The effects of the teacher–student relationship and academic press on student engagement and academic performance

International Journal of Educational Research 53 (2012) 330–340

Contents lists available at SciVerse ScienceDirect

International Journal of Educational Research journal homepage: www.elsevier.com/locate/ijedures

The effects of the teacher–student relationship and academic press on student engagement and academic performance Jung-Sook Lee University of New South Wales, Australia

A R T I C L E I N F O

A B S T R A C T

Article history: Received 28 January 2011 Received in revised form 19 April 2012 Accepted 19 April 2012 Available online 15 May 2012

This study examined relationships between students’ perceptions of the school social environment and student outcomes, using U.S. data from the Program for International Student Assessment 2000 (OECD, 2000). The sample comprised 3748 fifteen-year-old 9th and 10th graders from 147 schools. The two-dimensional approach of parenting typology was here applied to the school environment. The results partially supported the advantage of authoritative schools with high levels of both demandingness (academic press) and responsiveness (the teacher–student relationship). Supportive teacher–student relationships and academic press were significantly related to behavioral and emotional student engagement whereas only the teacher–student relationship was a significant predictor of reading performance. The effects of the teacher–student relationship on student outcomes were not contingent on academic press of the school. ß 2012 Elsevier Ltd. All rights reserved.

Keywords: Student engagement Academic performance Teacher–student relationship Academic press Authoritative style PISA

1. Introduction The present study investigated relationships between students’ perceptions of the school social environment and student outcomes (i.e., engagement at school and academic performance). To understand the relationships, the two-dimensional approach of parenting typology was applied to the school social environment. The two dimensions are demandingness (academic press) and responsiveness (the teacher–student relationship). 1.1. Literature review Research indicates that 25–60% of U.S. students are disengaged from school (Klem & Connell, 2004; Willms, 2003). This phenomenon is not unique to the United States and appears to be common and widespread. In a study using data from the Program for International Student Assessment 2000 (OECD, 2000), Willms (2003) found that 25% of students in the 43 countries studied reported a low sense of belonging, and 20% of students reported low participation. Lack of student engagement at school is a serious concern for educators and policy makers because disengaged students are more likely to struggle academically, to drop out of school, and to have problem behaviors (Fredricks, Blumenfeld, & Paris, 2004). Researchers and educators (e.g., Lee & Smith, 1993) have emphasized the significant influence of schools on student engagement and academic performance. Student engagement is a multifaceted concept. Researchers have identified several components of student engagement (e.g., behavioral, emotional/psychological, cognitive, and academic) (Appleton, Christenson, & Furlong, 2008; Fredricks et al., 2004). Although there is no consensus on which of these components is important, most studies have included behavioral

E-mail address: [email protected]. 0883-0355/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijer.2012.04.006

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

331

and emotional components. Behavioral engagement refers to the students’ participation in academic and nonacademic activities at school. Emotional engagement refers to the students’ sense of belonging at school and identification with school. Student engagement is a robust predictor of student success at school. Studies have reported a positive association between student engagement and academic achievement regardless of race, gender, and socio-economic status (SES) (e.g., Klem & Connell, 2004). Highly engaged students are also less likely to drop out of school (Finn & Rock, 1997). Taking a developmental perspective, academic failure and dropping out are not isolated events but the result of a long-term process of disengagement from school (Alexander, Entwisle, & Horsey, 1997; Randolph, Fraser, & Orthner, 2004). Thus, enhancing student engagement may help prevent these poor student outcomes. The reported magnitude of the effect of student engagement varies depending on the components of engagement that are examined. In a study by Willms (2003), behavioral engagement was defined as attendance and punctuality and it had a moderate correlation (.48–.51) with students’ literacy skills at the school level. Further, a positive relationship between behavioral engagement and academic performance was found in studies that examined efforts in learning (Carbonaro, 2005), attendance (Lamdin, 1996), and extracurricular activities (Jordan, 1999). The positive effect of behavioral engagement on academic performance seems to be more evident among academically resilient students (Borman & Overman, 2004; Finn & Rock, 1997). For example, in a study of 925 low-SES minority students, Borman and Overman (2004) found that greater engagement in academic activities was a characteristic shared by all students who were deemed to be academically resilient, defined by having higher than predicted math scores (predictions were based on previous math scores and individual SES). Evidence regarding the effect of emotional engagement on academic performance is mixed. Studies using measures of emotional engagement combined with behavioral engagement (Borman & Overman, 2004; Connell, Spencer, & Aber, 1994; Sirin & Rogers-Sirin, 2004) have generally found a positive relationship between engagement and academic performance. However, emotional engagement focusing on a sense of belonging or identification with school was not a strong predictor of academic performance in the PISA study (Willms, 2003) or in Finn’s study (1993). On the other hand, a study of 214 Mexican American high school students (Gonzalez & Padilla, 1997) found that the level of sense of belonging at school was significantly associated with the grade point average (GPA) of a student. Nonetheless, research has supported the significance of emotional engagement in the decision to drop out (e.g., Ekstrom, Goertz, Pollack, & Rock, 1986; Finn, 1989). For instance, in an ethnographic study of adolescents who dropped out of school (Fine, 1991), one of the primary reasons reported for dropping out was not being emotionally engaged with the school. Student engagement at school is influenced by various individual and family factors (e.g., gender, grade level, race/ ethnicity, language spoken at home, and socioeconomic status). For example, Finn (1989, as cited in Marks, 2000) found that girls were consistently more engaged than boys. Student engagement is also related to grade level. According to Klem and Connell (2004), students become less engaged with school as they progress from elementary to middle school and from middle to high school. There has been a consistent and significant achievement gap among racial/ethnic groups, and emotional disengagement of students of color has been offered as one explanation for this phenomenon. Steele (1997), for example, argued that African American students disidentify with school due to frustration caused by a ‘‘stereotype threat’’— the anxiety that their actions will confirm existing negative stereotypes about African Americans. For non-native English speakers, the language spoken at home can be related to student engagement because it may reflect the acculturation and English proficiency of students. Furthermore, it has been widely reported that family socioeconomic status (SES) is significantly related to student outcomes (Henderson & Berla, 1994; Henderson & Mapp, 2002). The current study, however, focused on the effect of the school environment, because school is the setting where student engagement actually occurs. Schools exert great influence on a student’s engagement by promoting or constraining an individual student’s opportunities for engagement. There has been a controversy over which classroom and school characteristics matter most, both for schools and for students (Gill, Ashton, & Algina, 2004; Lee & Smith, 1999). Some advocated academic press because they believed that valuing academic success and holding high academic expectations for students promote student achievement (e.g., Goddard, Sweetland, & Hoy, 2000; McDill, Natriello, & Pallas, 1986). However, this view was challenged by researchers who advocated a communal perspective that emphasizes shared values, supportive teacher–student relationships, and a caring atmosphere at school (e.g., Battistich, Solomon, Kim, Watson, & Schaps, 1995; Bryk & Driscoll, 1988). Academic press is the normative and behavioral environment of a school that emphasizes academic excellence and conformity to academic standards (Goddard et al., 2000; Lee & Smith, 1999). A normative environment emphasizing academic excellence creates expectations of behaviors of teachers and students and imposes sanctions when individuals deviate from the expectations (Goddard et al., 2000). Relevant terms used in the literature include academic emphasis, academic optimism, teacher expectations, and academic standards. Academic press has been reported to be positively associated with students’ sense of belonging, attendance, and academic performance (Goddard et al., 2000; Lee & Smith, 1999; Ma, 2003; Phillips, 1997). For example, in a study with 2429 students from 45 elementary schools, Goddard et al. (2000) found that academic emphasis by a school explained about 50% of between-school variability in mathematics and reading. Supportive teacher–student relationships are a critical aspect of the interpersonal climate in schools. Relevant concepts in the literature include teacher support, social capital in school, school as a community, and responsiveness of the school. Although different terms were used, studies found positive effects of supportive teacher–student relationships on various student outcomes. A supportive teacher–student relationship was positively related to social self-concept, school adjustment and grade, whereas it was negatively associated with externalizing behavioral problems, internalized symptoms, and school dropout (Baker, 2006; Brewster & Bowen, 2004; Hamre & Pianta, 2001; Meehan, Hughes, & Cavell,

