Effects of class size on alternative educational outcomes across disciplines

This is the first study to use self-reported ratings of student learning, instructor recommendations, and course recommendations as the outcome measure to estimate class size effects, doing so across 24 disciplines. Fixed-effects models controlling for heterogeneous courses and instructors reveal that increasing enrollment has negative and significant effects on student satisfaction in Sociology, Political Science, Computer Science and Engineering, and Mechanical and Aerospace Engineering. Educational outcomes in Linguistics, Psychology, Biological Sciences, Chemistry, Math, Physics, Cognitive Sciences, Visual Arts, History, and Philosophy are unaffected by class size. Other disciplines, including Economics, have inconclusive findings. No discipline benefits from increasing enrollment.     

Teaching effectiveness as measured by student ratings and instructor self-evaluation

The present study is designed to evaluate teaching effectiveness based on effective teacher behaviors. Operationally, it is measured by assessing levels of agreement between perceptions of instructors and students on reported employment rates and rated ability for specific instructional behavior attributes. Individually-based and organizationally-based factors are incorporated in the study for a possible contribution to teaching effectiveness variance. Findings indicate a lack of agreement between perceptions on abstract instructional attributes but that agreement exists for attributes of a concrete nature. Analysis suggests that individually-based and organizationally-based factors do not contribute to agreement variance.     

Estimating the impact of relative expected grade on student evaluations of teachers

Grade inflation over the past few decades has been a concern for many universities. Course evaluation scores are known to be positively correlated with students’ expected grades, and this paper tests whether or not there is an incentive for the instructor to “buy” higher evaluation scores by inflating grades. To test this hypothesis, I use unique data from the University of Washington's Office of Educational Assessment that includes a measure of each student's relative expected grade in the course. I find that there is an incentive for instructors to grade leniently after accounting for the potential endogeneity of the relative expected grade variable due to unobserved teacher productivity and unobserved heterogeneity of instructors and departments. Instructor fixed effects account for a significant part of the measured effect of relative expected grade on evaluations, and by not including them, the estimated impact of relative expected grade on evaluations is biased upwards.     

The effect of kindergarten classroom size reduction on second grade student achievement: Evidence from California

Because classroom size reduction (CSR) and standards based testing were implemented at the same time in California during the mid-1990s, it is difficult to isolate the effects of classroom size on outcomes from the effects of curriculum changes. As a result, the main comparison in this paper is very specific – the effect of reduced sized kindergarten classrooms on test performance in second grade. The main result of the paper is that the effect of classroom size reduction is small, especially when compared to the other determinants of student achievement at the second grade level. There is a very small effect of classroom size reduction on student achievement in reading and math, and no effect in language and spelling. Of the two offsetting effects of CSR controlling for spending – the direct effect of smaller classes and the indirect effect on teacher quality, the class-size effect is larger. Over time, the negative effect on teacher quality is reduced as new teachers gain experience and credentials.     

Twice considered: Charter schools and student achievement in Utah

A relatively small state, Utah presents an interesting case to study charter schools given its friendly policy environment and its significant growth in charter school enrollment. Based on longitudinal student-level data from 2004 to 2009, this paper utilizes two approaches to evaluate the Utah charter school effectiveness: (a) hierarchical linear growth models with matched sample, and (b) general methods of moments with student-fixed effects regressions. Both methods yield consistent results that charter schools on average perform slightly worse as compared to traditional public schools, a result that is primarily affected by the low effectiveness and high student mobility of newly opened charter schools. Interestingly, when charter schools gain more experience they become as effective as traditional public schools, and in some cases more effective than traditional public schools. This research has implications for local and state charter school policies, particularly policies that avoid “start-up” costs associated with new charter schools.     

