Motivation,Legitimation,or Both? Reciprocal Effects of Parental Meritocratic Beliefs and Children’s Educational Performance in China

Abstract

Different research traditions have long held that parental beliefs motivate children’s performance. However, regarding meritocratic beliefs, sociologists often argue that meritocratic narratives legitimize and make sense of societal inequalities as justly deserved. Using the case of China, I tested these competing hypotheses of the relationship between parental meritocratic beliefs and children’s educational achievement. Parental beliefs about skills and hard work as predictors of higher grades were used. I analyzed data from the first and second waves of the China Educational Panel Survey. Autoregressive cross-lagged structural models indicated that parental meritocratic beliefs do not affect children’s educational performance but, rather, meritocratic beliefs are affected by academic results, suggesting their justificatory role. This pattern is much sharper in rural China, where traditional Chinese culture is preserved. The implications of meritocratic beliefs for a broader discussion of citizens’ beliefs about social inequalities and stratification are discussed.

Keywords

China educational achievement meritocratic beliefs

In modern societies, meritocracy is a guiding principle for allocation of rewards based on achievement rather than ascription (Heiserman and Simpson 2017; Mijs 2016b; Young 1958). Defenders of meritocratic ideals (Herrnstein and Murray 1996; Saunders 1995) indicate that it enables scarce resources to be efficiently allocated and incentivizes effort. However, unequal and undeserved starting positions lead to the understanding that meritocracy violates its own merit principle. It legitimizes societal inequalities as justly deserved, and misfortune becomes a personal failure (Mijs 2016b). Nevertheless, the comparative efforts of legitimization researchers have not produced longitudinal evidence to rule out bidirectional effects (Mijs 2016a, 2019). Given this context, the aim of the present study is to test these two perspectives simultaneously in the case of the Chinese educational system: how are parents’ meritocratic beliefs and children’s educational performance affected by each other? By understanding parental meritocratic beliefs in the educational system, I examine how meritocratic beliefs shape and are shaped by success at school. Parental meritocratic beliefs are “rule-like structures” (DiMaggio 1997:265) that could make sense of children’s outcomes or “vocabularies of motive” (Vaisey and Lizardo 2010:1599) that could affect them. This is an attempt not to set aside the relevance of meritocratic beliefs in China but to better understand their relationships with children’s educational performance. In addition, I examine how this dynamic varies between rural and urban China as an important cultural cleavage.

This question arises from the classic debate in the sociology of culture on the relationship between culture and action (Frye 2012; Lizardo et al. 2016; Vaisey 2009). However, as Vaisey and Valentino (2018) argue, cultural sociologists typically sideline beliefs. In this study, parental beliefs about meritocracy are shared narratives of how rewards are allocated in the educational system. Hence, by discussing the reciprocal effects of parental beliefs and educational outcomes, I engage in a broader discussion about culture and action literature.

Moreover, parental meritocratic beliefs could be considered cultural orientations of parenting (Weininger, Lareau, and Conley 2015), consequential for inequality and social reproduction. Cross-cultural studies in social psychology (e.g., Kinlaw, Kurtz-Costes, and Goldman-Fraser 2001; Stevenson, Lee, and Stigler 1986) have long characterized attributional beliefs of Chinese parents as emphasizing effort, while American and European parents emphasize skills. For the first time, this study examines how parental beliefs about effort and talents are related to children’s educational outcomes in a large and representative sample of Chinese students.

Parental beliefs about determinants of grades are relevant constructs about their views on reward allocation. Previous studies across social sciences (Pulfrey, Darnon, and Butera 2013; Schuman et al. 1985; Verniers, Martinot, and Dompnier 2016) have acknowledged that grades are a valid performance-contingent reward for multiple reasons. They could be considered part of the “currency of schools” (Deci, Koestner, and Ryan 2001:1) and “provide a clear analogue to a merit system of salary or wages” (Schuman et al. 1985:946). As explained by Pulfrey et al. (2013), grades are a linear function of task performance. Grades can also be understood as performance-contingent rewards when their symbolic value is taken into account (Harackiewicz, Manderlink, and Sansone 1984) because they have a clue value that signals excellence.

This article makes several contributions to the literature. First, I simultaneously test two competing hypotheses about the role of parental meritocratic beliefs in the educational system, using a theoretical rationale and rich longitudinal data. Sociologists tend to propose that meritocratic narratives legitimize inequalities (e.g., Lamont 2019; Mijs 2019), distancing themselves from motivational arguments. However, this theory has not been accurately tested. In the present study, autoregressive cross-lagged structural models are implemented using data from the two waves of the China Educational Panel Survey with the aim of disentangling the theorized reciprocal effects. This theoretical argument and research design could be applied in different societies and to understand the role of meritocratic beliefs in a myriad of fields. Parental beliefs about two determinants of grades are considered: skills and hard work. Although these factors are the central constructs of a meritocratic system (Heiserman and Simpson 2017; Mijs 2016b; Young 1958), they are not necessarily the only ones. Thus, we do not expect to answer the extent to which the educational system is meritocratic because that will imply incorporating a large array of factors, including nonmeritocratic determinants. Rather, we focus on the extent that these two factors explain and are explained by educational performance.

Second, the Chinese case is a strategic site for studying meritocratic beliefs due to cultural traditions and contemporary structural changes that exacerbate them. Paradoxically, there is no equivalent of the word meritocracy in Chinese. It is often translated as elitism (), which ignores its antecedents. However, Chinese culture has long-standing meritocratic norms, grounded in mainstream Chinese philosophies. It links educational institutions and exam-based qualifications for highly valued civil service jobs (Hannum et al. 2019). The selection of public servants based on both moral character and talent has been attributed to the earliest period of Chinese history (Xiao and Li 2013). Moreover, administrative performance provides the basis of merit-based inequality (Xie 2016). It is widely believed that the ruling class deserves its benefits as they work for the public good.

In addition, a recent study has shown that individuals from Western countries with higher levels of inequality tend to explain success in meritocratic terms (Mijs 2019). A similar conclusion has been drawn for Latin American countries (Bucca 2016) and using experimental evidence (Heiserman and Simpson 2017; Molina, Bucca, and Macy 2019). In the case of China, a high level of inequality (Xie and Zhou 2014) and social fluidity (Zhou and Xie 2019), by international standards, brings the justificatory role of meritocratic narratives, alongside their hypothesized motivational effect, into sharper relief. Empirical evidence has also challenged the conventional portrayal of Chinese people as accepting inequalities (Lei 2020): in 2014, most people in contemporary China believe that inequalities are large and economic distribution unfair. I expand this literature and incorporate the Chinese case in the literature of meritocratic beliefs through a focus on the educational system.

By meritocratic beliefs, I refer to the perception of reward distribution in the educational system (as it is) in contrast to individual preferences for inequality (what ought to be), which is an important distinction made by previous studies in social justice beliefs (e.g., Castillo 2011). Although meritocratic ideals are important normative principles in Chinese tradition, it is not clear whether Chinese parents perceive educational systems as such and whether these beliefs have cultural efficacy in the explanation of children’s performance. Throughout this article, I will use, for the sake of simplicity, meritocratic beliefs and perceived meritocracy as synonyms.

Finally, this study contributes to the literature by clarifying the debate about the cultural explanation of educational inequality. Values and beliefs are common explanations of racial gaps in the United States. Asian Americans outperform white Americans due to their cultural differences regarding the association between effort and success (Chen and Stevenson 1995; Hsin and Xie 2014; Liu and Xie 2016). Thus, parental beliefs have an effect on children’s performance. However, Lee and Zhou (2015) have expressed their concerns about the misperception of culture and Asian American achievement. They are skeptical about findings showing that Confucian values of hard work have explained Asian Americans’ academic achievement, because some migrant groups, such as Chinese and Koreans, are hyperselected. The findings reported in this study are particularly insightful for reevaluating the cultural explanation under different social contexts and without self-selection by cultural repertoire.

This article is structured as follows. First, in the theoretical framework, I contrast motivational and legitimization hypotheses to discuss how different traditions understand the role of meritocratic beliefs. Second, based on this theoretical reasoning, the analytic strategy is proposed. Data and variables are presented in the third section, which is followed by the results reported in the fourth section. Finally, the conclusion discusses these results in relation to the theoretical reasoning of the first sections and the implications of meritocratic beliefs for a broader discussion of citizens’ beliefs about social inequalities and stratification.