332

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

2003; Miller, 2000; O’Connor, Dearing, & Collins, 2011). Students were more behaviorally and emotionally engaged when they had positive relationships with their teachers and this further contributed to their academic achievement (Hughes, Luo, Kwok, & Loyd, 2008; O’Connor & McCartney, 2007). Although studies reported that academic press and supportive teacher–student relationships each significantly predicted student outcomes, several studies examining multiple school factors simultaneously indicated the possible existence of optimal combinations of school factors (Gill et al., 2004; Pellerin, 2005; Shouse, 1996). For example, in a study with 30,000 sixth and eighth graders from 304 Chicago public elementary schools, Lee and Smith (1999) found that the relationship between social support and learning are contingent on the academic press of the school. The effects of social support on math and reading were greater in schools with higher levels of academic press. Shouse (1996), in a study of the mathematics achievement of 398 high school students, also found that high levels of communality in schools had a positive effect on academic performance in low-SES schools only when accompanied by high academic press and a disciplinary climate. Based on these findings, Luyten, Visscher, and Witziers (2005) suggested that examining meaningful combinations of school factors may be more important than studying a single factor and may, in fact, better reflect the complex school realities experienced by students. 1.2. Theoretical framework In an effort to identify the optimal combinations of school factors, a few researchers (Gill et al., 2004; Pellerin, 2005; Wentzel, 2002) have tested the typology of parenting style in school settings. An authoritative style, with both demandingness (e.g., academic press) and responsiveness (e.g., supportive relationship), is expected to provide the optimal conditions to achieve best student outcomes. The self-system processes model (Connell, 1990) and self-determination theory (Ryan & Deci, 2000a) help to explain the underlying mechanisms. Children learn socially desirable behaviors and rules through the process of socialization, which allows them to adequately participate in a society (Handel, Cahill, & Elkin, 2007). One of the well-known conceptual frameworks in the socialization research is Baumrind’s (1967) typology of parenting style: authoritative, authoritarian, and permissive. Maccoby and Martin (1983) enhanced the generalizability of Baumrind’s typology by creating a two-dimensional framework based on the levels of demandingness and responsiveness of the socialization agent—the parent. Demandingness is the socialization agent’s willingness to socialize children to facilitate their integration into society; responsiveness is the socialization agent’s recognition of a child’s individuality. Accordingly, Maccoby and Martin added one more style (i.e., indifferent/neglectful) to the Baumrind’s typology. Studies of parenting style have reported that authoritative parenting, characterized by high demandingness and high responsiveness, is the most effective among these parenting styles (see, e.g., Aunola, Stattin, & Nurmi, 2000; Darling & Steinberg, 1993). Authoritative parents show high levels of involvement in their children’s lives, behavioral control, and monitoring while also providing emotional support, open communication, trust of the child, and encouragement of psychological autonomy (Aunola et al., 2000; Slicker, 1998). As a result, children with authoritative parents have more adaptive achievement strategies (Aunola et al., 2000), fewer problem behaviors (Querido, Warner, & Eyberg, 2002), higher levels of performance, and higher levels of school engagement (Steinberg, Lamborn, Dornbusch, & Darling, 1992). Although socialization style was originally developed to explain interpersonal phenomena between children and parents, the two-dimensional approach of parenting style has also been applied to teachers and schools. Demandingness of teachers and schools was represented by academic press, high expectations, and an orderly disciplinary climate, whereas responsiveness included supportive teacher–student relationships, a supportive school climate, and shared values. Studies found that an authoritative style was also effective in school settings. Pellerin (2005), for example, found that 10th and 11th graders in authoritative schools showed the lowest levels of behavioral disengagement, defined as absenteeism, tardiness, or turning in unfinished homework, whereas students in indifferent or neglectful schools presented the highest levels of disengagement. The self-system processes model (Connell, 1990) and self-determination theory (Ryan & Deci, 2000a) help to explain the mechanism by which the social environment of a school influences student engagement and achievement. The essence of socialization is to develop self-regulation of one’s own behavior (Handel et al., 2007). In the development of self-regulation, it is important to help children internalize the values and rules of the society. It is believed that children are more likely to internalize and adopt values and rules when their relationships with the socialization agents are nurturing and supportive (Grusec & Hastings, 2007; Ryan & Deci, 2000a). Self-system processes are the result of a dialectic relationship between an individual’s psychological needs (i.e., competence, relatedness, and autonomy) and social context (Connell, 1990). Social contexts that provide individuals with structure, autonomy support, and involvement promote the development of self-system processes because they satisfy an individual’s psychological needs. As a result, individuals show a desired action, which in turn produces a desired outcome. When applied to the school setting, the social context of a school that satisfies an individual’s psychological needs encourages the desired action of student engagement, which in turn produces the desired outcome of improved academic performance and student achievement. In this process, student engagement is pivotal because it is the link through which social context and student self-system influence achievement (Tucker et al., 2002). According to the tenets of self-determination theory (SDT), social contexts that support an individual’s psychological needs (i.e., competence, relatedness, and autonomy) promote motivated actions by facilitating the internalization of

