathematics and literacy, mothers’ beliefs were positively related to the initial level of their children’s self-concepts. Thus, the more positive beliefs the mother had, the higher were the adolescent’s self-concepts of ability in the sixth grade, and vice versa. This result is consistent with the expectancy-value theory of Eccles et al. (Eccles Parsons et al., 1982; Wigfield & Cambria, 2010) and with previous empirical research on this topic (Frome & Eccles, 1998; Gniewosz et al., 2014; Lau & Pun, 1999; McGrath & Repetti, 2000) that maternal beliefs have a significant positive effect on students’ self-concept of ability. Mothers’ beliefs may affect students’ self-concept of ability through different kinds of mechanisms. First, it has been suggested that mothers may communicate their beliefs to their children by encouraging them to do better in school or by giving them positive feedback on their effort (Eccles, 1993). Second, mothers may also indirectly convey their beliefs to their children through the ways they behave with their children (Eccles, 1993). Overall, the result that the mothers’ beliefs were positively associated with the initial level of the adolescents’ self-concepts in mathematics and literacy, regardless of the level of the students’ performance, is in line with self-enhancement theory, which suggests that people have a tendency and a need to see themselves positively (Leary, 2007). Another explanation for the result is, however, that a third factor—adolescents’ actual ability level—might explain both positive maternal beliefs and adolescents’ own self-concepts. It has been previously found that both students and parents use the information from grades in forming their competence beliefs (Gniewosz et al., 2014). MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 23 Second, the results showed that mothers’ beliefs of their adolescent children’s ability were related to the changes in these adolescents’ self-concepts of ability. However, this effect was found to vary according to the adolescents’ performance level and the subject domain. In the domain of mathematics, mothers’ high beliefs were found to buffer against the decrease in adolescents’ self-concept of mathematics ability, but only among high-performing adolescents. Thus, the more positive beliefs the mothers had about high-performing adolescents’ abilities, the less these students’ self-concept in mathematics decreased during the transition to lower secondary school. In turn, the less positive the mothers’ beliefs, the steeper was the decrease in high-performing students’ self-concept of ability development. Among low-performing students, mothers’ beliefs were not related to the changes in self-concept of mathematics ability. Overall, this pattern of findings is in line with our first hypothesis for research question 3, which suggested that the positive association between mothers’ beliefs about their children’s abilities and students’ self-concept of ability development is stronger among high-performing adolescents than among low-performing adolescents. One explanation for this result draws on self-verification theory (Kwang & Swann, 2010), according to which people prefer others to see them in the same way that they see themselves, even when their self-perception is negative. Consequently, it may be that highperforming students benefit more from their mothers’ positive beliefs than low-performing students, and as such, positive beliefs fit better with these students’ perceptions of themselves. Similarly, negative maternal beliefs may be less detrimental to low-performing students than to high-performing students, as low maternal beliefs fit with low-performing students’ selfperceptions. Besides self-verification theory, there are also alternative explanations for the results. For example, one possible explanation is that students’ cognitive abilities affect the way MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 24 they perceive and interpret the beliefs of their mothers (Bohlmann & Weinstein, 2013). Highperforming students may have better cognitive ability to more accurately perceive and interpret their mothers’ beliefs and, consequently, benefit more from positive maternal beliefs (see also Pesu et al., 2016b). It may also be that above-average and high-performing students feel encouraged by their mothers’ positive beliefs, whereas low-performing students may feel discouraged or pressured by positive maternal beliefs. Further investigation is needed to examine the suggested differences in the thinking patterns and interpretation mechanisms between lowand high-performing students. Mothers’ high beliefs in adolescents’ literacy abilities, as opposed to mathematics abilities, were