Introduction: This study examines the evolution of student and physician interest in primary care from medical school matriculation to practice, focusing on student factors that potentiate primary care (PC) practice. Methods: We compiled a dataset of 2,047 Michigan State University College of Human Medicine graduates from 1991 to 2010. PC interest was assessed using the Association of American Medical Colleges (AAMC) matriculating student (MSQ) and graduation (GQ) questionnaires. PC practice was determined using AMA physician Masterfile data. C2 analyses and logistic regression were used to examine factors that predict PC practice. Results: PC interest at matriculation and at graduation were the factors most likely to predict PC practice. After controlling for URM status, gender, and rural origin, the odds of practicing PC among those with a sustained interest in PC (on both the MSQ and GQ) were 100 times higher than those with no interest in PC, on either survey (P < .01). Among those students who developed an interest in PC by graduation, the odds of practicing PC were 60 times higher than noninterested students (P < .01). Finally, among students who were interested in PC at matriculation, but not graduation, the odds of eventually practicing PC were 3.8 times higher than noninterested students (P < .01). Conclusions: Our study suggests that cultivating PC interest at any point during medical school may predict PC practice. Early and sustained interest in primary care was the most substantial predictor of PC practice in our study, highlighting the need for primary care education even before medical school matriculation.

We investigated the relationship between principal instructional leadership and teacher participation in multiple types of professional development in Japan, Singapore, and South Korea. Using the Teaching and Learning International Survey dataset of 2013, we employed two-level logistic regression models to estimate the rigorous effects of principal instructional leadership that were separated out from teacher-level effects. We found that the influence of principal instructional leadership on teachers’ participation in professional development varied across types of learning activities and countries. Our analysis suggests that principal instructional leadership can influence teachers’ participation in mentoring, mentoring, peer observation, and coaching compared to the other types of professional development. Our study builds on and extends research on cross-national characteristics of teacher learning by adding evidence about the relations between principal leadership and teacher professional development in the three Asian countries.

After a brief discussion of the history of IEA’s Trends in International Mathematics and Science Study (TIMSS), and the implications this has had for comparative research into the influence of teachers on student outcomes, this chapter details the key variables that are employed in the remainder of the book. These include several student-level control variables (such as gender, socioeconomic status, and language spoken in the home), teacher-reported measures (experience, education, gender, and time spent on teaching mathematics), and newly developed instruments based on TIMSS responses related to teacher self-efficacy and instructional alignment. All of the analyses in this report make use of TIMSS data, merging several distinct instruments: student assessments, teacher background surveys, and national background surveys. The design of TIMSS has a number of important implications. Relationships between variables should be treated as associations rather than strictly causal, thus a multi-model analytical strategy is needed to test the robustness of statistical results and improve confidence in the reliability of analyses. Relationships were examined across time, across different aggregations of data, and using different statistical procedures, and are discussed in more detail in later chapters. Many features of TIMSS and the topics included in the survey have changed considerably over time and some key variables of interest were not included in earlier versions of the TIMSS, greatly restricting the available sample when analyzing country-level trends (whether as means or regression coefficients). The complex sampling design has important implications for statistical modeling and analysis, and for purposes of analysis, the models in subsequent chapters ignore school effects. Multilevel and classroom-mean models treat each classroom as existing independently, rather than being clustered within schools. Given the acknowledged impact of schools, and the importance of within-school, between-classroom heterogeneity, any conclusions drawn from the analysis should be treated with caution.

Researchers agree that teachers are one of the most important school-based resources in determining students' future academic success and lifetime outcomes, yet have simultaneously had difficulties in defining what teacher characteristics make for an effective teacher. This chapter reviews the large body of literature on measures of teacher effectiveness, underscoring the diversity of methods by which the general construct of "teacher quality" has been explored, including experience, professional knowledge, and opportunity to learn. Each of these concepts comprises a number of different dimensions and methods of operationalizing. Single-country research (and particularly research from the United States) is distinguished from genuinely comparative work. Despite a voluminous research literature on the question of teacher quality, evidence for the impact of teacher characteristics (experience and professional knowledge) on student outcomes remains quite limited. There is a smaller, but more robust set of findings for the effect of teacher support on opportunity to learn. Five measures may be associated with higher student achievement: teacher experience (measured by years of teaching), teacher professional knowledge (measured by education and self-reported preparation to teach mathematics), and teacher provision of opportunity to learn (measured by time on mathematics and content coverage). These factors provide the basis for a comparative cross-country model.