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

333

extrinsic motivation (Ryan & Deci, 2000a). There is an important distinction between motivation and engagement. If ‘‘to be motivated’’ is ‘‘to be moved to do something’’ (Ryan & Deci, 2000a, p. 54), then ‘‘to be engaged’’ is ‘‘to do something’’ in a broader sense. Despite the distinction, the theory facilitates an understanding of the reasons behind student engagement. Support for the psychological needs of human beings (i.e., competence, relatedness, and autonomy) is critical in internalizing extrinsic motivation as well as maintaining intrinsic motivation (Deci, Vallerand, Pelletier, & Ryan, 1991). In other words, the primary reason that people are willing to do uninteresting activities is that these activities are valued by significant others with whom they feel connected or have a sense of relatedness. Authoritative schools with high levels of both demandingness and responsiveness can enhance student engagement and academic performance by providing optimal conditions to facilitate students’ self-system processes and their internalization of extrinsic motivation. The social environments of authoritative schools that satisfy a student’s psychological needs encourage the desired action of student engagement and produce the desired outcome of improved academic performance. Authoritative schools with high academic expectations allow students to develop the self-regulation and skills necessary to function adequately in society. Authoritative schools supporting positive interpersonal relationships may promote a student’s sense of relatedness. Further, supportive relationships lead students to internalize extrinsic motivation to learn, because learning is valued by teachers with whom they feel connected. Even in the same classroom and the same school, students have different relationships with teachers and they perceive academic press differently depending on their motivation to learn. Thus, this study focused on students’ perceptions of the school social environment. According to the two-dimensional approach of parenting style, it is expected that students who experience high levels of both responsiveness (i.e., the teacher–student relationship) and demandingness (i.e., academic press) present higher levels of engagement and academic performance. Therefore, the current study tested three following hypotheses: (1) The teacher–student relationship is positively associated with student engagement and academic performance. (2) Academic press is positively associated with student engagement and academic performance. (3) The associations between the teacher–student relationship and student outcomes differ by the level of academic press.

2. Methods 2.1. Study design Analyses were conducted on the U.S. data taken from an international cross-sectional data set, the Program for International Student Assessment 2000 (PISA 2000) (OECD, 2000). Literacy, mathematics, and science skills were assessed by the Organisation for Economic Co-operation and Development (OECD) at three different time points, placing emphasis on a different outcome domain each year. The present study uses PISA results from 2000, when reading literacy was the outcome domain of interest. Forty-three countries took part in the development and administration of PISA 2000. In addition to performance data, PISA 2000 collected detailed and comprehensive information from students and school principals about psychological, social, economic, cultural, and educational factors related to student performance. 2.2. Sample The target population of PISA 2000 (OECD, 2000) was 15-year-old students. In the United States, a three-stage stratified sampling procedure was used. The first-stage or primary sampling units were geographical areas, the secondstage units were schools, and the third-stage units were students. Fifty-two geographic areas were chosen, out of which 220 schools were sampled initially. In each participating school, up to 35 eligible students were randomly selected. Schools with student response rates of more than 50% were categorized as responding schools (Adams & Wu, 2002). Because less than 85% of initially selected schools achieved more than 50% student response rates, replacement schools were used to achieve an acceptable school response rate. With 116 initially selected schools and 29 replacement schools, 145 schools participated. As a result, the U.S. data includes 3700 students in 145 schools. Weighted student response rates were 85% and weighted school response rates were 70% after replacement (Adams & Wu, 2002). With the addition of 146 students from the partially responding schools, 3846 U.S. students were included in the international data. As the unit of analysis is students, the current study used the U.S. sample in the international data to make full use of data collected. For the present study, the sample was limited to 9th and 10th graders (n = 3754), who comprised the vast majority (98%) of the U.S. sample. Of the 9th- and 10th-grade students, 6 students without data on all three outcome variables were dropped from the current study. As a result, the final study sample included 3748 students from 147 schools. 2.3. Measures Three outcome measures were behavioral engagement, emotional engagement, and academic performance. Behavioral engagement was defined in this study as effort and perseverance in learning activities. Behavioral engagement scores were