Theory

Meritocratic Beliefs as Motivation

Parental beliefs exert a large influence on children’s beliefs, attitudes, and behaviors (e.g., Bandura et al. 2001; Frye 2012). As Frye (2012:1592) argues, “familiar slogan can enter into cultural models and shape individual cognition.” Theories in sociology and social psychology posit that parents’ perceptions about meritocratic assessments and rewards may promote motivation and effort. As I will show in this section, several studies have suggested that meritocratic beliefs drive academic behaviors and outcomes.

Parental beliefs are rule-like structures (DiMaggio 1997) that set high educational expectations for their children and enhance educational achievement. In cultural sociology, declarative beliefs are predictors of future individual outcomes (Vaisey 2010; Vaisey and Lizardo 2010) and even bypass, interrupt, and overread unconscious dispositions (Martin and Desmond 2010). Regarding educational outcomes, Vaisey (2010) presents evidence of the explanatory power of educational aspirations on students’ decision making. Thus, educational ideals, expectations, and worldviews are consequential for individuals’ choices. This theoretical development brings back the motivational role of culture on the explanation of action.¹ In this framework, parental beliefs can be understood as vocabularies of motive (Vaisey and Lizardo 2010).

Different social psychological theories suggest an effect of parental meritocratic beliefs on achievement outcomes (for a review, see Wigfield et al. 2006). Bandura’s (1997) self-efficacy theory proposes that parents provide social models and vicarious experiences, influencing individuals’ level of effort, persistence, and choice of activities. Parental meritocratic beliefs are resources for student behavior and posit the perceived payoffs of abilities and effort in direct relation with students’ own potential. Parents’ persuasion affects self-efficacy beliefs through reaffirmation of capabilities in mastering certain activities. Verbal persuasion leads children to mobilize and sustain greater effort by setting a climate for achievement-oriented behavior. Empirical evidence has consistently shown that parents’ beliefs about schooling; general parental child-rearing beliefs, values, and goals; and culturally based beliefs are linked to children’s school achievement (Wigfield et al. 2006). Overall, it is plausible to think that parental meritocratic beliefs may influence children’s educational outcomes.

In addition, meritocratic beliefs are causal attributions without a specific subject. They refer to how society allocates rewards. Thus, the cultural orientation provided by the perceived meritocracy of the educational system is related to Weiner’s (1989) classic attributional model of achievement motivation, one of the most influential theories of achievement-related behavior (Glasgow et al. 1997). In this model, causal attributions generate affective reactions and expectations for future educational success, guiding academic behaviors and activities. Indeed, causal beliefs about effort and ability are held to produce the most optimal responses. Both beliefs are internal to students, and in the case of effort, they are volitional and can change over time. Internal, controllable, and changeable causes have been found to affect educational performance (Van Overwalle and De Metsenaere 1990; Perry et al. 2010). Overall, theories in social psychology suggest that parental meritocratic beliefs may enhance children’s educational performance either directly or indirectly through effort.

Meritocratic Beliefs as Justification

Researchers in sociology often turn the motivational hypothesis upside down. They argue that individual status positively predicts the perception of society as meritocratic because individuals seek to legitimize their superiority through a narrative of success (Bucca 2016; Khan 2011). Moreover, high-status individuals reject beliefs that challenge their social position (Kreidl 2000). Thus, instead of endorsing structural explanations of social stratification, they endorse individual explanations, such as skills and abilities. Since parents with high-performing children may be understood as winners in educational systems, the theories of social status literature may be applied to educational performance.

Sociologists debate the role of meritocratic beliefs in justifying educational outcomes, which has strived to generalize the role of meritocratic beliefs as legitimizing inequalities. Based on the theory on the heterogeneity and homogeneity of social structure, Mijs (2016a) showed that ability tracking shapes attributions of failure. Students in more homogeneous groups (mixed-ability classes) are more likely to attribute their failure to meritocratic factors, such as (lack of) hard work and academic (in)ability. Thus, stratification provides a context for cognitive processes that legitimize educational inequality. Students interpret their scholastic failure based on their position in the educational stratification. The same role has been suggested for meritocratic beliefs legitimizing inequality at the country level (Bucca 2016; Mijs 2019).

A large body of experimental evidence in social sciences has supported the self-serving bias as an explanation of the effect of performance on normative beliefs (Greenberg, Pyszczynski, and Solomon 1982; Rodriguez-Lara and Moreno-Garrido 2012; Ubeda 2014). In this kind of bias, winners overstate the role of talent, while losers attribute the outcome to external factors rather than meritocratic causes. In a recent experimental study, Molina et al. (2019) showed the effect of unequal outcomes on cognitive, normative, and affective responses. Their findings indicated that winners are more likely to perceive outcomes as fair relative to losers of a seven-round card game. Thus, we might expect that students with better performance will be more likely to attribute their educational outcomes to their merit. Molina and colleagues also manipulate the level of opportunities by allowing the exchange of a certain number of cards. When more cards are allowed, the scenario is more redistributive. In all the conditions of a low level of redistribution, winners were more likely than losers to see the outcome as fair and attributable to talent. In more egalitarian conditions, differences are reduced but not eliminated. These conditions are highly relevant in the Chinese context since the level of inequality has been increasingly high in the last several years (Xie and Zhou 2014). In the educational system, inequality has increased according to the positional approach (Hannum et al. 2019). Thus, the differences in normative beliefs of success between winners and losers in the educational system might sharpen in the China case.

Overall, the literature suggests that meritocratic narratives are justifiers of success and social inequalities, leading to the suggestion that parental meritocratic beliefs could justify students’ academic performance. Parents of students with higher grades, as winners of the educational tournaments, will be more likely to attribute their success to talent than parents of students with lower grades.

Integration of Reciprocal Effects and Hypotheses

The culture-as-rationalization approach criticizes the traditional understanding of culture as guiding action by ends and values (Swidler 1986). Instead, this approach conceives culture as providing a repertoire of justifications that limits and makes sense of the available possible strategies of action (Swidler 1986, 2001). This fashionable reversal of the sociological tradition was sustained by the observation that people seldom give consistent explanations of their behavior (Swidler 2001). Therefore, culture may not be a motivator of action.

However, advances in cognitive science have been echoed in sociology (Lizardo et al. 2016; Martin and Desmond 2010; Vaisey 2009) to vindicate the motivational role of culture (Bourdieu 1979; Ortner 2006). The dual-process model of culture proposes that individuals are both influenced by cultural schemas and capable of rationalizing when required by social interaction (Vaisey 2009; Vaisey and Lizardo 2010). This model enables investigation into how parental meritocratic narratives shape children’s educational performance in the form of vocabularies of motive (Vaisey and Lizardo 2010) as well as how parental meritocratic beliefs could be ad hoc rationalizations of achievement. Parental beliefs about meritocracy are semantic knowledge as impersonal propositions about the world. They are part of the discursive consciousness of parents (Lizardo 2017). In a dual-process model, these rule-like structures are resources that affect children’s’ outcomes either being stored in their nondeclarative semantic memory or directly through behavior (Martin and Desmond 2010). At the same time, they could rationalize and make sense of children’s academic achievement through meritocratic beliefs. Thus, reciprocal effects of performance and meritocratic beliefs are theoretically plausible.²

Hypothesis 1a: Students whose parents hold beliefs about skills as a predictor of academic achievement will have better educational performance.

Hypothesis 1b: Students whose parents hold beliefs about hard work as a predictor of academic achievement will have better educational performance.

Hypothesis 2a: Parents of students with higher academic performance are more likely to hold beliefs about skills as a predictor of academic achievement.

Hypothesis 2b: Parents of students with higher academic performance are more likely to hold beliefs about hard work as a predictor of academic achievement.

The relevance of this theoretical reasoning and the analytical strategy proposed in the following section may be illustrated with the recent evidence presented by Mijs (2019). His study aims to explain the paradox where citizens of the most unequal countries believe that their society is the paragon of meritocracy. His findings show that inequality is legitimated by meritocratic narratives. However, “it could be that the relationship between beliefs and inequality goes in the other direction” (Mijs 2019:16), or the two factors could be part of a dual process.

