|
Activity Number:
|
170
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #304672 |
|
Title:
|
Difference in Causal Effects of Class Size on Academic Achievement Between Black and Other Students: Multivariate Instrumental Variable Estimators with Tennessee Class Size Data MAR
|
|
Author(s):
|
Yongyun Shin*+
|
|
Companies:
|
Virginia Commonwealth University
|
|
Address:
|
Department of Biostatistics, Room 3048, Richmond, VA, 23298-0032,
|
|
Keywords:
|
Causal Effect ; Class Size ; Missing At Random ; Instrumental Variable ; Maximum Likelihood ; Simultaneous Equation Model
|
|
Abstract:
|
Does reduced class size cause higher academic achievement for black and other students in reading, math, listening, and word recognition skills? Are the causal effects different across the two ethnic groups? Does the magnitude of such differences vary across schools? The causal questions are addressed by analysis of data from Tennessee's Student/Teacher Achievement Ratio study where students are randomly assigned to a small or regular class type. The data are analyzed by a three-level simultaneous equation model (SM) via maximum likelihood. The class type as an instrumental variable causes class size to vary which, by hypothesis, influences academic achievement overall and moderates a disparity in academic achievement between black and other students. The SM handles missing data with a general missing pattern at all levels. Policy implication on the ethnic achievement gap is discussed.
|
