Testing the Difference Between Two Alpha Coefficients With Small Samples of Subjects and Raters.
A study was conducted to test the difference between two alpha coefficients with small samples of subjects and raters. Data were obtained from a simulated repeated application of Feldt's test using six population matrices of part-core variances and covariances . Findings revealed that the test wor...
| Publicado en: | Educational & Psychological Measurement Vol. 66; no. 4; pp. 589 - 601 |
|---|---|
| Autor principal: | |
| Formato: | Artículo |
| Publicado: |
Sage Publications Inc.
August 2006
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| Sumario: | A study was conducted to test the difference between two alpha coefficients with small samples of subjects and raters. Data were obtained from a simulated repeated application of Feldt's test using six population matrices of part-core variances and covariances . Findings revealed that the test worked well in its control of Type 1 error in such situations. However, findings indicated that power is limited with small samples of subjects, and sampling errors in the coefficients tend to be large. Findings are discussed in detail. |
|---|