Effect of teacher support on students’ math attitudes: Measurement invariance and moderation of students' background characteristics

https://doi.org/10.1016/j.cedpsych.2021.101988Get rights and content

Highlights

  • A split-sample CFA provided rigorously tested validity evidence for measures used.

  • Measures were metric invariant for sex, minority status, and parent education.

  • Teacher support had an effect on students' math attitudes.

  • Investigation of moderation effect in light of measurement invariance.

  • Benefit of teacher support on math attitudes does not appear to differ based on student background.

Abstract

We examine the effect of students' perceptions of teacher support on attitudes towards math and whether the association varies between students from historically underrepresented groups in STEM. Participants included high school students enrolled in an AP Statistics course (N = 585, Mage = 16.75 years, SDage = 0.88). Measurement invariance conducted on the basis of student background characteristics (i.e., biological sex, underrepresented race/ethnicity status, and parental educational attainment) provided evidence of metric invariance or greater. Standardized regression coefficients suggested potential differences, such that for students who were female, from an underrepresented racial/ethnicity group in STEM, or had parents with lower levels of education, the effect of teacher support appeared not as strong compared to their counterparts. However, scaled chi-square difference tests comparing nested latent path models did not suggest a moderation effect of these three characteristics. These findings have implications for understanding barriers many students face in receiving the benefits of teacher support.

Introduction

Teachers undoubtedly influence students' learning, by guiding the acquisition of knowledge and skills, as well as providing the emotional-social support for students to feel capable of learning. The quality of students' interactions with teachers may influence their attitudes towards certain subject matter, with past findings indicating that positive teacher-student interactions can promote more favorable math attitudes among students (Rice, Barth, Guadagno, Smith, & McCallum, 2013). However, whether the effect of teacher support on students’ math attitudes is consistent across student subgroups and how differences in certain background characteristics may influence the strength of this association has yet to be considered. In the present study, there are two specific aims. As a primary aim, we test for an association between teacher support and math attitudes, and whether the strength of the effect of teacher support on math attitudes differs between potentially minoritized individuals within science, technology, engineering, and mathematics (STEM) disciplines given a sample of high school students enrolled in an Advanced Placement (AP) course. We focus particularly on group differences based on biological sex, race/ethnicity, and parental educational attainment. In order to achieve this aim, we first attempted to establish measurement invariance of the self-report measures across subgroups by conducting measurement invariance. We thus serve an applied and methodological aim towards understanding the measurement of teacher support and students’ attitudes towards math.

Interactions with teachers can positively affect a student’s attitudes towards math (Liu et al., 2018). Neale (1969) describes math attitudes as a general liking of math, a tendency to engage in math activities, a belief that one is good at math and a belief that math is useful. Math attitudes therefore encompass one’s enjoyment of math, views of their own ability to solve math problems, and the values placed on learning math. In line with this perspective, past research has found that students' attitudes towards math is associated with their achievement in learning math (Chen et al., 2018, Lipnevich et al., 2015), and their intentions to continue an education where knowledge of math is essential (Ing and Nylund-Gibson, 2013, Nugent et al., 2015, Wang, 2013).

There is empirical evidence that the quality of academic and interpersonal support provided by teachers influences students' attitudes and behaviors towards learning, as well as academic performance. Teacher support refers to the academic and affective support teachers provide students (Patrick, Ryan, & Kaplan, 2007) by offering students a balance between structure and autonomy to optimize their learning while also demonstrating caring for the individual needs of students (Klem & Connell, 2004). Though academic and affective support appear to be distinct, past research has found them to be so highly intercorrelated they can be treated as part of a unitary construct among middle (Wentzel, 1997) and high school students (Yu & Singh, 2018).