Students' ratings of instructors revisited: Interactions among class and instructor variables

Previous studies concerning students' ratings of instruction have traditionally used the class as the unit of analysis and the ratings have been analyzed in one of two ways: (1) regression analysis, wherein the amount of variability in instructor ratings can be attributed to a set of variables; or (2) analysis of variance, wherein the effect of some selected independent variable on instructor ratings is measured. While both approaches have provided valuable information about the evaluation of instruction, little attention has been given to the interactions among the variables selected. In order to determine how situational variables influence the student at the time an evaluation is performed, the present study used the individual student as the unit of analysis and focused principally on the interactions between three variables related to the class (type, level, and size) and three variables related to the instructor (reputation, rank, and sex). The data were analyzed through 15 two-way factorial analyses of variance, with 23 main effects and 12 interactions reaching significance. The implications of these findings are discussed in terms of their effect on the student rating process.     

Heterogeneity,comparative advantage,and return to education: The case of Taiwan

By considering heterogeneity in abilities and self-selection in educational choice, this paper adopts the heterogeneous human capital model to estimate rate of return to university education using data from the 1990 and 2000 Taiwan's Manpower Utilization Surveys. The Taiwan empirical study shows that significant heterogeneous return to education does exist, and that the educational choice was made according to the principle of comparative advantage. The estimated rates of return for attaining university were 19% and 15%, much higher than the average rate of return of 11.55 and 6.6%, for 1990 and 2000, respectively. The declining trend of return to university education may have been caused by the rapid expansion of the number of colleges and universities and the increasing supply of college graduates in the 1990s.     

Education inputs,student performance and school finance reform in Michigan

This paper estimates the impact of the Michigan school finance reform, Proposal A, on education inputs and test scores. Using a difference-in-difference estimation strategy, I find that school districts in Michigan used the increase in educational spending generated through Proposal A to increase teacher salaries and reduce class size to a smaller extent. Then, using the foundation allowance created by Proposal A as an instrument, I estimate the causal effect of increased spending on 4th and 7th grade math scores for two test measures – a scaled score and a percent satisfactory measure – and find positive effects of increased spending on 4th grade test scores. A 60% increase in spending increases the percent satisfactory score by one standard deviation. The positive impact of expenditures on test performance seems largely due to higher teacher salaries.     

The effect of instructor race and gender on student persistence in STEM fields

The objective of this study is to determine if minority and female students are more likely to persist in a science, technology, engineering, and math (STEM) major when they enroll in classes taught by instructors of their own race or gender. Using data from public 4-year universities in the state of Ohio, I analyze first semester STEM courses to see if the race or gender of the instructor effects persistence of initial STEM majors in a STEM field after the first semester and first year. Results indicate that black students are more likely to persist in a STEM major if they have a STEM course taught by a black instructor. Similar to previous findings, female students are less likely to persist when more of their STEM courses are taught by female instructors.     

Income contingent student loans for Thailand: Alternatives compared

There is significant irresolution in many countries concerning the design of student loan schemes. In no country recently has there been more uncertainty as to the form that loans should take than Thailand. The Student Loans Fund (SLF), a conventional approach to financing, was introduced in 1996, discontinued at the end of 2005, and re-introduced in 2007. In its place an income contingent loan (ICL) was implemented for one year only, 2006. As part of this debate we contribute to an understanding of the repayment burdens associated with the SLF in Chapman, Lounkaew, Polsiri, Sarachitti and Sitthipongpanich (in this issue).     

The cost structure of higher education in further education colleges in England

This paper examines the cost of the increased provision of higher education courses within further education colleges in England. We believe this to be the first attempt to fit a cost function specifically to the further education sector. Cost functions for a sample of 96 colleges over a 2-year period, from 2000 to 2002, are fitted using a panel data methodology as well as stochastic frontier analysis. We compare and contrast our findings with a sample of 959 US colleges. Our findings indicate that most further education colleges are able to benefit from economies of scale. Results from both methodologies suggest the presence of product-specific economies of scale, substantial ray economies of scale and indicate that higher education classroom-based courses, such as business studies, as well as vocational courses display substantial economies of scope.     