Insofar as declarative meritocratic beliefs may become internalized dispositions, they might explain positions in the social structure, which is made manifest in different levels of inequality. The same problem is encountered when examining the relationship between academic success and parental meritocratic narratives, which is a criticism of cultural explanations of ethnoracial gaps (Lizardo 2017). Parental beliefs may be encultured by children to affect their actions. Therefore, if parental meritocratic beliefs are coupled with children’s dispositions that foster scholastic achievement, I expect a stronger effect of meritocratic beliefs when parent–child relationships provide the conditions for such acquisition of culture. Similarly, children’s educational performance may send stronger signals when parents and children have a close relationship. Therefore, I hypothesize the following difference between groups:

Hypothesis 3: Reciprocal effects of meritocratic beliefs and educational performance are stronger for children with a close relationship with their parents.

Moreover, due to rapid economic growth and urbanization in China, one of the most important cultural cleavages is the rural–urban divide. Unlike global Chinese cities, such as Beijing and Shanghai, rural areas preserve traditional values (Chen and Chiu 2010; Wu 2017). In rural areas, filial piety, loyalty and self-sacrifice, and a general Confucian consciousness (Yang et al. 2006) still explain individual beliefs, attitudes, and behaviors. In addition, besides promoting thrift and respect for authority, Confucian values also advocate for educational achievement and hard work (Lim 2003). In addition, this faith in effort is coupled with the belief that educational success is not only seen as an individual achievement but is also associated with family honor (Peng 2019). Thus, in rural schools, parental meritocratic beliefs may have stronger efficacy and are salient cultural repertoires for making sense of achievement.³ Consequently, this literature leads to the following hypothesis:

Hypothesis 4a: Reciprocal effects of meritocratic beliefs and educational performance are stronger in rural areas than in urban areas.

However, modernization theories state that ascribed characteristics weaken their effects on status attainment, increasing the effectiveness of individual achievement in this process (Treiman 1970). In this context, the modernization of China may also cultivate norms, values, and beliefs consistent with this meritocratic understanding of development. Past studies have shown that the level of meritocracy in certain countries predicts support of meritocratic principles as well as a more advantaged individual social position (Bucca 2016; Kunovich and Slomczynski 2007). Individuals realize and internalize the idea that meritocracy explains rewards in rationalized and modern societies. Therefore, modernization reinforces meritocratic beliefs available as an explanation of success in urban areas and a motivator of actions that promote better performance. A competing hypothesis is thus derived:

Hypothesis 4b: Reciprocal effects of meritocratic beliefs and educational performance are stronger in urban areas than in rural areas.

Analytic Strategy

From the theoretical arguments proposed in the previous sections, I derive reciprocal effects between children’s educational performance and parental perception of meritocracy in the educational system. Autoregressive cross-lagged panel models (ACLPM) are a specific type of structural equation model that can be used if two or more variables have been measured longitudinally and the interest is in their reciprocal effects. ACLPM allows both latent and observed variables to be included. Moreover, the standardization of coefficients enables effects to be compared between paths.

As Huijsmans and colleagues (2019) explain, two kinds of effects can be estimated using ACLPM: first, autoregressive effects examining the average stability of a variable in students from one point in time to the next. For instance, the autoregressive path from beliefs about skills at Time 1 to beliefs about skills at Time 2 shows how stable this belief is between these two waves of the panel. These effects are represented by the solid lines in Figure 1.

Figure 1.

Theoretical Model

Second, cross-lagged effects indicate the effect of one variable on another measured at the following time. These effects represent the theoretical hypotheses of motivational and justificatory effects of meritocratic beliefs. They are represented in Figure 1 by dotted lines and dashed lines, respectively. For example, the effect of educational performance in Wave 1 on parental beliefs about skills in Wave 2 refers to the justification path for that belief. Due to the inclusion of the autoregressive effects, this path is independent of the belief at Time 1. This makes it possible to rule out that the effect of academic achievement on beliefs exists simply because these variables are correlated at the first measurement point. Controlling cross-lagged effects by autoregressive paths enables the direction of the potential causality to be tested. In addition, dash-dotted lines represent the effects between parental beliefs about skills and parental beliefs about hard work, which are intended to test the consistency of these beliefs in our context. Last, covariances are included for all the variables in the same period. It reflects the assumption that they share at least one omitted cause.

The control variables at Time 1 are included in two sets. One set of endogenous categorical or continuous variables and another set of exogenous binary variables. Following a standard procedure in ACLPM (Allison 2002; Huijsmans et al. 2019), the control variables are separated in these two sets for conducting full-information maximum-likelihood estimation rather than a listwise deletion of cases. This method provides better estimates of the parameters in comparison to listwise deletion, where missing data are assumed to be random. Dummy variables (hukou, migrant status, male, urban, close relationship) are specified as exogenous variables, which means that cases with missing values on these measurements are excluded. This procedure assumes that all omitted causes of endogenous variables are unrelated to the exogenous variables. In addition, educational performance is assumed to be a reflective measurement. That is, following Kline (2012), test scores reflect the influence of an underlying performance construct. I use a weighted least square parameter estimator for binary outcomes.⁴

Two additional multigroup analyses were conducted to test the equality of the parameters across groups. A multigroup model uses the same principle as interactions but applied across the structural equation model. All the analyses are conducted using Mplus 8.4.

Method

I analyze data from the first and second waves of the China Educational Panel Survey. The sample comprises 10,279 seventh graders surveyed in 2013–14 with a follow-up in eighth grade. This survey is a longitudinal, large-scale, probabilistic and representative sample of seventh and ninth graders from 438 classes at 112 schools in 28 counties in mainland China. Of the 10,279 seventh graders in the baseline, 9,449 students were surveyed in the follow-up in eighth grade. Total attrition in the follow-up of seventh graders is about 8.1 percent of the original sample. The analyses in this work consider only parent questionnaires answered by either the biological father or biological mother. Therefore, the final analytic sample includes 9,163 observations.

Key characteristics of the Chinese educational system are relevant to understand the role of meritocratic beliefs. The pre-tertiary education system can be divided into compulsory and postcompulsory education. First, six years of primary education and three years of junior high school are part of compulsory education. Second, there are three years of senior high school. Compulsory education is free, which aims to reduce heterogeneity within compulsory educational resources. The baseline sample students are in the first year of junior high school, with an average age of 13. Students in China are assigned to middle schools based on their area of residence. Elites schools are open to all the students within district boundaries, with lottery systems used by oversubscribed elite schools (Zhang 2013). Students sit their first citywide standardized exam during junior high school to compete for admission opportunities to senior high schools at the end of ninth grade. Therefore, both students and parents experience an increase in competitive intensity. Grades during junior high school are based on tests that seek to replicate the standardized examination. Thus, a good academic performance in junior high school and, therefore, a higher test score could increase the chances of entering into a high-quality senior school (i.e., key-point school), an important predictor of college enrollment (Ye 2015). Compared to Western contexts, where admission to higher education is multidimensional, the test score in the standardized entry examination (gaokao) is the most salient factor. Besides test-centeredness being a central characteristic of the Chinese educational system (Hannum et al. 2019), grades are important for within-school ability tracking, signals of future success in standardized examinations, and Chinese culture places a high social value on them (Li et al. 2018).

Main Variables

ACLPM uses variables that are both predictors and outcomes at different time points. First, parents were asked, among a list of factors, whether “the extent of hardworking” and “talent and capability” have effects on students’ grades, where 1 = yes and 0 = no. These are parents’ beliefs about hard work and skills, respectively.

Second, educational performance is measured as a latent variable based on the grade ranking in math, Chinese, and English midterm exams for each year.⁵ I compute students’ ordinal rank—the higher the rank, the higher the grades—as their decile in each exam session evaluation across classmates in the same grade. Since the absolute rank measure does not account for different sizes of students’ groups, the decile rank (Ω) is standardized to size:

Ω_{ist} = \frac{α_{ist} - 1}{N_{ist}} \times 10,

(1)

where α is the absolute rank of student i within school s for wave t. Wave 1 contained two grades, while Wave 2 only one. Therefore, each subject grade rank varies in a range from 0 to 10, where 0 refers to the student at the bottom of the distribution and 10 to the students at the top. Thus, each unit of increase or decrease represents a change of 10 percentiles. After oblique rotation, factor loadings are above 1, indicating a positive and strong relationship between the latent variable performance and the three observed measures in each wave. An advantage of using grade rank instead of class rank is that it prevents biases due to tracking effect. This potential bias is relevant, considering the supported effect of tracking on causal attributions of failure and success (Mijs 2016a).