related to a steeper decrease in low-performing adolescents’ self-concept of ability. In other words, decreases in low-performing students’ self-concept were less dramatic if their mothers had less positive beliefs about their abilities. Among students with average or high literacy performance, mothers’ beliefs had no effect on the change in the self-concept of literacy ability. This result was somewhat unexpected and, in contrast to our hypothesis, suggests that high-performing students are more prone to being impacted by maternal beliefs than lowperforming students. One explanation for this result might be that, as suggested by selfverification theory, mothers’ positive beliefs concerning low-performing students are incongruent with the students’ perceptions of their own abilities and that this disparity then has negative consequences for the development of these students’ self-concept of literacy ability. However, when interpreting this particular finding, it should be noted that since the variance in the trend of the self-concept of literacy ability was not statistically significant, the impact of maternal beliefs on that trend should be interpreted with caution and as tentative only. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 31 on young adolescents’ experiences in schools and in families. American Psychologist, 48, 90–101. doi: 10.1037/0003-066X.48.2.90 Eccles Parsons, J., Adler, T. F., & Kaczala, C. M. (1982). Socialization of achievement attitudes and beliefs: Parental influences. Child Development, 53, 310–321. doi: 10.2307/1128973 Eccles, J. S., & 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 Fredricks, J. A., & Eccles, J. S. (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 Frome, P. M., & Eccles, J. S. (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 Gniewosz, B., Eccles, J. S., & Noack, P. (2012). Secondary school transition and the use of different sources of information for the construction of the academic self-concept. Social Development, 21, 537–557. doi: 10.1111/j.1467-9507.2011.00635.x Gniewosz, B., Eccles, J. S., & Noack, P. (2014). Early adolescents’ development of academic self-concept and intrinsic task value: The role of contextual feedback. Journal of Research on Adolescence, 25, 1–15. doi: 10.1111/jora.12140. Holopainen, L., Kairaluoma, L., Nevala, J., Ahonen, T., & Aro, M. (2004). Lukivaikeuksien seulontamenetelmä nuorille ja aikuisille [Dyslexia Screening Test for Youth and Adults]. Jyväskylä: University of Jyväskylä. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 32 Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1-55. doi: 10.1080/10705519909540118 Jacobs, J. E., Lanza, S., Osgood, D. W., Eccles, J. S., & 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 Kiuru, N., Haverinen, K., Salmela-Aro, K., Nurmi, J-E., Savolainen, H. & Holopainen, L. (2011). Students with reading and spelling disabilities: Peer groups and educational attainment in secondary education. Journal of Learning Disabilities, 44, 556–569. doi: 10.1177/0022219410392043 Kwang, T., & Swann, W. B. (2010). Do people embrace praise even when they feel unworthy? A review of critical tests of self-enhancement versus self-verification. Personality and Social Psychology Review, 14, 263–280. doi: 10.1177/1088868310365876 Landerl, K., Wimmer, H., & Moser, E. (1997). Salzburger Leseund rechtschreibtest [Salzburg Reading and Spelling Test]. Bern: Huber. Lau, S., & Pun, K-T. (1999). Parental evaluations and their agreement: Relationship with children’s self-concepts. Social Behavior and Personality, 27, 639–650. doi: 10.2224/sbp.1999.27.6.639 Leary, M. R. (2007). Motivational and emotional aspects of the self. Annual Review of Psychology, 58, 317-344. doi: 10.1146/annurev.psych.58.110405.085658 Little, R. J. A. (1988). A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association, 83, 1198–1202. doi: 10.1080/01621459.1988.10478722 MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 33 Marsh, H. W., & Martin, A. J. (2011). Academic self‐concept and academic achievement: Relations and causal ordering. British Journal of Educational Psychology, 81, 59–77. doi: 10.1348/000709910X503501 Marsh, H. W., 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.14678624.2005.00853.x McGrath, E. P., & Repetti, R. L. (2000). Mothers’ and fathers’ attitudes toward their children’s academic performance and children’s perceptions of their academic competence. Journal of Youth and Adolescence, 29, 713–723. doi: 10.1023/A:1026460007421 Midgley, C., & Edelin, K. C. (1998). Middle school reform and early adolescent well-being: The good news and the bad. Educational Psychologist, 33, 195–106. doi: 10.1207/s15326985ep3304_4 