On the relationship between student tastes and motivations,higher education decisions,and annual earnings

We examine the degree to which measures of student tastes and motivations are associated with the outcomes of three important higher education decisions and subsequent annual earnings. Within a sample of nearly 9000 students from the Baccalaureate and Beyond, we find that these measures are correlated with college type, college major, and highest postgraduate degree earned in generally predictable ways. For instance, students claiming it important to be well-off financially are significantly more likely to attend top public universities and major in Business or Engineering while students claiming it important to live near family are significantly less likely to attend top quality private institutions and significantly more likely to major in education.     

Social capital and student learning: Empirical results from Latin American primary schools

This paper presents an empirical analysis of the relationship between social capital and student math and language achievement and the probability of promotion, using data gathered from fourth grade classrooms in public schools in four Latin American cities. The results suggest that social capital among teachers in a school, between teacher and students, and among the students in a classroom contribute significantly to learning achievement and the probability of promotion. Furthermore, social capital between the students matters at least as much as the teacher's social capital. Children learn from each other and the networks allowing that to happen can be very important. The current pressure on teachers to achieve results on reading and math scores has tended to push teachers to “teach to the test”. Ironically, this study's results indicate that spending time in creating social capital within the classroom environment is associated with higher language and math test scores.     

The spatial geography of teacher labor markets: Evidence from a developing country

An unequal distribution of teacher quality is a problem underlying the unequal distribution of educational outcomes in developing countries. However, we know little about how the labor market produces such a distribution. Using data from two regions in Peru, we investigate whether there is a national teacher market or smaller regional markets. We estimate discrete-choice multinomial models to identify variables (including teacher characteristics, institutional features and geographical factors) associated with the location of teachers in the first jobs of their careers. Results indicate that teacher markets are regional in scope. Being born in a certain province (sub-area of a region) substantially increases the probability of having a first teaching position in that same province. We also find evidence that the geographic mobility of teachers is quite limited. Results suggest that policies to strengthen teacher educational systems and reduce inequities should focus on the regional level.     

Meta-analysis of faculty's teaching effectiveness: Student evaluation of teaching ratings and student learning are not related

Student evaluation of teaching (SET) ratings are used to evaluate faculty's teaching effectiveness based on a widespread belief that students learn more from highly rated professors. The key evidence cited in support of this belief are meta-analyses of multisection studies showing small-to-moderate correlations between SET ratings and student achievement (e.g., Cohen, 1980, Cohen, 1981; Feldman, 1989). We re-analyzed previously published meta-analyses of the multisection studies and found that their findings were an artifact of small sample sized studies and publication bias. Whereas the small sample sized studies showed large and moderate correlation, the large sample sized studies showed no or only minimal correlation between SET ratings and learning. Our up-to-date meta-analysis of all multisection studies revealed no significant correlations between the SET ratings and learning. These findings suggest that institutions focused on student learning and career success may want to abandon SET ratings as a measure of faculty's teaching effectiveness.     

Validity research on student evaluations of teaching effectiveness: Individual student observations versus class mean observations

Among the several substantive methodological issues in statistical research on possible relationships between student evaluations of teaching effectiveness and student achievement is the issue of whether it is preferable to use individual students or class means as observations. The problem is illuminated in this paper through Monte Carlo simulation experiments employing a general model which may be used to represent validity or invalidity of student evaluations of teaching through appropriate specification of the numerical parameters of the model. The results suggest the superiority of class means, because statistics based on class mean observations do better at manifesting the true condition under both cases: validity and invalidity.     

Expenditures and postsecondary graduation: An investigation using individual-level data from the state of Ohio

Using detailed individual-level data from public universities in the state of Ohio, I estimate the effect of various institutional expenditures on the probability of graduating from college. Using a competing risks regression framework, I find differential impacts of expenditure categories across student characteristics. I estimate that student service expenditures have a larger impact on students with low SAT/ACT scores, while instructional expenditures are more important for high test score students and those majoring in scientific/quantitative fields. The individual-level nature of these data allows me to address measurement error and endogeneity concerns the previous literature has been unable to deal with.