To understand whether there are identifiable trends in teacher quality and instructional metrics over time, it is important to first recognize that measures of teacher effectiveness may vary by education system. Descriptive results for the teacher quality measures related to mathematics drawn from twenty years of the IEA's Trends in International Mathematics and Science Study (TIMSS) data provide a general picture of the variation in such measures; additional consideration is given to student-level variables. Combining information from multiple cycles of TIMSS, with a specific emphasis on teacher variables, greater focus is given to the post-2003 TIMSS data, largely because these later cycles contain additional teacher-related measures (such as preparedness to teach mathematics topics). Changes in education system means may be due to changes in underlying teacher characteristics. The data suggests that, among the pool of countries that participated in all the TIMSS between 2003 and 2015, there have been increases in the international averages for the number of years of experience that teachers possess, their level of educational qualification to teach mathematics, and classroom time spent on mathematics instruction. However, instructional alignment with national standards has remained surprisingly modest (at around 55-60%) and the international average for instructional alignment declined for grade eight over twenty years of TIMSS. The results suggest that, in many countries, teachers (especially grade four teachers) retain a substantial degree of discretion in what to teach. Variability in teacher behaviors related to opportunity to learn suggests it is highly advisable to consider this factor in studies of teacher effectiveness.

The central aim of this report was to investigate what international comparative assessments could reveal about the role of teachers in influencing student outcomes. A series of analyses attempted to identify potential trends in teacher quality and instructional metrics over time and their relationship with student achievement, but found no strong evidence for consistent predictable relationships between commonly-employed indicators of teacher effectiveness and student outcomes (both within and across educational systems). Further, although inequality in student outcomes remains pervasive across education systems, teacher quality is less subject to inequality, and, indeed, many countries have seen a rise in formal preparation to teach mathematics over the twenty years of IEA's Trends in International Mathematics and Science Study (TIMSS). While improving conventional measures of teacher quality may not have a significant impact on educational inequality, there is some evidence that increasing the average amount of time spent on teaching mathematics may reduce inequalities in student achievement. Researchers and policymakers should be extremely cautious about applying the associations identified in one education systems to a very different educational context. The results indicate that teachers with similar experience, credentials, and instructional strategies may regardless produce quite different results, indicating that parents, policymakers, community institutions, and cultural context likely play a powerful role in determining student outcomes.

Inequalities in teacher quality may be examined using several different lenses. To better investigate the relationships between student equity and teacher quality and instructional metrics, educational inequality in student performance across countries and time was explored as (1) variation in student mathematics scores, and (2) differences in socioeconomic status using descriptive, regression, and fixed-effects techniques. Measures of teacher quality (as measured by experience, education, preparedness, time spent on teaching mathematics, and instructional alignment) appeared to have only a limited effect on aggregate or within-classroom variation in student outcomes. The results also suggest that teacher quality may be more equitable at grade four than at grade eight, as measured by differences between higher and lower socioeconomic status classrooms.

Opportunity to learn (measured by time spent on teaching mathematics and mathematics content coverage) potentially acts as a mediator of instructional quality and teacher characteristics. Previous research used a multilevel structural equation model (SEM) to explore the relationship of teacher instructional quality (identified using a latent variable) to student mean outcomes at grade four. Here, this model is extended by incorporating two instructional quality variables related to opportunity to learn: teacher alignment with national curriculum expectations, and time spent on mathematics. This model was applied to grade four and grade eight data from the 2011 cycle of IEA’s Trends in International Mathematics and Science Study (TIMSS). Replicating the analysis at grade eight determined whether the mathematics performance of lower secondary students showed similar relationships with instructional quality as those observed for grade four students. Teacher instructional alignment and time on mathematics was found to have a weak mediating effect on student outcomes.

​This open access book examines the interrelationship of national policy, teacher effectiveness, and student outcomes with a specific emphasis on educational equity. It investigates trends in key teacher characteristics, teacher behaviors, and their relationship to student outcomes.

Internationally, high-quality teachers are often considered to play an essential role in good student outcomes, but there is little research on the relationship between measures of teacher effectiveness and student achievement. Variation across countries in how diverse measures of teacher quality are related to student outcomes has made cross-country transfer of educational ideas difficult, since, without understanding if or why policies work in different national contexts, policymakers and researchers have no solid basis for devising the best strategies for improving teacher effectiveness. IEA's Trends in International Mathematics and Science Study (TIMSS) collects extensive information about teacher background and practices across countries, and therefore can potentially be used to address important questions about the role of teachers in influencing student outcomes. In this study, analysis of the TIMSS data prompts a series of research questions, designed to investigate whether there are identifiable international trends in teacher quality and instructional metrics over time while focusing on issues of educational equity, and examine the association between teacher characteristics and behaviors and mean student outcomes. The chapter concludes with a brief outline of the remainder of the report.

IEA's Trends in International Mathematics and Science Study (TIMSS) can potentially be used to address important questions about the role of teachers in influencing student outcomes. To establish the potential relationships between student achievement and different types of teacher quality and instructional metrics and how stable these relationships may be across time, a variety of different statistical approaches were implemented and compared. The results from a single-level (unclustered) model, a multilevel model, and a classroom-mean regression model were compared, as was the stability of within-country estimates across time. The potential utility of country-level fixed-effects analysis was also explored. The complex sampling design of TIMSS enables robust standard errors to be generated. The analyses suggested only a weak relationship between teacher quality measures and student outcomes across countries and grade levels, and considerable instability in within-country estimates across time, especially at grade eight.