Control Variables

As mentioned, several control variables were included from the survey baseline as exogenous and endogenous variables. First, three binary indicators were included as exogenous in the models. A dummy variable indicating whether the student was male (1 = yes, 0 = no) was included because there is evidence suggesting gender differences in the effect of parental beliefs about success on children skills (Mägi et al. 2011). In addition, binary indicators were incorporated, denoting whether the student holds agricultural hukou (1 = yes, 0 = no), the geographic area of the schools (1 = urban, 0 = rural), and migration status (1 = yes, 0 = no). These variables have been suggested as important sources of educational disadvantage (Hannum et al. 2019). Finally, I used a dichotomous indicator of the parent–child relationship where 1 denotes children declaring a “very close” relationship with the parent who responded to the questionnaire. Thus, 0 corresponded to children with a “not close” or “not too close nor too far” relationship.

An additional block of five endogenous control variables was included in the model, of which two were predictors intended to control for objective conditions of skills and hard work beliefs: cognitive ability and average hours spent on homework. First, because of its logic effect on academic performance, cognitive ability was measured by cognitive ability test score estimated using an item response theory model. Second, regarding homework, students were asked how much time they had spent on homework assigned by their teachers at school the previous week. The hours spent from Monday to Friday were added to the hours spent on weekends. I did not consider minutes because of the unreliability of this measure. Due to its skewness, a logarithmic transformation of this variable was used. The third endogenous variable, in order to take account of the children’s development, was student age. The fourth variable was the educational level of the father and mother, measured on a range from 1 to 9, where 1 = no education and 9 = master’s degree or higher. It has been reported that the educational level of both parents affects the path from parental beliefs to children’s outcomes (Sigel and McGillicuddy-De Lisi 2002). The fifth variable was the number of siblings (including the respondents), intended to address the effect of sibship size on educational outcomes (Lu and Treiman 2008). In addition, the control by parents’ educational level, number of siblings, and hukou enables us to show that it is the parents’ beliefs that influence student performance, not other resources they can offer.

Results

Descriptive Results

Before reporting the testing of the hypotheses, Table 1 shows the descriptive statistics involved in the analyses. As expected, the mean value of grade rank at every midterm for Waves 1 and 2 is around the 50th percentile. Regarding meritocratic beliefs, there is an important descriptive result to be highlighted. For each wave, the proportion of parents who consider that skills lead to better performance is smaller than those who perceive the payoffs of hard work. Three of every 10 parents believe that talent and capabilities are predictors of better academic achievement. In contrast, 8 of every 10 believe that students who make a greater effort will obtain higher grades. This suggests that beliefs about skills and hard work do not go hand in hand in the Chinese context. This is confirmed by low unweighted tetrachoric correlations when students were seventh grade (r = .16) and eighth grade (r = .23). Table 1 also shows the proportion of parents who changed their beliefs between Wave 1 and Wave 2, with 36 percent changing their beliefs about skills, while 22 percent changed their endorsement of hard work as a predictor of success. Both are substantive proportions of change to be explained in this study. The between-wave variation of grade ranks is even more considerable: one out of each four students changed their performance in at least one decile between waves in math, Chinese, or English. It is worth noting the high level of consensus about the perception of hard-work payoffs, which indicates this narrative seems to be a strong belief in Chinese public culture. Despite the lower level of variance of this variable, the large sample size enables the theoretical hypotheses to be accurately tested.

Table 1.

Descriptive Statistics of Main Variables

Variable	M	SD	N	Min.	Max.
Main predictors
Grade rank in math_t1	4.79	2.80	9,163	0	10
Grade rank in math_t2	4.96	2.86	9,163	0	10
Grade rank in Chinese_t1	4.79	2.80	9,163	0	10
Grade rank in Chinese_t2	4.95	2.86	9,163	0	10
Grade rank in English_t1	4.79	2.80	9,163	0	10
Grade rank in English_t2	4.95	2.86	9,163	0	10
Belief about hard work_t1	0.82	0.38	9,064	0	1
Belief about hard work_t2	0.85	0.35	8,284	0	1
Belief about skills_t1	0.31	0.46	9,064	0	1
Belief about skills_t2	0.37	0.48	8,284	0	1
Between-wave variation
Grade rank in math^a	0.77	0.42	9,163	0	1
Grade rank in Chinese^a	0.77	0.42	9,163	0	1
Grade rank in English^a	0.74	0.44	9,163	0	1
Belief about hard work	0.22	0.41	8,197	0	1
Belief about skills	0.36	0.48	8,197	0	1
Covariates
Age	0.79	12.69	8,963	11	17
Male	0.50	0.53	9,163	0	1
Agricultural hukou	0.49	0.62	9,163	0	1
Urban school	0.50	0.52	9,163	0	1
Migrant status	0.32	0.12	9,098	0	1
Mother’s educational level	1.84	3.51	9,154	1	9
Father’s educational level	1.86	3.88	9,154	1	9
Number of siblings	0.88	1.89	9,163	1	6
Cognitive skills test	0.87	–0.11	9,163	–2.03	2.33
Time spent on homework (logarithm)	0.75	6.76	8,294	2.30	9.22
Close parent–child relationship	0.43	0.76	9,130	0	1

Note: Weighted statistics. Following Hsieh (1989), sample size and distributions of binary variables provide sufficient statistical power for our estimations. t1 = Time 1; t2 = Time 2.

Variation in at least one decile.

Multivariate Results

Table 2 shows the results of the cross-lagged model testing the hypotheses of the dual process of parental meritocratic beliefs. In addition, this model includes stability paths and the relationships between meritocratic beliefs (consistency path). Overall, the goodness-of-fit indices indicate that our model fits the observations accurately (root mean square error of approximation [RMSEA] = .031, comparative fit index [CFI] = .964, Tucker-Lewis index [TLI] = .926). Standardized and unstandardized coefficients are reported. Effects on binary outcomes are reported in log odds. The standardization of variables predicting binary outcomes also addresses the limitations of logistic regressions underlined by Mood (2010). Sociologists usually ignore the fact that logistic regressions are biased by omitted variables that are unrelated to independent variables when using nonexperimental data (i.e., performance and beliefs are not random treatments), unlike linear regressions. Therefore, coefficients across models in the same sample are incomparable. Doing so requires a y-standardization (Winship and Mare 1984), which divides the coefficients by the standard deviation of the latent variable for each equation of the structural model (sdY*). These standardized coefficients are reported in Table 2 for binary outcomes. First, stability paths show divergent results for educational performance and meritocratic beliefs. Every variable in Wave 2 is significant and positively related to its value in Wave 1. However, while educational performance is more stable (β = .782, p < .001), autoregressive effects of beliefs about skills (β = .347, p < .001) and hard work (β = .342, p < .001) are smaller. Once again, this suggests a substantive degree of variability between waves.

Table 2.

Cross-Lagged Panel Results for Dual-Process Model of Meritocratic Beliefs

	Outcome variables at t₂
	BHW		BS		EP
Predictor at t₁	B (SE)	β	B (SE)	β	B (SE)	β
BHW	.342*** (.036)	.329	.143*** (.030)	.141	.065 (.039)	.030
BS	.093*** (.033)	.085	.347*** (.026)	.328	.055 (.035)	.024
EP	.116*** (.013)	.221	.030** (.011)	.059	.782*** (.024)	.703

Note: Control variables are not reported. Dummy controls are assumed as exogenous to reduce missing data. Standardized coefficients reported in log odds. Model fit: root mean square error of approximation = .031, comparative fit index = .964, Tucker-Lewis index = .926. Residual covariances between variables within waves and baseline controls are included. Model has been weighted. t1 = Time 1; t2 = Time 2; BHW = parents’ belief about hard work; BS = parents’ belief about skills; EP = educational performance.

p < .05. **p < .01. ***p < .001.