334

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

derived from four items: working hard, working despite difficulty, trying one’s best to acquire knowledge and skills, and putting forth one’s best effort. The internal consistency reliability was a = .83. Emotional engagement refers to a sense of belonging at school (a = .86). This measure was derived from students’ responses to questions about the degree to which they agree that the school is a place where they feel that they belong, where they make friends easily, and where other students seem to like them, or conversely, where they feel awkward and out of place, feel like an outsider, or feel lonely. Academic performance was measured as reading literacy. PISA defines reading literacy as ‘‘capacity to understand, use and reflect on written texts, in order to achieve one’s goals, to develop one’s knowledge and potential, and to participate in society’’ (Kirsch et al., 2002). Reading literacy includes three domains: processing skills, knowledge and understanding, and the context of application. The measure of the teacher–student relationship (a = .83) was derived from five items about the degree to which students agree that teachers get along well with students, are interested in students’ well-being, really listen to what students have to say, will provide extra help when students need it, and treat students fairly. The measure of academic press (a = .54) was derived from four items about the students’ English teachers: the teacher wants students to work hard, tells students that they can do better, does not like students’ careless work, and expects students to learn a lot. All questions had four ordinal response categories (never, some lessons, most lessons, or every lesson; strongly disagree, disagree, agree, or strongly agree). According to the PISA 2000 technical report (Adams & Wu, 2002), the measures used in this study have moderate to high reliabilities, with Cronbach’s alphas (a) above .80; the exception is academic press (a = .54). The reliability of academic press falls in the range of poor (a < 60) but not unacceptable (a < 50) (George & Mallery, 2005). Factor analyses have shown that the measures have adequate construct validity (Adams & Wu, 2002). The present study used indexes of student and school characteristics that were provided by PISA 2000 (OECD, 2000). Index scale scores were created by PISA researchers using Rasch item response model techniques. All scale scores were weighted likelihood estimates (Warm, 1989). Grade, gender, race/ethnicity, language spoken at home, individual SES, school-mean SES, and school type were covariates in the analysis. Tenth grade was coded as 1 whereas ninth grade was the reference group. Female was coded as 1 and male was coded as 0. Using European American as the reference group, three dummy variables of race were created to indicate African American, Latino/Hispanic, and other race. Language spoken at home was categorized into English (coded as 1) and other languages (coded as 0). To measure students’ individual SES, PISA used the International Socio-Economic Index of Occupational Status (ISEI). Students’ ISEI scores were derived from information they provided about parental occupation and were designed to capture the attributes of occupation that convert parents’ education to income (Kirsch et al., 2002). The highest occupational status of parents (HISEI) is the highest ISEI value of a student’s father, mother, or guardian(s). In the present study, HISEI divided by 10 was used as a measure of individual SES, to reduce the difference in scale. School-mean SES was added to separate the contextual effect from the effect of individual SES. School-mean SES represents the average SES in each school. Using private schools as the reference group, two dummy variables of school type were created to indicate public schools and unknown type schools. 2.4. Data analysis Multilevel analysis was run in SAS 9.2 (SAS Institute Inc., 2008). Multilevel models were estimated to adjust standard errors for intraclass correlations because students were nested within schools. Although a three-stage sampling procedure was used for the PISA study in the U.S., the primary sampling units—geographical regions—were not included as a level in the models because the effect of geographical regions was not a substantive interest of the present study. Moreover, in a simulation study using PISA 2000, Rabe-Hesketh and Skrondal (2006) reported that ‘‘taking [primary sampling units] into account did not necessarily increase the standard errors’’ (p. 818). Thus, the present study utilized two-level models, with students at Level 1 and schools at Level 2. Intraclass correlations for each of the three components of student outcomes were calculated to examine the variation among schools. Random intercepts and slopes were adopted to include the unmeasured heterogeneity between schools. Continuous variables were grand-mean centered to adequately control for the Level 1 covariates. Main effect models. Dependent variables (Yij) were behavioral engagement, emotional engagement, and reading performance. The Level 1 model for each outcome included the individual-level demographic covariates of gender, grade, race/ethnicity, language spoken at home, SES, the teacher–student relationship, and academic press. The Level 1 equation was as follows:

Y i j ¼ p0 j þ p1 j ðFEMALEÞi j þ p2 j ð10thÞi j þ p3 j ðAFRÞi j þ p4 j ðLATINOÞi j þ p5 j ðOTHERÞi j þ p6 j ðENGLISHÞi j þ p7 j ðSESÞi j þ p8 j ðRELATIONÞi j þ p9 j ðPRESSÞi j þ r i j ; where Yij is the outcome for student i in school j and p0j is the mean outcome in school j; p1j is the gender difference; p2j is the grade difference (9th and 10th); p3j, p4j, and p5j, are the racial differences of outcomes for African American, Latino/Hispanic, and other race when European American is the reference group; p6j is the difference due to the language spoken at home; p7j is the degree to which differences in students’ SES related to outcomes; p8j and p9j are the effects of the teacher–student relationship and academic press on outcome, respectively; and rij is the residual variance within schools.

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

335

In the Level 2 equations, the mean of each outcome from the Level 1 equations (p0j) was regressed on school type and the variable for school-mean SES. The Level 2 equations were modeled as follows:

p0 j ¼ b00 þ b01 ðSSESÞ j þ b02 ðPUBLICÞ j þ b03 ðUNKNOWNÞ j þ u0 j p p j ¼ b p0 þ u p j ;

p > 0;