Past research has directly connected the quality of teacher support with students’ attitudes towards math (De Lourdes Mata et al., 2012, Rice et al., 2013). Though past findings largely lend evidence of a correlation, there is compelling evidence that teacher support directly influences the positive development of students’ attitudes towards math, particularly students’ self-efficacy beliefs (Schunk & Pajares, 2002; Siegle & McCoach, 2007). The association between teacher support and math attitudes appears robust with speculation that it is a causal association (Allen et al., 2006, Dorman and Fraser, 2009, Garrison et al., 2010); however, it may be difficult to draw general conclusions given that most past research has drawn from experiences of elementary and middle school students. Further, there is some evidence that the effect of perceived teacher support on math attitudes may vary as a function of students' age. For example, De Lourdes Mata et al. (2012) found that students in high school reported feeling less supported by teachers and having less positive attitudes towards math compared to students in lower grade levels. Rice et al. (2013) also found that perceived support was highest in the youngest students and in college level students compared to students of all other grade levels. Thus, there is a gap in understanding the effect of teacher support on students' math attitudes among adolescents. In particular, there have been few efforts to examine the effect of teacher support among adolescents enrolled in AP coursework.

The association between teacher support and students’ math attitudes may systematically differ between students due to variation in students’ past educational experiences and socialization. First, students’ experiences are likely to differ slightly as a function of their demographic background, and these past experiences shape their attitudes toward math. Thus, the association between teacher support and math attitudes may be hampered or enhanced due to past students’ experiences and socialization (McGraw, Lubienski, & Strutchens, 2006). Second, students from different demographic backgrounds may hold different expectations about what types of behaviors are supportive, and thus may perceive teachers’ behaviors as nurturing to different extents (Fouad et al., 2010). Third, teachers may actually provide different levels of support to students from different demographic groups (Riegle-Crumb & Humphries, 2012). Fourth, other contextual factors or a dynamic interaction among multiple factors could influence this association (Garrison et al., 2010). Understanding how teacher support may differ based on student demographics and how it may also differentially influence high school students’ attitudes towards math subjects has implications for understanding persistence among underrepresented groups in math and other STEM disciplines. In this paper, we focus on groups on the basis of three critical background factors: biological sex, underrepresented racial/ethnic status in STEM disciplines, and parental educational attainment. To the best of our knowledge, very few past studies have examined these factors as possible moderators of the association between teacher support and math attitudes, but there is evidence that these factors explain some variation in both constructs.

There has been considerable past interest in comparing math achievement and math attitudes between male and female students. Though there is inconsistent evidence of a discrepancy in math achievement based on biological sex, there is compelling evidence suggesting robust differences in math attitudes. Else-Quest, Mineo, and Higgins (2013) examined math attitudes and math achievement based on sex and found that males and females earned similar end-of-year grades in their math and science courses. However, male students reported higher attitudes towards math, particularly higher self-confidence, than female students. De Lourdes Mata et al. (2012) also found that among female students, positive attitudes towards math continually declined as they advanced through school, while male students’ math attitudes declined earlier in the course of their education and remained stable throughout secondary school. A student’s biological sex may have a moderating effect on the association between perceived teacher support and their math attitudes. Among a sample of children and adolescents spanning elementary (grade 5), middle (grade 8), and high school, as well as early college, Rice et al. (2013) found a significant difference between male and female students in the association between affective teacher support and students’ math attitudes. In addition, teachers may unknowingly hold a biased view that math is a subject in which males are more likely to excel than females (Wang & Degol, 2013).

