Challenges to drawing generalized causal inferences in educational research : methodological and philosophical considerations
Abstract/Contents
- Abstract
- This work is concerned with external validity in educational research. The first part is an empirical work that examines the levels of two kinds of bias that can figure into the estimate of a program effect, where the intervention is being administered at multiple sites. The first (Bias 1) is due to differences in average performance between sites in the absence of treatment, the second (Bias 2) is due to variation in the treatment effect across sites. Data from the Tennessee STAR multisite trial of the effects of reduced class size is used to investigate both forms of bias. Findings include that both kinds of bias are significantly different from zero, Bias 1 is larger than Bias 2, regression adjustments are more effective at reducing Bias 1 than Bias 2, and in the case of reading but not math outcomes, certain covariates account for both forms of bias. The main conclusion is that both types of bias are large and difficult to reduce to a level where a local school official can feel confident that the impact estimates based on information from other location apply locally. The second work is a critique of the heterogeneity of replication approach to establishing generalized causal inferences. The argument is made that generalized causal inferences stemming from meta-analyses of randomized trials of nominally the same intervention are deficient for several reasons, the main one being that they ignore context. The empirical nature of causal generalizations in the social sciences makes them tentative and more akin to working hypotheses than immutable laws. Alternatives are considered, and the argument is made for an approach that admits a broader range of evidence for seeking appropriate and adequate warrants for drawing generalized causal inferences.
Description
Type of resource | text |
---|---|
Form | electronic; electronic resource; remote |
Extent | 1 online resource. |
Publication date | 2010 |
Issuance | monographic |
Language | English |
Creators/Contributors
Associated with | Jaciw, Andrew P |
---|---|
Associated with | Stanford University, School of Education. |
Primary advisor | Reardon, Sean F |
Thesis advisor | Reardon, Sean F |
Thesis advisor | Bhattacharya, Jay |
Thesis advisor | Phillips, D. C. (Denis Charles), 1938- |
Thesis advisor | Shavelson, Richard J, 1942- |
Advisor | Bhattacharya, Jay |
Advisor | Phillips, D. C. (Denis Charles), 1938- |
Advisor | Shavelson, Richard J, 1942- |
Subjects
Genre | Theses |
---|
Bibliographic information
Statement of responsibility | Andrew Peter Jaciw. |
---|---|
Note | Submitted to the School of Education. |
Thesis | Ph.D. Stanford University 2010 |
Location | electronic resource |
Access conditions
- Copyright
- © 2010 by Andrew P Jaciw
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...