Musu-Gillette, L. E., Wigfield, A., Harring, J. R., & Eccles, J. S. (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 Official Statistics of Finland (2016a). Educational structure of population (e-publication). Helsinki, Finland: Statistics Finland. Retrieved from www.stat.fi/til/vkour/2014/vkour_2014_2015-11-05_tie_001_en.html Official Statistics of Finland (2016b). Families. Appendix Table 3: Families with underage children by type in 1950–2014 (e-publication). Helsinki, Finland: Statistics Finland. Retrieved from www.stat.fi/til/perh/2014/perh_2014_2015-05-28_tau_003_en.html MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 34 Paulson, S. E., & Sputa, C. L. (1996). Patterns of parenting during adolescence: Perceptions of adolescents and parents. Adolescence, 31, 369-382. Pesu, L., Viljaranta, J., & Aunola, K. (2016a). The role of parents’ and teachers’ beliefs in children’s self-concept development. Journal of Applied Developmental Psychology, 44, 63–71. doi: 10.1016/j.appdev.2016.03.001 Pesu, L., Aunola, K., Viljaranta, J., & Nurmi, J.-E. (2016b). The Development of Adolescents’ Self-concept of Ability through Grades 7-9 and the Role of Parental Beliefs. Frontline Learning Research, 4, 92–109. doi: 10.14786/flr.v4i3.249 Pichler, C., & Wimmer, L. (2006). Das salzburger Lesescreening 2-9. Handreichnung für Lehrerinnen und Lehrer. Based on Mayringer, H., & Wimmer, H. (2003). Salzburger Lesescreening für die Klassenstufen 1-4 and Auer, M., Gruber, G., Mayringer, H., & Wimmer, H. (2005). Salzburger Lesescreening für die Klassenstufen 5-8. Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent curve analysis. Journal of Educational and Behavioral Statistics, 31, 437-448. doi:10.3102/10769986031004437 Ruble, D. N. (1994). A phase model of transitions: Cognitive and motivational consequences. Advances in Experimental Social Psychology, 26, 163–214. doi: 10.1016/S0065-2601(08)60154-9 Räsänen, P., Salminen, J., Wilson, A. J., Aunio, P., & Dehaene, S. (2009). Computer-assisted intervention for children with low numeracy skills. Cognitive Development, 24, 450–472. doi: 10.1016/j.cogdev.2009.09.003 Savolainen, H., Ahonen, T., Aro, M., Tolvanen, A., & Holopainen, L. (2008). Reading comprehension, word reading and spelling as predictors of school achievement and MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 35 choice of secondary education. Learning and Instruction, 16, 201–210. doi: 10.1016/j.learninstruc.2007.09.017 Sedikides, C., & Gregg, A. P. (2008). Self-enhancement: Food for thought. Perspectives on Social Psychology, 3, 102-116. doi: 10.1111/j.1745-6916.2008.00068.x Simpkins, S. D., Fredricks, J. A., & Eccles, J. S. (2012). Charting the Eccles’ expectancy-value model from mothers’ beliefs in childhood to youths’ activities in adolescence. Developmental Psychology, 48, 1019–1032. doi: 10.1037/a0027468 Spinath, B., & Spinath, F. M. (2005). Longitudinal analysis of the link between learning motivation and competence beliefs among elementary school children. Learning and Instruction, 15, 87–102. doi: 10.1016/j.learninstruc.2005.04.008 Spinath, B., & Steinmayr, R. (2008). Longitudinal analysis of intrinsic motivation and competence beliefs: Is there a relation over time? Child Development, 79, 1555–1569. doi: 10.1111/j.1467-8624.2008.01205.x Starkey, P., & Klein, A. (2000). Fostering parental support for children's mathematical development: An intervention with Head Start families. Early Education and Development, 11, 659-81. doi: 10.1207/s15566935eed1105_7 Valentine, J. C., DuBois, D. L., & Cooper, H. (2004). The relation between self-beliefs and academic achievement: A meta-analytic review. Educational Psychologist, 39, 111– 133. doi: 10.1207/s15326985ep3902_3 Viljaranta, J., Lazarides, R., Aunola, K., Räikkönen, E., & Nurmi, J. E. (2015). The different role of mothers' and fathers' beliefs in the development of adolescents' mathematics and literacy task values. International Journal of Gender, Science and Technology, 7. 297– 317. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 36 Wigfield, A., & Cambria, J. (2010). Expectancy-value theory: Retrospective and prospective. In The decade ahead: Theoretical perspectives on motivation and achievement (pp. 3570). Emerald Group Publishing Limited. Wigfield, A., & Eccles, J. S. (2000). Expectancy-value theory of achievement motivation. Contemporary Educational Psychology, 25, 68–81. doi: 10.1006/ceps.1999.1015 Wigfield, A., Eccles, J. S., Schiefele, U., Roeser, R. W., & Davis-Kean, P. (2006). Development of achievement motivation. In N. Eisenberg, W. Damon, & R. M. Lerner (Eds.), Handbook of child psychology: Vol. 3, Social, emotional, and personality development (6th ed., pp. 933–1002). Hoboken, NJ: John Wiley. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 37 Figure captions. Figure 1. Latent Growth Curve Model with Predictors for Children’s Self-concept of Mathematics Ability (Standardized Estimates). Note. ***p < .001, **p < .01, *p < .05. Figure 2. The effect of mothers’ beliefs on the level of children’s self-concept of mathematics as a function of students’ level of math performance. The X-axis depicts a continuous range of student performance (moderator variable) values. The Y-axis depicts a continuous range of values for the adjusted effect of mothers’ beliefs on the initial level of students’ self-concept. The straight plot line represents values of the adjusted effect that correspond to the full range of all continuous values of student performance requested (i.e., -2 to +2). The curved lines above and below the plot line represent 95% confidence bands around the adjusted effect of mothers’ beliefs on the initial level of self-concept in mathematics. Figure 3. The effect of mothers’ beliefs on the linear change of children’s self-concept of mathematics as a function of students’ level of math performance. The X-axis depicts a continuous range of student performance (moderator variable) values. The Y-axis depicts a continuous range of values for the adjusted effect of mothers’ beliefs on the linear change of student self-concept. The straight plot line represents values of the adjusted effect that correspond to the full range of all continuous values of student performance requested (i.e., -2 to +2). The curved lines above and below the plot line represent 95% confidence bands around the adjusted effect of mothers’ beliefs on the linear change of self-concept in mathematics. When the standardized score of students’ level of math performance is over the value 0.2 (dashed line), the effect of mothers’ beliefs on the linear change of self-concept is positive and statistically significant, whereas when the value of math performance is lower than 0.2, mothers’ beliefs have no effects on the change of students’ self-concept. Figure 4. Latent Growth Curve Model with Predictors for Children’s Self-concept of Literacy Ability (Standardized Estimates). Note. ***p < .001, **p < .01, *p < .05. Figure 5. The effect of mothers’ beliefs on the linear change of children’s self-concept of literacy as a function of students’ reading performance. The X-axis depicts a continuous range of students’ reading performance (moderator variable) values. The Y-axis depicts a continuous range of values for the adjusted effect of mothers’ beliefs on the linear change of students’ selfconcept. The straight plot line represents values of the adjusted effect that correspond to the full range of all continuous values of student performance requested (i.e., -2 to +2). The curved lines above and below the plot line represent 95% confidence bands around the adjusted effect of mothers’ beliefs on the linear change of self-concept in literacy. When the standardized score of students’ level of reading performance is below the value -1.4 (dashed line), the effect of mothers’ beliefs on the linear change of self-concept is negative and statistically significant, whereas when the value of reading performance is higher than -1.4, mothers’ beliefs have no effects on the change of students’ self-concept. MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 38 MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 39 Table 1 Intercorrelations, Means, and Standard Deviations for the Study Variables Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 1. Self-concept literacy T1 2. Self-concept literacy T2 .71*** 3. Self-concept literacy T3 .68*** .74*** 4. Self-concept literacy T4 .65*** .66*** .72*** 5. Self-concept math T1 .35*** .25*** .20*** .21*** 6. Self-concept math T2 .29*** .31*** .23*** .23*** .78*** 7. Self-concept math T3 .32*** .28*** .33*** .29*** .72*** .76*** 8. Self-concept math T4 .26*** .23*** .25*** .32*** .63*** .71*** .76*** 9. Performance literacy T1 .49*** .48*** .45*** .50*** .20*** .20*** .22*** .19*** 10. Performance math T1 .21*** .21*** .15*** .20*** .47*** .49*** .47*** .42*** .47*** 11. Mother belief literacy T1 .47*** .44*** .41*** .43*** .16*** .19*** .21*** .23*** .52*** .30*** 12. Mother belief math T1 .27*** .22*** .17*** .22*** .59*** .58*** .56*** .50*** .36*** .55*** .55*** MOTHERS’ BELIEFS AND SELF-CONCEPT DEVELOPMENT 40 13. Gender -.12** -.17*** -.16*** -.13*** .23*** .20*** .16*** .11** -.22*** .10** -.33*** .07 M 3.89 3.88 3.89 3.78 3.84 3.79 3.82 3.74 .00 .00 .00 .00 SD 0.52 0.52 0.54 0.56 0.73 0.72 0.73 0.79 1.00 1.00 1.00 1.00 Note. *** p < .001. ** p < .01. * p < .05. T1 = time 1, T2= time 2, T3 = time 3, T4 = time 4. Literacy and mathematics performance and mothers’ beliefs are standardized variables.