The effect of teacher support on math attitudes may differ for students from underrepresented racial/ethnic groups relative to peers. In the context of STEM education, underrepresented minority status is typically defined as belonging to a group whose number is substantially below the comparable figure for scientists and engineers who are not considered a member of a racial or ethnic minority group (U.S. Code, 2011). Based on this definition, individuals with underrepresented minority (URM) status in STEM are those who identify as Black/African-American, Hispanic/Latinx, American Indian, Hawaiian or Alaskan Native (NSF, 2019). Individuals with URM status face multiple obstacles in pursuing STEM education. Some research has found URM students may have a diminished belief of self-efficacy and increased anxiety towards math (Ahmed, 2018). While socialization and past experiences are likely contributing factors, situational factors (e.g., stereotype threat) may also undermine academic attitudes and achievement (Nguyen & Ryan, 2008). Teachers’ perceptions of a student’s mathematics ability may also differ as a function of race/ethnicity (Copur-Gencturk et al., 2019, Marks and Garcia Coll, 2018, Seyfried, 1998). Potential racial/ethnic biases in teachers’ perceptions of students’ math abilities likely affect the support provided to students, and thus also students’ attitudes towards the subject matter. In mathematics, teacher shortages coupled with the already low proportion of math teachers with backgrounds from minoritized groups (Taie & Goldring, 2017) implies that many minoritized students will not match the race/ethnicity of their teacher and thus may feel alienated in school settings.

The connection between race/ethnicity and socioeconomic status (SES) further complicates associations between teacher support and students’ academic attitudes and achievement (Riegle-Crumb & Grodsky, 2010). For example, students from URM backgrounds are more likely to attend schools serving students from predominantly low-income households (Crosnoe, 2009) and which also have higher turn-over among teaching staff and a teaching staff with fewer years of teaching experience on average (Darling-Hammond, 2001). As such, the association between teacher support and math attitudes may differ on the basis of URM status for any of the previously mentioned reasons.

Additional factors associated with the students’ home context may also influence the association between teacher support and math attitudes. Parental educational attainment is often used as an indicator of SES (Sirin, 2005); however, it also provides a proxy for the support students receive from caregivers with respect to their own educational attainment (Faas, Benson, & Kaestle, 2013). Students who have parents with a higher level of educational attainment tend to enroll in a greater number of college-preparatory (Crosnoe & Muller, 2014) and STEM (Svoboda, Rozek, Hyde, Harackiewicz, & Destin, 2016) courses in high school. Higher parental educational attainment has also been found to positively correlate with students’ achievement in and attitudes toward math (Eccles & Jacobs, 1986), particularly with respect to increased self-confidence (Schreiber, 2002) and diminished math anxiety (Soni & Kumari, 2017). Teachers’ own perceptions of the degree of parents’ involvement in their child’s education has been found to positively correlate with parents’ educational attainment, even while parents’ own self-reported involvement did not (Bakker, Denessen, & Brus-Laeven, 2007). Thus, if there is an inclination to view more educated parents as more involved in their child’s education, there may also be a tendency to offer different levels of support and hold different expectations for certain students. There is also evidence that perceived teacher support varies by socioeconomic background. For example, Barile et al. (2012) found students’ socioeconomic background was positively associated with student–teacher relationships, even when controlling for other factors, such as biological sex and race/ethnicity.

Several theories may explain how teacher support is differentially associated with students’ attitudes towards math. Distancing theory (see Sigel, 2002) has been proposed as one framework to understand how such differences in background characteristics influence teacher and student relationships (Rasheed, Brown, Doyle, & Jennings, 2019). Following the logic of distancing theory, cultural misunderstanding or unfamiliarity precipitates teachers’ reliance on simplistic and stereotypical perceptions of their students (Irvine, 1990, Saft and Pianta, 2001). Teacher biases about student ability have even been speculated as a potential factor that perpetuates self-fulfillment of underachievement, commonly referred to as the Pygmalion effect (Rosenthal & Jacobsen, 1968). Students from minoritized groups who experience stereotype threat activation towards certain subjects may be more likely to have negative attitudes towards that subject (Spencer, Steele, & Quinn, 1999). Stereotype threat refers to a state in which a negative stereotype about one's group is particularly salient within a given context (Steele & Aronson, 1995). Therefore, teacher bias may have a compounding negative impact on students’ attitudes towards the subject and perceptions of their ability in math (Friedrich, Flunger, Nagengast, Jonkmann, & Trautwein, 2015).