where b00 is the grand mean of outcome; b01 is the degree to which differences in outcomes between schools related to school-mean SES; b02 and b03 are the mean differences of outcome for public schools and unknown type schools when private schools serve as the reference group; and u0j is the residual variance among schools. Interaction effect models. To investigate whether the associations between the teacher–student relationship and student outcomes differed according to the level of academic press (Hypothesis 3), interaction effects were examined. The interaction models differed from the main effect models in that the interaction terms between the teacher–student relationship and academic press were added to the Level 1 equations. Explained variance. Explained variance indicated how much variability of the dependent variable was accounted for by explanatory variables in the model (Snijders & Bosker, 1999). To calculate the explained variance, the present study used the proportional reduction of prediction error suggested by Snijders and Bosker. To reduce possible bias due to missing values, the multiple imputation procedure was employed. The percentage of missingness varied from 0 (grade) to 15.3 (socioeconomic status). Therefore, the current study conducted data analyses using the multiple imputation method because it allowed for valid statistical inference (Fichman & Cummings, 2003). Note, however, that outcome variables were not imputed. 3. Results 3.1. Sample description Behavioral engagement, emotional engagement, the teacher–student relationship, and academic press were index scale scores with means around zero and standard deviations close to 1. The mean reading score was 508.82 (SD = 100.00). Slightly more than half of the students were female and about 60% were in 10th grade. A majority of the students were European American (62.66%) and spoke English at home (90.16%). Individual SES scores varied widely ranging from 1.6 to 9.0 with a mean of 5.27 (SD = 1.65). This meant that the average parental occupational level was equivalent to sales workers or engineering technicians (SES = 5.3). Schools also varied greatly on the level of school-mean SES, although the range was narrower than the individual SES. The school-mean SES ranged from 3.15 to 6.9 and the average was 5.07 (SD = 0.72). About 75% of the schools were public and 5% were private; approximately 20% of the schools did not report school type. Table 1 presents descriptive statistics using standardized final weights. Intraclass correlations were .04 for behavioral engagement, .03 for emotional engagement, and .21 for reading performance. Although intraclass correlations were relatively small, all were statistically significant at the .01 level. In the literature, using multilevel analysis was suggested with intraclass correlations as low as .02 (Kreft & de Leeuw, 1998). Moreover, conducting multilevel analysis was necessary due to the multilevel data structure. 3.2. Main effect models As shown in Table 2, the teacher–student relationship had significantly positive associations with all three student outcomes. When all covariates were held at their means, a one-unit increase in the teacher–student relationship predicted a 0.33-unit increase in behavioral engagement (p < .001), a 0.29-unit increase in emotional engagement (p < .001) and a 10.96-unit increase in reading performance (p < .001). In other words, when students perceived that they had positive relationships with teachers, they presented higher levels of effort and perseverance in learning, were more likely to feel that they belonged to a school, and had higher reading scores. Academic press was a significant predictor for behavioral engagement (b = 0.13, p < .001) and emotional engagement (b = 0.06, p < .01) but not for reading performance. When students perceived that their teachers showed greater emphasis on academic excellence, they were more likely to present higher levels of effort and perseverance in learning and felt they belonged to a school. The coefficients of other covariates generally showed expected directions. The Level 1 explained variance was about 17%, 10%, and 29% for behavioral engagement, emotional engagement, and reading performance, respectively. The Level 2 explained variance was about 6%, 8%, and 65% for behavioral engagement, emotional engagement, and reading performance in that order. 3.3. Interaction effect models Some studies suggested that the effects of the teacher–student relationship on student outcomes differed by the level of academic press (e.g., Lee & Smith, 1999; Shouse, 1996). To verify this hypothesis, I tested interaction effects between the

336

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

Table 1 Descriptive statistics.

Outcome variables Behavioral engagement Emotional engagement Reading performance Student level predictors Gender Male Female Grade 9th grade 10th grade Race/ethnicity European American African American Hispanic/Latino Other race Language at home English Other language Socioeconomic status (SES) Perceived school environment Teacher–student relationship Academic press School level predictors School-mean SES School type Public school Private school Unknown type

n

Mean (SD)/percent

3278 3509 3747

0.07 (1.10) 0.05 (1.11) 508.82 (100.00)

1786 1961

47.66 52.34

1547 2201

41.27 58.73

2348 478 607 314

62.66 12.75 16.20 8.39

3184 348 3174

90.16 9.84 5.27 (1.65)

3512 3536

0.21 (1.02) 0.41 (1.06)

142

5.07 (0.72)

110 7 30

74.83 4.76 20.41

Note. Other race includes Asian American, American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or Multiracial students. SD = standard deviation. Descriptive statistics using standardized final weights were reported. Number of students = 3748 and number of schools = 147.

teacher–student relationship and academic press on behavioral engagement, emotional engagement, and reading performance. None of the tested interactions was significant at the .05 level. 4. Discussion Using the two-dimensional approach of parenting style, the current study examined relationships between the perceived social environment of the school (the teacher–student relationship and academic press) and student outcomes (behavioral engagement, emotional engagement, and academic performance). The current study partially supported the advantage of authoritative schools with both demandingness and responsiveness. Interaction effects between the teacher–student relationship and academic press on student outcomes were also tested but none was found to be significant. As expected by the two-dimensional approach of parenting style, both the teacher–student relationship and academic press were significant predictors of behavioral engagement. In the present study, academic press was measured by students’ perceptions about the existence of pressure toward academic excellence, and behavioral engagement was measured by effort and perseverance in learning. It is understandable that students showed higher levels of effort and perseverance in learning when they felt pressured to perform well academically. A normative environment emphasizing academic excellence creates expectations that students will work hard to achieve academically and imposes sanctions for those who deviate from the expectations. Thus, students were more likely to behaviorally engage when they were in classrooms and schools with higher levels of academic press. However, an individual’s willingness to act in accordance with the expectations may differ between students. According to the self-determination theory (Ryan & Deci, 2000b), individuals become willing to perform desired actions when they internalize the values and expectations of socializing agents. An individual’s interpersonal relationship with socializing agents is believed to be critical to the process of internalization. Thus, it is understandable that students who had positive relationships with teachers made more effort and persevered in learning because they internalized the academic values and expectations appreciated by teachers. Similarly, emotional engagement was significantly predicted by the teacher–student relationship and academic press. Emotional engagement, defined as a sense of belonging at school, is an affective component and it can easily be influenced by the interpersonal relationships in the school, especially by the relationships with teachers. According to the self-system processes model (Connell, 1990), an individual develops felt security with self and others when social context satisfies his or her psychological need for relatedness. When students feel supported by their teachers, they also feel secure within themselves and with others. This will lead them to feel that they are part of the classroom and school community. In a sense, academic press can be also considered as a form of teacher support. In particular, when it is delivered in a caring manner,

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

337

Table 2 Effects of the perceived school social environment on student outcomes. Behavioral engagement Estimate Fixed effects Intercept Gender (male) Female Grade (9th grade) 10th grade Race (European American) African American Latino/Hispanic Other race Language at home (other) English SES Teacher–student relationship Academic press School-mean SES School type (private) Public school Unknown type Random effects Intercept English SES Teacher–student relationship Academic press Residuals Deviance Percent of variance explained Within schools Between schools