Table 2 also shows the relationship between meritocratic beliefs. If the classic conceptualization of merit as a function of skills and effort is accurate, a high association between meritocratic narratives is expected. Although they have a positive and significant relationship, the effect size is smaller in comparison to other paths. Parents who believe that hard work leads to better educational performance are 1.15 (odds ratio [OR] = exp[.143]) times more likely to hold beliefs about skills. In a similar fashion, parents who believe in skills as a predictor of academic success are 1.10 (OR = exp[.093]) times more likely to be believers in hard work. As well as descriptive statistics, this indicates that beliefs about skills and hard work as predictors of students’ success work differently in the Chinese context.

Finally, Table 2 reports the results for the testing of the hypotheses of the dual process of meritocratic beliefs. The last column indicates that neither parental beliefs about skills nor those about hard work are significantly related to educational performance in Wave 2. Parental perception of meritocracy does not influence children’s educational performance. Hence, these results do not support the hypotheses on motivation (Hypotheses 1a and 1b). Conversely, educational performance in Wave 1 is a significant predictor of narratives about both skills (β = .059, p < .01) and hard work (β = .221, p < .001). These coefficients could be expressed in terms of ORs as a magnitude of their effect size. They indicate that an increase of one standard deviation in the class rank of educational performance leads to a 1.03-times increase in the likelihood of holding skills beliefs, while for beliefs about hard work, this increase is 1.12 times. The difference in magnitude between these two coefficients is a straightforward indicator of a substantive effect on hard work beliefs and a less strong effect on attributional beliefs about skills. Although the absolute effect could be considered small following standard thresholds, it is also possible that these effects accumulate over an educational career. Thus, if students increase or decrease their performance consistently, parental beliefs could be affected to a large extent.

In addition, differences between motivational and justificatory paths are formally tested and reported in Tables 3 and 4. Table 3 indicates that the justification role of attributional beliefs (educational performance → skills belief; educational performance → hard work) is not significantly different (p > .05) from its motivational path (skills belief → educational performance; hard work belief → educational performance), as suggested by the Wald chi-square tests. Table 4 shows that the effect of educational performance on narratives about skills and hard work are heterogeneous. In the Chinese context, the justificatory effect of educational performance on beliefs about hard work is stronger than the effect on beliefs about skills (p < .001). Therefore, as suggested earlier, skills and hard work narratives do not go hand in hand in this dual process of parental meritocratic beliefs. This finding is a crucial contribution to the literature on attributional beliefs that emphasize Chinese parents’ effort as a predictor of academic achievement. Here, we show not only a descriptive difference between beliefs but also a difference in its functioning.

Table 3.

Wald Chi-Square Tests for Reciprocal Effects Differences

Motivation paths		Justification paths
Predictor	Outcome	Predictor	Outcome	χ² Wald test (1)	p
BHW	EP	EP	BHW	1.561	.212
BS	EP	EP	BS	0.476	.490

Note: BHW = parents’ belief about hard work; BS = parents’ belief about skills; EP = educational performance.

Table 4.

Wald Chi-Square Tests for Meritocratic Beliefs Differences

Justification path of skills		Justification path of hard work
Predictor	Outcome	Predictor	Outcome	χ² Wald test (1)	p
EP	BS	EP	BHW	28.945	.000

Note: BHW = parents’ belief about hard work; BS = parents’ belief about skills; EP = educational performance.

Another form of testing whether beliefs do not have an effect on educational performance is constraining the motivational paths to zero (skills belief → educational performance; hard work belief → educational performance). The goodness-of-fit indices for this model improved slightly in comparison with the previous model (RMSEA = .031, CFI = .964, TLI = .928). This is another argument for disregarding the motivational effects of parental beliefs on performance. In addition, different specifications of the model were estimated. It is possible that cognitive ability and time spent on homework could mediate the effect of beliefs on performance. I have tested this possibility, and the results are not significant and near zero. Besides, Table 1A in the online supplement shows the model without control variables. Despite the changes in the coefficients, the conclusion is the same: the effect of beliefs on performance is nonsignificant and close to zero. In contrast, and consistently with the main models, performance has a positive effect. In the online supplement, Table 2A reports the model without hours of homework and cognitive ability. It answers the question of whether the time spent on homework controls away the effect. The results are consistent with the original findings. Thus, the evidence does not support the direct effect of beliefs or the indirect effect through effort in these specifications.

Overall, the findings indicate that parents’ meritocratic beliefs have no significant effect on children’s educational performance. However, educational performance has significant effects on both beliefs about skills and hard work as predictors of children’s achievement. Therefore, parents’ meritocratic narratives justify educational performance.

Multigroup Analyses

Multigroup analyses are reported in Table 5. First, the justificatory effect of performance on beliefs about hard work is significant for parents with close (B = .238, p < .001) and nonclose relationships (B = .162, p < .001). For skills, it is significant only in the case of close relationships (B = .069, p < .01). Although the magnitude of the effects is larger for close parent–child relations, the differences are not significant according to the Wald tests. However, what is important in this case is that the quality of the parent–child relationship does not influence the potential motivational effect of meritocratic beliefs. Thus, the evidence does not support Hypothesis 3. The null hypothesis is highly relevant because, as a propitious condition for the internalization of parental declarative culture, a close parent–child relationship does not increase the motivational effect of parental meritocratic beliefs on children’s educational performance.

Table 5.

Multigroup Analyses of Motivational and Justificatory Paths

		Group
		Close parent–child relationship		Geographic area of school
		Yes	No	Rural	Urban
Dual process	Path	β	β	β	β
Motivational effects	BHW → EP	.024	.058	.001	.051
Motivational effects		(.021)	(.036)	(.026)	(.025)
	BS → EP	.033	-.016	.018	.022
		(.017)	(.036)	(.023)	(.020)
Justificatory effects	EP → BHW	.238***	.162***	.291***;	.164***
Justificatory effects		(.027)	(.018)	(.034)	(.034)
	EP → BS	.069**	.025	.051	.053***
		(.024)	(.044)	(.032)	(.028)

Note: Model in Table 2 has been estimated for each group. BHW = parents’ belief about hard work; BS = parents’ belief about skills; EP = educational performance.

p < .05. **p < .01. ***p < .001.

Finally, the results regarding the competing hypotheses of Confucian tradition and modernization theories are also shown in Table 5. Parental meritocratic beliefs do not motivate action in either rural or urban schools. Moreover, in both cases, higher educational performance predicts the endorsement of beliefs about hard work. A Wald test on the difference between these two coefficients indicates that the effect is stronger in schools located in rural areas (p < .01, χ² = 6.796). In addition, the justificatory effect of the belief about hard work as a predictor of success is stronger than its motivational effect (p < .05, χ² = 5.876) in rural schools. This supports Hypothesis 4a on the Confucian tradition but specified only to its justificatory effect. The argument that urbanization makes meritocratic beliefs more salient due to modernization is not sustained by the results.

Sensitivity Analyses

Different robustness checks were conducted to assess the sensitivity of the findings. First, alternative measurements of academic performance were included in the model, replacing grade rank. Although Chinese parents show a high level of engagement with children’s education, there is still possibility of informational bias.⁶ Thus, we replaced the grade rank by parents’ declared rank of students. Parents were asked, “How does this child’s academic record rank in his/her class at present?” Their answers could range from 1 to 5, where 1 was “near the bottom” and 5 “around the top.” Once again, Table 3A in the online supplement confirms our results. In this case, belief about hard work increases the subjective assessment of children’s performance (B = .087, p < .000), in the same way as objective assessment. However, the Wald test indicates that the justificatory path of belief about hard work is stronger than its motivational path (p < .05, χ² = 5.868).

Second, class rank cannot be used because 15 percent of the schools have rearranged students in the second wave, according to school principals. Researchers could argue that class rank is a more direct indicator of educational performance than a more abstract grade rank. Thus, the analyses were reproduced using only the subsample of schools without rearrangement and class ranks. As shown in Table 4A in the online supplement, the main effects are confirmed and in a similar magnitude. Only the justificatory path of effort is significant, but it could be attributed to the reduction of the sample size. The difference between justificatory paths is also significant (p < .001), like results using grade rank.

