Measuring teacher effectiveness for equity: value-added model of teachers’ distributive effects on classroom achievement gaps

ABSTRACT

Extending conventional value-added models (VAM) of teacher effects that focus on the estimation of average effects for excellence, this study examines “distributive” effects for equity. Using multilevel regression models, the study addresses two research questions to inform educational policy for teacher improvement and accountability. First, do more effective teachers contribute not only to improving student achievement overall (excellence) but also narrowing achievement gaps among student groups (equity)? The results show highly mixed relationships between average effect and distributive effects; effective teachers on average narrowed the achievement gaps among academic groups but not the gaps among racial and socioeconomic groups in their classrooms. Second, what attributes and practices of teachers are associated with desirable teacher effects for both excellence and equity? The conventional measures of teacher characteristics and practices help account for the average teacher effects but not the distributive effects. Implications are discussed for improving the measures of teacher effectiveness.

KEYWORDS:

• Teacher effectiveness
• accountability
• equity
• value-added model
• distributive effects

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Nine percent in Goldhaber et al. (1999); 13% (mathematics) and 7% (reading) in Nye et al. (2004). The latter was evaluated from Tennessee STAR data.

2 In measuring students’ poverty status, we used the free or reduced-priced lunch (FRPL) information that serves as a proxy measure for poverty. The educational background and occupational status of students’ parents should have been taken into account to improve the quality of the variable Poverty.

3 Since the teacher distributive effect with respect to the overall dispersion (TDE-R) model (see Equation 7(7) $(TDE-R{)}_j=\ln [Var\left(\widehat{r_{ij}}\right)]=\mathrm{l}\mathrm{n}\left[\frac{{\sum }_{i=1}^{n_j}{\left(\widehat{r_{ij}}-\overline{\widehat{r_{.j}}}\right)}^2}{n_j-1}\right]$(7) ) was used in this study to produce the overall value-added dispersion metric (without using any predictors except the prior test scores), the analysis here focused on the base model in its interpretation of HLM results.

Additional information

Funding

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea under Grant NRF-2019S1A5C2A04081101.

Notes on contributors

Taeyoung Kim

Taeyoung Kim is an assistant professor at Dong Eui University in the Dignitas Institute for Liberal Education, Busan, South Korea. His research interests include quantitative research method, educational effectiveness (in regard to liberal education), and statistical literacy.

Jaekyung Lee

Jaekyung Lee is a professor in the Graduate School of Education at the University at Buffalo, SUNY, USA. He specializes in educational policy research and evaluation.