SE

Emotional engagement

Reading performance

Estimate

Estimate

SE

SE

0.40

0.24

0.05

0.21

329.68***

22.14

0.29***

0.04

0.05

0.04

22.24***

2.76

0.04

0.04

0.03

0.04

42.29***

3.02

***

***

0.27 0.09 0.08

0.06 0.06 0.07

0.01 0.02 0.20**

0.06 0.06 0.07

46.93 31.73*** 14.14*

5.06 4.81 5.49

0.13 0.07*** 0.33*** 0.13*** 0.09*

0.07 0.01 0.02 0.02 0.04

0.19** 0.03* 0.29*** 0.06*** 0.03

0.07 0.01 0.02 0.02 0.03

11.59* 10.43*** 10.96*** 0.67 29.89***

5.79 1.19 1.49 1.57 3.64

0.14 0.10

0.10 0.11

0.05 0.08

0.09 0.09

0.02*

*

0.02 0.01* 0.96*** 9679.4 16.70 5.53

16.81 10.89

9.83 10.55

0.01

0.01 0.01 0.03

***

0.01

1.08***

0.03

0.03

10,728.3 9.70 8.21

535.87*** 30.78*

102.10 15.36

6584.18***

159.50

44,213.9 29.00 65.24

Note. SE = standard error. Reference groups are in parentheses. Sample sizes are 3278 for behavioral engagement, 3509 for emotional engagement, and 3747 for reading performance. * p < .05. ** p < .01. *** p < .001.

teachers’ high expectations for all students can create a positive social environment at school. In contrast, when teachers do not expect a lot from students, this may create a social environment that alienates some students, especially low performing students. This may explain the positive relationship between academic press and emotional engagement of students. However, the hypothesized advantage of the authoritative school environment in relation to academic performance was not supported in that only the teacher–student relationship was a significant predictor of reading performance. As explained above, a possible explanation is that when students have positive relationships with teachers, they internalize the academic values and expectations appreciated by teachers. Consequently, these students perform better on reading. However, it is also possible that students with higher academic performance maintain good relationships with their teachers. Unexpectedly, the effect of academic press on reading performance was not significant. Low reliability of the measure of academic press could be the source of a non-significant association. Other possible explanations are cumulative effects of previous learning, dominant influence of individual and family background on learning, and mediated effects through student engagement. Further research is warranted to ascertain any of these reasons. There was no significant interaction effect between the teacher–student relationship and academic press on any outcome examined. This means that the teacher–student relationship and academic press had independent effects on student engagement and academic performance. The result is different from findings of other studies (Lee & Smith, 1999; Shouse, 1996), which have argued that the positive effect of communality of school was contingent on the academic press of the school. The finding suggests that a nurturing school environment could be critical in itself, regardless of school demandingness (i.e., academic press). Overall, the current study partially supported the application of the two-dimensional approach of parenting typology to the school environment: demandingness and responsiveness. Both the teacher–student relationship and academic press were positive predictors of behavioral and emotional engagement at school. In other words, students who perceived higher levels of both demandingness and responsiveness of teachers (i.e., authoritative style) presented the highest levels of effort and perseverance in learning and sense of belonging. Thus, an authoritative school environment seems to provide optimal conditions to facilitate a student’s behavioral and emotional engagement. Although these are based on students’ perceptions of the school social environment, the findings provide some insights into the optimal school conditions to promote student engagement.

338

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

Although creating a school environment with both demandingness and responsiveness seems to be most beneficial to enhance student outcomes, the significance of a nurturing school environment cannot be overly emphasized. In the current study, the teacher–student relationship was a significant predictor of all student outcomes examined. Recent educational reforms that focus narrowly on test scores and academic achievement have raised serious concerns for educators. Schools may have difficulty providing a nurturing environment when they are under pressure from an accountability system that focuses solely on the scores of standardized tests. Improving test scores is indeed important; however, it needs to be recognized that the behavioral, emotional, and social development of children is as important as their intellectual development and can, in fact, influence their learning. 5. Limitations Despite the rigorous methods used, the present study had limitations related to measures, data, and analysis. PISA 2000 (OECD, 2000) included measures covering various aspects of students’ lives and school conditions as well as academic performance. Most of these measures have good reliability and validity. However, the measure of academic press had low reliability although it was not unacceptable. This could be the source of a non-significant association with reading performance. More importantly, low reliability raises a concern about the consistency of the measure. One possible reason is that the indicators contain various dimensions. Academic press can be expressed in a positive and/or negative way and teachers can focus on the quantity and/or quality of a student’s work. This needs to be further investigated. Academic press is one of the major concerns for researchers who are interested in the school climate and environment. Unfortunately, low reliability of the measure reduces the potential of PISA data to be used in this area of research. Given the low reliability, it is desirable to develop better measures of academic press in future studies. The main focus of the present study was the effect of the perceived school social environment, controlling for student demographics. What happens at school indeed distinguishes high-performing schools from low-performing schools (Craig et al., 2005). Nonetheless, the influence of a student’s individual characteristics and his/her family also should be considered. Studies have reported, for example, that individual characteristics (e.g., self-efficacy or educational aspiration) and familyrelated factors other than a student’s demographics (e.g., parental expectation or educational involvement) influence student engagement (Henderson & Berla, 1994; Henderson & Mapp, 2002; Maton, Hrabowski, & Greif, 1998). In the future, it would be worthwhile to examine the joint effects of individual, family, and school factors. A final limitation is related to the cross-sectional nature of the data. It is not clear whether the school social environment is an influence or a consequence. Students may be highly engaged because of a supportive social environment at the school; alternatively, teachers may maintain good relationships with students and stress academic excellence because students are highly engaged. With the limited information provided by PISA 2000, it was also not possible to control for previous engagement and performance. High school students are under the influence of various educational institutions before they enter their high schools. Without controlling for previous student engagement and performance, it may be difficult to isolate the effect of the current school. Longitudinal studies may help resolve these problems. 6. Conclusion There has been a debate over the importance of academic press and supportive school environments to achieve best student outcomes. The findings of the current study partially supported the advantage of authoritative schools with both demandingness (i.e., academic press) and responsiveness (i.e., the teacher–student relationship). It was also found that the effects of the teacher–student relationship on student outcomes were not contingent on academic press of the school. The findings of the current study provide useful information to resolve the debate over the optimal school environment and to help educators and practitioners develop effective interventions to promote student engagement. Acknowledgments This paper is largely based on my doctoral dissertation. I would like to express my gratitude to Associate Professor Natasha K. Bowen for her guidance throughout the doctoral program at the University of North Carolina at Chapel Hill. I extend thanks to committee members for their advice. I gratefully acknowledge anonymous reviewers, Emeritus Professor Ralph Hall, Ms. Karen Heycox, and Dr. Wendy Foote for their helpful suggestions. References Adams, R., & Wu, M. (Eds.). (2002). PISA 2000 technical report. Paris: Organisation for Economic Co-operation and Development (OECD). Alexander, K. L., Entwisle, D. R., & Horsey, C. S. (1997). From first grade forward: Early foundations of high school dropout. Sociology of Education, 70, 87–107. Appleton, J., Christenson, S., & Furlong, M. (2008). Student engagement with school: Critical conceptual and methodological issues of the construct. Psychology in the Schools, 45, 369–386. Aunola, K., Stattin, H., & Nurmi, J. E. (2000). Parenting styles and adolescents’ achievement strategies. Journal of Adolescence, 23, 205–222. Baker, J. (2006). Contributions of teacher–child relationships to positive school adjustment during elementary school. Journal of School Psychology, 44, 211–229. Battistich, V., Solomon, D., Kim, D. I., Watson, M., & Schaps, E. (1995). Schools as communities, poverty levels of student populations, and students attitudes, motives, and performance—a multilevel analysis. American Educational Research Journal, 32, 627–658. Baumrind, D. (1967). Child care practices anteceding 3 patterns of preschool behavior. Genetic Psychology Monographs, 75, 43–88.