Finally, the comparison of logistic regression is difficult not only between models, as mentioned before, but also across samples (Allison 1999; Williams 2009). The robustness of the multigroup analysis requires addressing the potential bias of unobserved heterogeneity across samples. To assume that all the variables predict the outcome equally well across groups (e.g., close/nonclose, urban/rural) is unrealistic and difficult to support theoretically. As Breen, Karlson, and Holm (2018) explain, the comparison of coefficients across groups from nonlinear probability models rests on the common residual variance assumption, which is untestable. I use linear probability models (LPM) for each multigroup analysis, as proposed by Mood (2010). Differences between groups’ coefficients using LPM may not correspond to differences using relative terms, such as odds ratios (Breen et al. 2018). With these models, the conclusions are not substantively altered.

Discussion

The aim of this study was to understand the reciprocal effects of the perception of meritocracy in the educational system and children’s educational performance in China. The findings indicate that parental narratives of skills and hard work as predictors of success do not have an effect on the educational performance of children. Instead, these meritocratic beliefs are affected by academic results, suggesting they have a justificatory role. This pattern is much clearer in rural China, where traditional Chinese culture is preserved. Beyond the statistical significance of these findings, their implications provide insights into their substantive significance. Recent studies in cultural sociology have suggested that personal culture does not usually change (Kiley and Vaisey 2020). By contrast, political and policy issues change within person due to contemporaneous external factors (Vaisey and Lizardo 2016). However, these results show that changes in children’s performance could modify parental worldviews, suggesting potential intrafamily sources of cultural change. Thus, the results open an avenue to further understand whether these changes persist in the long run.

In addition, this study sought to explore the consistency of skills and hard-work beliefs. The results indicate that in China they are moderately related, and beliefs about hard work are used to make sense of children’s better educational performance in a more salient manner. This finding is surprising due to its lower level of variance in comparison to beliefs about skills. It is consistent with the idea that Confucian tradition promotes willpower as a mean of success (Lee 1999) and that abilities may be rewarded if effort is made. For Confucius, any man might become a “gentleman,” and in education, there should be no class distinctions (Creel 1953). The prominence of beliefs about hard work in China is consistent with the so-called meritocratic component of meritocracy in Western societies (Bucca 2016; Mijs 2019). Ample literature in social psychology indicates that Chinese parents believe in hard work as leading to success to a larger extent relative to natural ability (Chen and Stevenson 1995; Kinlaw et al. 2001; Stevenson et al. 1986). Here, I have shown that these beliefs are also different in how they work with educational performance.

Another interpretation of the difference between beliefs about skills and hard work is the difference established by the attributional theory (Weiner 1989), used to predict the effect of beliefs on children’s achievement. Following the attributional argument, hard work should have a stronger effect on achievement than skills, because in addition to being internal, it is more controllable and changeable than skills. Therefore, it has a greater effect on emotions and behaviors. In the case of this study, the justificatory effect of hard work is stronger than skills. Therefore, using hard work as a narrative of success enhances the deservedness of the individual even more because it is more volitional than skills.

The findings of the present study have important implications for the growing body of research on citizens’ inequality beliefs worldwide (Castillo 2011; Hunt 2007; McCall 2013; Mijs 2019) and with a focus on China (Chen, Tam, and Chiang 2019; Lei 2020; Xian and Reynolds 2017; Xie 2016). This study has shown that the relationship between academic success and meritocratic beliefs is not bidirectional as theorized, but rather, these factors justify such a relationship. I have provided a theoretical model and a longitudinal test of the legitimation hypothesis suggested by this tradition and the motivational hypothesis, which is used to explain ethnoracial gaps in sociology (Hsin and Xie 2014; Liu and Xie 2016; Ogbu 2003). In a dual-process model of parental beliefs, only culture as rationalization (DiMaggio 1997; Swidler 1986, 2001; Vaisey 2009) is supported. This evidence joins skeptical views of Confucian values as affecting higher educational performance in the United States (Lee and Zhou 2015). Future studies may extend this model, considering the reciprocal effects in different societies and systems. The study proposes a potential causal effect of educational performance on parental beliefs that might be further tested using experimental or quasiexperimental methods.

The lack of effect of parental beliefs on children’s educational performance should not be interpreted as the impossibility of motivating students. Instead, these findings could suggest a decoupling between parents’ declarative culture and children’s dispositions that could enhance educational performance. As Lizardo (2017) argues, a superficial understanding of these cultural processes has led motivational studies toward a declarative culture bias. Further studies might address whether parental beliefs fail to motivate educational performance or whether there is a decoupling between declarative and nondeclarative forms of culture. In addition, this study has focused on the perceived meritocracy of the educational system. As suggested by previous studies, it is important to distinguish between preferences about inequality and perceived inequality (Castillo 2011). More research is needed to reveal any potential motivational effects of these alternative cultural orientations. Besides, the China Educational Panel Survey does not include measurements of children’s beliefs or specific providence of residence. In general, there is a lack of data measuring shared patterns of thinking or talking of children in China. Studies in cultural sociology and sociology of education might move forward the understanding of culture’s role by expanding the data availability.

This study has also demonstrated the role of parental meritocratic narratives in an East Asian country. To the best of my knowledge, this is the first study simultaneously considering meritocratic beliefs and educational performance in this region. Meritocratic selection of civil servants is widespread across countries such as Korea, Japan, and China, where it provides the building blocks of an exam-oriented educational culture (Hannum et al. 2019). In these countries, Confucian philosophy has led to long-lasting competitive pressures on families, schools, teachers, and students. However, the descriptive relevance of meritocratic beliefs in the belief system says nothing about their relations with action or structure. Meritocracy is a dominant normative principle, and parental meritocratic narratives justify success in the Chinese exam-oriented system. As once stated by Weber (1978:953), “every highly privileged group develops the myth of its natural superiority,” and in the Chinese educational system, the myth seems to be that of skills and hard work.

Supplemental Material

sj-docx-1-spq-10.1177_0190272520984730 – Supplemental material for Motivation, Legitimation, or Both? Reciprocal Effects of Parental Meritocratic Beliefs and Children’s Educational Performance in China

Supplemental material, sj-docx-1-spq-10.1177_0190272520984730 for Motivation, Legitimation, or Both? Reciprocal Effects of Parental Meritocratic Beliefs and Children’s Educational Performance in China by Francisco Olivos in Social Psychology Quarterly

Footnotes

Acknowledgements

I thank professor Tony Tam and attendees at the postgraduate seminar at the Chinese University of Hong Kong for their insightful comments. I am grateful to Juan Carlos Castillo, Twan Huijsmans, Jonathan Mijs, and Jacqueline Chen Chen for their suggestions on previous drafts of this article. Preliminary versions of this paper were presented at the American Sociological Association 2020 virtual conference, the Hong Kong Sociological Association 21st annual conference, and the NZASIA 23rd biennial international conference at Victoria University of Wellington.

Funding

The author thanks the support of the Hong Kong PhD Fellowship Scheme of the Research Grants Council, Hong Kong SAR.

ORCID iD

Francisco Olivos

Supplemental Material

Supplemental material is available with the online version of the article:

Notes

Bio

Francisco Olivos is a PhD candidate in sociology at the Chinese University of Hong Kong. He is interested in quantitative cultural sociology, subjective well-being, and computational methods. His work has been previously published in Poetics and the Journal of Happiness Studies, among others.

References

Allison

Paul

. 1999. “Comparing Logit and Probit Coefficients across Groups.” Sociological Methods & Research 28(2):186–208.

Allison

Paul David

. 2002. Missing Data. Thousand Oaks, CA: Sage.

Bandura

Albert

. 1997. Self-Efficacy in Changing Societies. Cambridge, UK: Cambridge University Press.

Bandura

Albert

Barbaranelli

Claudio

Caprara

Gian Vittorio

Pastorelli

Concetta

. 2001. “Self-Efficacy Beliefs as Shapers of Children’s Aspirations and Career Trajectories.” Child Development 72(1):187–206.

Bourdieu

Pierre

. 1979. Distinction: A Social Critique of the Judgement of Taste. Cambridge, MA: Harvard University Press.

Breen

Richard

Karlson

Kristian Bernt

Holm

Anders

. 2018. “Interpreting and Understanding Logits, Probits, and Other Nonlinear Probability Models.” Annual Review of Sociology 44(1):39–54.

Bucca

Mauricio

. 2016. “Merit and Blame in Unequal Societies: Explaining Latin Americans’ Beliefs about Wealth and Poverty.” Research in Social Stratification and Mobility 44:98–112.