For instructors in STEM disciplines, such negative biases, particularly towards already minoritized groups, may be readily perceived by students, thus further barring opportunities for effective learning within the course. Understanding whether, and to what extent, teacher support differentially influences students’ attitudes towards math may thus have implications for improving access to instructional equity for otherwise minoritized students.

Section snippets

Current investigation

The current investigation was motivated by both substantive and methodological aims. The substantive goal of this investigation is to examine the relationship between teacher support and math attitudes. We anticipated that more positive levels of perceived teacher support will be correlated with positive math attitudes. Further, we predicted that the demographic characteristics of biological sex, URM status, and parental education levels will influence the strength of the relationship between

Study 1: Measurement models and measurement invariance

We attempted to establish the measurement model and examine measurement invariance of the latent constructs of teacher support and math attitudes. First, we determined whether the unidimensional model for the teacher support scale fit the data well. This model was based on a preliminary analysis of the scale. Second, we determined whether the factor model for the math attitudes suggested by Lim and Chapman (2013) fit the data. Once the factor models for both teacher support and math attitudes

Study 2: Moderation analyses

After establishing measurement models and at least metric invariance for teacher support and math attitudes, we wanted to examine the effects of different demographic characteristics as moderators on the association between teacher support and math attitudes. We ran separate moderation analyses on the structural equation model to answer three research questions, one for each demographic variable: biological sex, URM status, and parental education level.

Discussion

The purpose of this study was to examine the effect of students' perceptions of teacher support on attitudes towards math, and to further determine whether the association varied between students based on certain key background characteristics. To pursue this objective, we first sought to establish some evidence that the measures of teacher support and math attitudes were valid and invariant. Given that AP courses are similar to each other across schools, any potential curriculum effects are

Limitations

Despite our efforts to be rigorous in establishing validity evidence before conducting the moderation analyses, there are several potential limitations of the present study. The present investigation examined the associations between teacher support and math attitudes specifically within a sample of students enrolled in AP Statistics. There are benefits of examining learning within a specific population such as this, particularly given that the AP test framework (CollegeBoard, 2020) is likely

Conclusions

Past research has found that greater support from math teachers is associated with more positive attitudes towards math (Rice et al., 2013), which are in turn associated with math achievement (Lipnevich et al., 2015), and even distal outcomes such as persistence in STEM (Ing and Nylund-Gibson, 2013, Nugent et al., 2015, Wang, 2013). We sought to understand this association, and particularly what background characteristics (i.e., biological sex, underrepresented race/ethnicity status, and

Acknowledgements

We would like to thank the graduate and undergraduate research assistants in the Learning Analytics and Measurement in Behavioral Sciences (LAMBS) Lab at the University of Notre Dame for their contributions to the broader discussion of the topic, and the high school statistics teachers and students who contributed to this project. A portion of this work presented at the Association for Psychological Science 2020 Annual Meeting held virtually.

Funding

This work was supported by the National Science Foundation CAREER award (Grant #DRL-1350787) to Dr. Ying Cheng.

References (84)

  • D. Witteveen et al.

    Reconsidering the ‘meritocratic power of a college degree’

    Research in Social Stratification and Mobility

    (2020)
  • M. Allen et al.

    The role of teacher immediacy as a motivational factor in student learning: Using meta-analysis to test a causal model

    Communication Education

    (2006)
  • D. Baker et al.

    Mother’s strategies for children’s school achievement: Managing the transition to high school

    Sociology of Education

    (1986)
  • J. Bakker et al.

    Socio-economic background, parental involvement and teacher perceptions of these in relation to pupil achievement

    Educational studies

    (2007)
  • J.P. Barile et al.

    Teacher-student relationship climate and school outcomes: Implications for educational policy initiatives

    Journal of Youth and Adolescence

    (2012)
  • A. Beauducel et al.

    On the performance of maximum likelihood versus means and variance adjusted weighted least squares estimation in CFA

    Structural Equation Modeling: A Multidisciplinary Journal

    (2006)
  • P.E. Brosseau-Liard et al.

    Adjusting incremental fit indices for nonnormality

    Multivariate Behavioral Research

    (2014)
  • P.E. Brosseau-Liard et al.

    An investigation of the sample performance of two nonnormality corrections for RMSEA

    Multivariate Behavioral Research

    (2012)
  • G.W. Cheung et al.

    Evaluating goodness-of-fit indexes for testing measurement invariance

    Structural Equation Modeling: A Multidisciplinary Journal

    (2002)
  • L. Chen et al.

    Positive attitude toward math supports early academic success: Behavioral evidence and neurocognitive mechanisms

    Psychological Science

    (2018)
  • CollegeBoard. (2019). National and State Summary Reports for 2019. Retrieved from...
  • CollegeBoard. (2020). AP Central: AP Statistics. Retrieved from...
  • Y. Copur-Gencturk et al.

    Teachers’ bias against the mathematical ability of female, black, and Hispanic students

    Educational Researcher

    (2019)
  • R. Crosnoe

    Low-income students and the socioeconomic composition of public high schools

    American Sociological Review

    (2009)
  • R. Crosnoe et al.

    Family socioeconomic status, peers, and the path to college

    Social Problems

    (2014)
  • L. Darling-Hammond

    The challenge of staffing our schools

    Educational Leadership

    (2001)
  • M. De Lourdes Mata et al.

    Attitudes towards mathematics: Effects of individual, motivational, and social support factors

    Child Development Research

    (2012)
  • C. DiStefano et al.

    A comparison of diagonal weighted least squares robust estimation techniques for ordinal data

    Structural Equation Modeling: A Multidisciplinary Journal

    (2014)
  • C.V. Dolan

    Factor analysis of variables with 2, 3, 5 and 7 response categories: A comparison of categorical variable estimators using simulated data

    British Journal of Mathematical and Statistical Psychology

    (1994)
  • J.P. Dorman et al.

    Psychosocial environment and affective outcomes in technology-rich classrooms: Testing a causal model

    Social Psychology of Education

    (2009)
  • J.S. Eccles et al.

    Social forces shape math attitudes and performance

    Signs: Journal of Women in Culture and Society

    (1986)
  • N.M. Else-Quest et al.

    Math and science attitudes and achievement at the intersection of gender and ethnicity

    Psychology of Women Quarterly

    (2013)
  • C. Faas et al.

    Parent resources during adolescence: Effects on education and careers in young adulthood

    Journal of Youth Studies

    (2013)
  • X. Fan et al.

    Effects of potential confounding factors on fit indices and parameter estimates for true and misspecified SEM models

    Educational and Psychological Measurement

    (1998)
  • X. Fan et al.

    Effects of sample size, estimation methods, and model specification on structural equation modeling fit indexes

    Structural Equation Modeling: A Multidisciplinary Journal

    (1999)
  • J. Grossman et al.

    Perceived gender and racial/ethnic barriers to STEM success

    Urban Education

    (2014)
  • N. Guenole et al.

    The consequences of ignoring measurement invariance for path coefficients in structural equation models

    Frontiers in Psychology

    (2014)
  • Y.Y. Hsiao et al.

    The impact of partial measurement invariance on testing moderation for single and multi-level data

    Frontiers in Psychology

    (2018)
  • L.T. Hu et al.

    Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives

    Structural Equation Modeling: A Multidisciplinary Journal

    (1999)
  • M. Ing et al.

    Linking early science and mathematics attitudes to long-term science, technology, engineering, and mathematics career attainment: Latent class analysis with proximal and distal outcomes

    Educational Research and Evaluation

    (2013)
  • J.J. Irvine

    88 Post Road West

    Black students and school failure. Policies, practices, and prescriptions

    (1990)
  • E. Jung et al.

    Comparisons of three empirical methods for partial factorial invariance: Forward, backward, and factor-ratio tests

    Structural Equation Modeling: A Multidisciplinary Journal

    (2016)
  • Cited by (0)

    View full text