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

339

Borman, G. D., & Overman, L. T. (2004). Academic resilience in mathematics among poor and minority students. Elementary School Journal, 104, 177–195. Brewster, A., & Bowen, G. (2004). Teacher support and the school engagement of Latino middle and high school students at risk of school failure. Child and Adolescent Social Work Journal, 21, 47–67. Bryk, A. S., & Driscoll, M. E. (1988). The school as community: Theoretical foundations, contextual influences, and consequences for students and teachers. Chicago: University of Chicago, Benton Center for Curriculum and Instruction. Carbonaro, W. (2005). Tracking, students’ effort, and academic achievement. Sociology of Education, 78, 27–49. Connell, J. P. (1990). Context, self, and action: A motivational analysis of self-system processes across the life span. In D. Cicchetti & M. Beeghly (Eds.), The self in transition: Infancy to childhood (pp. 61–98). Chicago, IL: University of Chicago. Connell, J. P., Spencer, M. B., & Aber, J. L. (1994). Educational risk and resilience in African-American youth: Context, self, action, and outcomes in school. Child Development, 65, 493–506. Craig, J., Butler, A., Cairo, L., Wood, C., Gilchrist, C., Holloway, J., et al. (2005). A case study of six high-performing schools in Tennessee. Charleston, WV: Appalachia Educational Laboratory at Edvantia. Darling, N., & Steinberg, L. (1993). Parenting style as context: An integrative model. Psychological Bulletin, 113, 487–496. Deci, E. L., Vallerand, R. J., Pelletier, L. G., & Ryan, R. M. (1991). Motivation and education: The self-determination perspective. Educational Psychologist, 26, 325–346. Ekstrom, R. B., Goertz, M. E., Pollack, J. M., & Rock, D. A. (1986). Who drops out of high school and why? Findings from a national study. Teachers College Record, 87, 356–373. Fichman, M., & Cummings, J. N. (2003). Multiple imputation for missing data: Making the most of what you know. Organizational Research Methods, 6, 282–308. Fine, M. (1991). Framing dropouts: Notes on the politics of an urban high school. Albany: State University of New York Press. Finn, J. D. (1989). Withdrawing from school. Review of Educational Research, 59, 117–142. Finn, J. D. (1993). School engagement and students at risk. Washington, DC: National Center for Education Statistics. Finn, J. D., & Rock, D. A. (1997). Academic success among students at risk for school failure. Journal of Applied Psychology, 82, 221–234. Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of the concept, state of the evidence. Review of Educational Research, 74, 59–109. George, D., & Mallery, P. (2005). SPSS for Windows step by step: A simple guide and reference, 12.0 update (5th ed.). Boston: Pearson Education. Gill, M. G., Ashton, P., & Algina, J. (2004). Authoritative schools: A test of a model to resolve the school effectiveness debate. Contemporary Educational Psychology, 29, 389–409. Goddard, R., Sweetland, S., & Hoy, W. (2000). Academic emphasis of urban elementary schools and student achievement in reading and mathematics: A multilevel analysis. Educational Administration Quarterly, 36, 683–702. Gonzalez, R., & Padilla, A. M. (1997). The academic resilience of Mexican American high school students. Hispanic Journal of Behavioral Sciences, 19, 301–317. Grusec, J. E., & Hastings, P. D. (Eds.). (2007). Handbook of socialization. New York: Guilford. Hamre, B., & Pianta, R. (2001). Early teacher–child relationships and the trajectory of children’s school outcomes through eighth grade. Child Development, 72, 625–638. Handel, G., Cahill, S. E., & Elkin, F. (2007). Children and society: The sociology of children and childhood socialization. Los Angeles, CA: Roxbury Publishing. Henderson, A. T., & Berla, N. (1994). A new generation of evidence: The family is critical to student achievement. Washington, DC: National Committee for Citizens in Education. Henderson, A. T., & Mapp, K. L. (2002). A new wave of evidence: The impact of school, family, and community connections on student achievement. Austin, TX: National Center for Family & Community Connections with Schools, Southwest Educational Development Laboratory. Hughes, J., Luo, W., Kwok, O., & Loyd, L. (2008). Teacher–student support, effortful engagement, and achievement: A 3-year longitudinal study. Journal of Educational Psychology, 100, 1–14. Jordan, W. J. (1999). Black high school students’ participation in school-sponsored sports activities: Effects on school engagement and achievement. Journal of Negro Education, 68, 54–71. Kirsch, I., De Jong, J., Lafontaine, D., McQueen, J., Mendelovits, J., & Monseur, C. (2002). Reading for change. Paris: Organisation for Economic Co-operation and Development (OECD). Klem, A. M., & Connell, J. P. (2004). Relationships matter: Linking teacher support to student engagement and achievement. Journal of School Health, 74, 262–273. Kreft, I., & de Leeuw, J. (1998). Introducing multilevel modeling. Thousand Oaks, CA: Sage. Lamdin, D. J. (1996). Evidence of student attendance as an independent variable in education production functions. Journal of Educational Research, 89, 155–162. Lee, V. E., & Smith, J. B. (1993). Effects of school restructuring on the achievement and engagement of middle-grade students. Sociology of Education, 66, 164–187. Lee, V. E., & Smith, J. B. (1999). Social support and achievement for young adolescents in Chicago: The role of school academic press. American Educational Research Journal, 36, 907–945. Luyten, H., Visscher, A., & Witziers, B. (2005). School effectiveness research: From a review of the criticism to recommendations for further development. School Effectiveness and School Improvement, 16, 249–279. Ma, X. (2003). Sense of belonging to school: Can schools make a difference? Journal of Educational Research, 96, 340–349. Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family: Parent–child interaction. In P. H. Mussen & E. M. Hetherington (Eds.), Handbook of child psychology: Vol. 4. Socialization, personality, and social development (4th ed., pp. 1–101). New York: Free Press of Glencoe. Marks, H. M. (2000). Student engagement in instructional activity: Patterns in the elementary, middle, and high school years. American Educational Research Journal, 37, 153–184. Maton, K. I., Hrabowski, F. A., & Greif, G. L. (1998). Preparing the way: A qualitative study of high-achieving African American males and the role of the family. American Journal of Community Psychology, 26, 639–668. McDill, E. L., Natriello, G., & Pallas, A. (1986). A population at risk: Potential consequences of tougher school standards for student dropouts. American Journal of Education, 94, 135–181. Meehan, B., Hughes, J., & Cavell, T. (2003). Teacher–student relationships as compensatory resources for aggressive children. Child Development, 74, 1145–1157. Miller, S. (2000). Falling off track: How teacher–student relationships predict early high school failure rates. Paper presented at the annual meeting of the American educational research association, New Orleans, LA. O’Connor, E., Dearing, E., & Collins, B. (2011). Teacher–child relationship and behavior problem trajectories in elementary school. American Educational Research Journal, 48, 120–162. O’Connor, E., & McCartney, K. (2007). Examining teacher–child relationships and achievement as part of an ecological model of development. American Educational Research Journal, 44, 340–369. OECD (2000). U.S. national data, program for international student assessment (PISA), 2000 [CD-ROM] (NCES 2004-006). Paris, France: Organisation for Economic Cooperation and Development [producer], 2000. Washington, DC: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics [distributor], 2004. Pellerin, L. A. (2005). Student disengagement and the socialization styles of high schools. Social Forces, 84, 1159–1179. Phillips, M. (1997). What makes schools effective? A comparison of the relationships of communitarian climate and academic climate to mathematics achievement and attendance during middle school. American Educational Research Journal, 34, 633–662. Querido, J. G., Warner, T. D., & Eyberg, S. M. (2002). Parenting styles and child behavior in African American families of preschool children. Journal of Clinical Child and Adolescent Psychology, 31, 272–277. Rabe-Hesketh, S., & Skrondal, A. (2006). Multilevel modelling of complex survey data. Journal of the Royal Statistical Society: Series A (Statistics in Society), 169, 805–827. Randolph, K. A., Fraser, M. W., & Orthner, D. K. (2004). Educational resilience among youth at risk. International Journal of Substance Use and Misuse, 39, 747–767. Ryan, R. M., & Deci, E. L. (2000a). Intrinsic and extrinsic motivations: Classic definitions and new directions. Contemporary Educational Psychology, 25, 54–67.

340

J.-S. Lee / International Journal of Educational Research 53 (2012) 330–340

Ryan, R. M., & Deci, E. L. (2000b). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68–78. SAS Institute Inc.. (2008). SAS 9.2. Cary, NC: SAS Institute Inc. Shouse, R. (1996). Academic press and sense of community: Conflict and congruence in American high schools. Research in Sociology of Education and Socialization, 11, 173–202. Sirin, S. R., & Rogers-Sirin, L. (2004). Exploring school engagement of middle-class African American adolescents. Youth & Society, 35, 323–340. Slicker, E. K. (1998). Relationship of parenting style to behavioral adjustment in graduating high school seniors. Journal of Youth and Adolescence, 27, 345–372. Snijders, T., & Bosker, R. (1999). Multilevel analysis: An introduction to basic and advanced multilevel modeling. Thousand Oaks, CA: Sage. Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance. American Psychologist, 52, 613–629. Steinberg, L., Lamborn, S. D., Dornbusch, S. M., & Darling, N. (1992). Impact of parenting practices on adolescent achievement: Authoritative parenting, school involvement, and encouragement to succeed. Child Development, 63, 1266–1281. Tucker, C. M., Zayco, R. A., Herman, K. C., Reinke, W. M., Trujillo, M., Carraway, K., et al. (2002). Teacher and child variables as predictors of academic engagement among low-income African American children. Psychology in the Schools, 39, 477–488. Warm, T. A. (1989). Weighted likelihood estimation of ability in item response theory. Psychometrika, 54, 427–450. Wentzel, K. R. (2002). Are effective teachers like good parents? Teaching styles and student adjustment in early adolescence. Child Development, 73, 287–301. Willms, J. D. (2003). Student engagement at school: A sense of belonging and participation: Results from PISA 2000. Paris: Organisation for Economic Co-operation and Development (OECD).