Castillo

Juan Carlos

. 2011. “Legitimacy of Inequality in a Highly Unequal Context: Evidence from the Chilean Case.” Social Justice Research 24(4):314–40.

Chen

Chuansheng

Stevenson

Harold W.

1995. “Motivation and Mathematics Achievement: A Comparative Study of Asian-American, Caucasian-American, and East Asian High School Students.” Child Development 66(4):1215–34.

10.

Chen

Jacqueline Chen

Tam

Tony

Chiang

Yen Sheng

. 2019. “The Rise of Merit-Based Inequality Acceptance after Exposure to Competition: Experimental Evidence among Chinese University Students.” Social Indicators Research 144(2):707–28.

11.

Chen

Xia

Chi

Chiu Yue

. 2010. “Rural-Urban Differences in Generation of Chinese and Western Exemplary Persons: The Case of China.” Asian Journal of Social Psychology 13(1):9–18.

12.

Creel

Herrlee Glessner

. 1953. Chinese Thought, from Confucius to Mao Tsê-Tung. Chicago: University of Chicago Press.

13.

Deci

Edward L.

Koestner

Richard

Ryan

Richard M.

2001. “Extrinsic Rewards and Intrinsic Motivation in Education: Reconsidered Once Again.” Review of Educational Research 71(1):1–27.

14.

DiMaggio

Paul

. 1997. “Culture and Cognition.” Annual Review of Sociology 23(1):263–87.

15.

Frye

Margaret

. 2012. “Bright Futures in Malawi’s New Dawn: Educational Aspirations as Assertions of Identity.” American Journal of Sociology 117(6):1565–1624.

16.

Glasgow

Kristan L.

Dornbusch

Sanford M.

Troyer

Lisa

Steinberg

Laurence

Ritter

Philip L.

1997. “Parenting Styles, Adolescents’ Attributions, and Educational Outcomes in Nine Heterogeneous High Schools.” Child Development 68(3):507–29.

17.

Greenberg

Jeff

Pyszczynski

Tom

Solomon

Sheldon

. 1982. “The Self-Serving Attributional Bias: Beyond Self-Presentation.” Journal of Experimental Social Psychology 18(1):56–67.

18.

Hannum

Emily

Ishida

Hiroshi

Park

Hyunjoon

Tam

Tony

. 2019. “Education in East Asian Societies: Postwar Expansion and the Evolution of Inequality.” Annual Review of Sociology 45(1):625–47.

19.

Harackiewicz

Judith M.

Manderlink

George

Sansone

Carol

. 1984. “Rewarding Pinball Wizardry: Effects of Evaluation and Cue Value on Intrinsic Interest.” Journal of Personality and Social Psychology 47(2):287–300.

20.

Heiserman

Nicholas

Simpson

Brent

. 2017. “Higher Inequality Increases the Gap in the Perceived Merit of the Rich and Poor.” Social Psychology Quarterly 80(3):243–53.

21.

Herrnstein

Richard J.

Murray

Charles A.

1996. The Bell Curve: Intelligence and Class Structure in American Life. New York: Simon & Schuster.

22.

Hsieh

F. Y.

1989. “Sample Size Tables for Logistic Regression.” Statistics in Medicine 8(7):795–802.

23.

Hsin

Amy

Xie

. 2014. “Explaining Asian Americans’ Academic Advantage over Whites.” Proceedings of the National Academy of Sciences of the United States of America 111(23):8416–21.

24.

Huijsmans

Twan

Nivette

Amy E.

Eisner

Manuel

Ribeaud

Denis

. 2019. “Social Influences, Peer Delinquency, and Low Self-Control: An Examination of Time-Varying and Reciprocal Effects on Delinquency over Adolescence.” European Journal of Criminology. https://doi.org/10.1177/1477370819838720.

25.

Hunt

Matthew O.

2007. “African American, Hispanic, and White Beliefs about Black/White Inequality, 1977-2004.”American Sociological Review 72(3):390–415.

26.

Khan

Shamus Rahman

. 2011. Privilege: The Making of an Adolescent Elite at St. Paul’s School. Princeton, NJ: Princeton University Press.

27.

Kiley

Kevin

Vaisey

Stephen

. 2020. “Measuring Stability and Change in Personal Culture Using Panel Data.” American Sociological Review 85(3):477–506.

28.

Kinlaw

C. Ryan

Kurtz-Costes

Beth

Goldman-Fraser

Jenifer

. 2001. “Mothers’ Achievement Beliefs and Behaviors and Their Children’s School Readiness: A Cultural Comparison.” Journal of Applied Developmental Psychology 22(5):493–506.

29.

Kline

Rex B.

2012. “Assumptions in Structural Equation Modeling.” Pp. 111–25 in Handbook of Structural Equation Modeling, edited by Hoyle

New York: Guilford Press.

30.

Kreidl

Martin

. 2000. “Perceptions of Poverty and Wealth in Western and Post-Communist Countries.” Social Justice Research 13(2):151–76.

31.

Kunovich

Sheri

Slomczynski

Kazimierz M.

2007. “Systems of Distribution and a Sense of Equity: A Multilevel Analysis of Meritocratic Attitudes in Post-Industrial Societies.” European Sociological Review 23(5):649–63.

32.

Lamont

Michèle

. 2019. “From ‘Having’ to ‘Being’: Self-Worth and the Current Crisis of American Society.” British Journal of Sociology 70(3):660–707.

33.

Lee

Jennifer

Zhou

Min

. 2015. The Asian American Achievement Paradox. New York: Russell Sage Foundation.

34.

Lee

Wing.

1999. “The Cultural Context for Chinese Learners: Conceptions of Learning in the Confucian Tradition.” Pp. 25–41 in The Chinese Learner: Cultural, Psychological and Contextual Influences, edited by Watkins

Biggs

Hong Kong and Melbourne: Comparative Education Research Centre and Australian Council for Educational Research.

35.

Lei

Ya-Wen

. 2020. “Revisiting China’s Social Volcano: Attitudes toward Inequality and Political Trust in China.” Socius: Sociological Research for a Dynamic World 6:1–21.

36.

Fan

Loyalka

Prashant

Hongmei

Shi

Yaojiang

Johnson

Natalie

Rozelle

Scott

. 2018. “Ability Tracking and Social Trust in China’s Rural Secondary School System.” School Effectiveness and School Improvement 29(4):545–72.

37.

Lim

Vivien

. 2003. “Money Matters: An Empirical Investigation of Money, Face and Confucian Work Ethic.” Personality and Individual Differences 35(4):953–70.

38.

Liu

Airan

Xie

. 2016. “Why Do Asian Americans Academically Outperform Whites? The Cultural Explanation Revisited.” Social Science Research 58:210–26.

39.

Lizardo

Omar

. 2017. “Improving Cultural Analysis: Considering Personal Culture in Its Declarative and Nondeclarative Modes.” American Sociological Review 82(1):88–115.

40.

Lizardo

Omar

Mowry

Robert

Sepulvado

Brandon

Stoltz

Dustin S.

Taylor

Marshall A.

Van Ness

Justin

Wood

Michael

. 2016. “What Are Dual Process Models? Implications for Cultural Analysis in Sociology.” Sociological Theory 34(4):287–310.

41.

Yao

Treiman

Donald J.

2008. “The Effect of Sibship Size on Educational Attainment in China: Period Variations.” American Sociological Review 73(5):813–34.

42.

Machamer

Peter

Darden

Lindley

Craver

Carl F.

2000. “Thinking about Mechanisms.” Philosophy of Science 67(1):1–25.

43.

Mägi

Katrin

Lerkkanen

Marja-Kristiina

Poikkeus

Anna-Maija

Rasku-Puttonen

Helena

Nurmi

Jari-Erik

. 2011. “The Cross-Lagged Relations between Children’s Academic Skill Development, Task-Avoidance, and Parental Beliefs about Success.” Learning and Instruction 21(5):664–75.

44.

Martin

John Levi

Desmond

Matthew

. 2010. “Political Position and Social Knowledge.” Sociological Forum 25(1):1–26.

45.

McCall

Leslie

. 2013. The Undeserving Rich: American Beliefs about Inequality, Opportunity, and Redistribution. New York: Cambridge University Press.

46.

Mijs

Jonathan

. 2016a. “Stratified Failure: Educational Stratification and Students’ Attributions of Their Mathematics Performance in 24 Countries.” Sociology of Education 89(2):137–53.

47.

Mijs

Jonathan

. 2016b. “The Unfulfillable Promise of Meritocracy: Three Lessons and Their Implications for Justice in Education.” Social Justice Research 29(1):14–34.

48.

Mijs

Jonathan

. 2019. “The Paradox of Inequality: Income Inequality and Belief in Meritocracy Go Hand in Hand.” Socio-Economic Review. https://doi.org/10.1093/ser/mwy051.

49.

Molina

Mario

Bucca

Mauricio

Macy

Michael

. 2019. “It’s Not Just How the Game Is Played, It’s Whether You Win or Lose.” Science Advances 5(7).

50.

Mood

2010. “Logistic Regression: Why We Cannot Do What We Think We Can Do, and What We Can Do about It.” European Sociological Review 26(1):67–82.

51.

Ogbu

John U.

2003. Black American Students in an Affluent Suburb: A Study of Academic Disengagement. New York: Routledge.

52.

Ortner

Sherry

. 2006. Anthropology and Social Theory: Culture, Power, and the Acting Subject. Durham, NC: Duke University Press Books.

53.

Peng

Yinni

. 2019. “Getting Rural Migrant Children into School in South China: Migrant Agency and Parenting.” Asian Population Studies 15(2):172–89.

54.

Perry

Raymond P.

Stupnisky

Robert H.

Hall

Nathan C.

Chipperfield

Judith G.

Weiner

Bernard

. 2010. “Bad Starts and Better Finishes: Attributional Retraining and Initial Performance in Competitive Achievement Settings.” Journal of Social and Clinical Psychology 29(6):668–700.

55.

Pulfrey

Caroline

Darnon

Céline

Butera

Fabrizio

. 2013. “Autonomy and Task Performance: Explaining the Impact of Grades on Intrinsic Motivation.” Journal of Educational Psychology 105(1):39–57.

56.

Rodriguez-Lara

Ismael

Moreno-Garrido

Luis

. 2012. “Self-Interest and Fairness: Self-Serving Choices of Justice Principles.” Experimental Economics 15(1):158–75.

57.

Saunders

Peter

. 1995. “Might Britain Be a Meritocracy?” Sociology 29(1):23–41.

58.

Schuman

Howard

Walsh

Edward

Olson

Camille

Etheridge

Barbara

. 1985. “Effort and Reward: The Assumption That College Grades Are Affected by Quantity of Study.” Social Forces 63(4):945–66.

59.

Sigel

Irving E.

Lisi

Ann McGillicuddy-De

. 2002. “Parent Beliefs Are Cognitions: The Dynamic Belief Systems Model.” Pp. 485–508 in Handbook of Parenting: Being and Becoming a Parent, edited by Bornstein

Mahwah, NJ: Lawrence Erlbaum.

60.

Stevenson

Harold W.

Lee

Shin Ying

Stigler

James W.

1986. “Mathematics Achievement of Chinese, Japanese, and American Children.” Science 231(4739):693–99.

61.

Swidler

Ann

. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(2):273.

62.

Swidler

Ann

. 2001. Talk of Love: How Culture Matters. Chicago: University of Chicago Press.

63.

Treiman

Donald J.

1970. “Industrialization and Social Stratification.” Pp. 207–34 in Social Stratification: Research and Theory for the 1970s, edited by Laumann

Indianapolis: BobbsMerrill.

64.

Ubeda

Paloma

. 2014. “The Consistency of Fairness Rules: An Experimental Study.” Journal of Economic Psychology 41:88–100.

65.

Vaisey

Stephen

. 2009. “Motivation and Justification: A Dual-Process Model of Culture in Action.” American Journal of Sociology 114(6):1675–1715.

66.

Vaisey

Stephen

. 2010. “What People Want: Rethinking Poverty, Culture, and Educational Attainment.” ANNALS of the American Academy of Political and Social Science 629(1):75–101.

67.

Vaisey

Stephen

Lizardo

Omar

. 2010. “Can Cultural Worldviews Influence Network Composition?” Social Forces 88(4):1595–1618.

68.

Vaisey

Stephen

Lizardo

Omar

. 2016. “Cultural Fragmentation or Acquired Dispositions? A New Approach to Accounting for Patterns of Cultural Change.” Socius: Sociological Research for a Dynamic World2.

69.

Vaisey

Stephen

Valentino

Lauren

. 2018. “Culture and Choice: Toward Integrating Cultural Sociology with the Judgment and Decision-Making Sciences.” Poetics 68:131–43.

70.

Van Overwalle

Frank

Metsenaere

Machteld De

. 1990. “The Effects of Attribution-Based Intervention and Study Strategy Training on Academic Achievement in College Freshmen.” British Journal of Educational Psychology 60(3):299–311.

71.

Verniers

Catherine

Martinot

Delphine

Dompnier

Benoît

. 2016. “The Feminization of School Hypothesis Called into Question among Junior and High School Students.” British Journal of Educational Psychology 86(3):369–81.

72.

Weber

Marx

. 1978. Economy and Society: An Outline of Interpretive Sociology. Berkeley: University of California Press.

73.

Weiner

Bernard

. 1989. Human Motivation. Hillsdale, NJ: Lawrence Erlbaum.

74.

Weininger

Elliot B.

Lareau

Annette

Conley

Dalton

. 2015. “What Money Doesn’t Buy: Class Resources and Children’s Participation in Organized Extracurricular Activities.” Social Forces 94(2):479–503.

75.

Wigfield

Allan

Eccles

Jacquelynne

Schiefele

Ulrich

Roeser

Robert

Davis-Kean

Pamela

. 2006. “Development of Achievement Motivation.” Pp. 933–1002 in Handbook of Child Psychology: Social, Emotional, and Personality Development, edited by Damon

Lerner

R. M.

Hoboken, NJ: Wiley.

76.

Williams

Richard

. 2009. “Using Heterogeneous Choice Models to Compare Logit and Probit Coefficients across Groups.” Sociological Methods & Research 37(4):531–59.

77.

Winship

Christopher

Mare

Robert D.

1984. “Regression Models with Ordinal Variables.” American Sociological Review 49(4):512.

78.

Jinting

. 2017. Fabricating an Educational Miracle: Compulsory Schooling Meets Ethnic Rural Development in Southwest China. New York: State University of New York Press.

79.

Xiaogang

. 2019. “Inequality and Social Stratification in Postsocialist China.” Annual Review of Sociology 45(1):363–82.

80.

Xian

Reynolds

Jeremy

. 2017. “Bootstraps, Buddies, and Bribes: Perceived Meritocracy in the United States and China.” Sociological Quarterly 58(4):622–47.

81.

Xiao

Hong

Chenyang

. 2013. “China’s Meritocratic Examinations and the Ideal of Virtuous Talents.” Pp. 340–62 in The East Asian Challenge for Democracy: Political Meritocracy in Comparative Perspective, edited by Bell

Cambridge, UK: Cambridge University Press.

82.

Xie

. 2016. “Understanding Inequality in China.” Chinese Journal of Sociology 2(3):327–47.

83.

Xie

Zhou

Xiang

. 2014. “Income Inequality in Today’s China.” Proceedings of the National Academy of Sciences of the United States of America 111(19):6928–33.

84.

Yang

Chung Fang

Chiu

Chi Yue

Chan

Kin Man

King

Ambrose

Cheung

Tak Sing

Chan

Hoi Man

. 2006. “How Confucian Are Contemporary Chinese? Construction of an Ideal Type and Its Application to Three Chinese Communities.” European Journal of East Asian Studies 5(2):157–80.

85.

Hua

. 2015. “Key-Point Schools and Entry into Tertiary Education in China.” Chinese Sociological Review 47(2):128–53.

86.

Young

Michael Dunlop

. 1958. The Rise of the Meritocracy. New York: Transaction.

87.

Zhang

Honglian

. 2013. The Mirage of Elite Schools: Evidence from Lottery-Based School Admissions in China. http://hongliangzhang.weebly.com/uploads/3/1/9/7/31972829/the_mirage_of_elite_schools_hongliangzhang.pdf

88.

Zhou

Xiang

Xie

. 2019. “Market Transition, Industrialization, and Social Mobility Trends in Postrevolution China.” American Journal of Sociology 124(6):1810–47.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB