Ed230C Experiment #2 - Completely Randomized Factorial Design

Using the data set NAEP ANOVA or a dataset of your own choosing (with approval of the instructor). The NAEP ANOVA dataset is also available as a stata file (naepanv.dta) and can be accessed using the Stata command:

use http://www.gseis.ucla.edu/courses/data/naepanv

If you use the NAEP ANOVA, use TOTAL as the dependent variable. Explore the effects of the different levels of MTYPE and LUNCH using a completely randomized factorial design.

If you do any manual computations include them on your computer printout. Be sure to include the following:

1. Draw a schematic representation of the design.

2. Present the linear model, explaining all terms.

3. Present the statistical hypotheses in an appropriate form.

4. State the assumptions underlying this design. For each assumption, present logical and/or empirical evidence which bears upon whether the assumption is tennable. Present descriptive statistics and statistical tests when appropriate. Be sure to show your work.

5. Report the results of the ANOVA. Be sure to include the ANOVA summary table, measures of strength of association.

6. Present the expected mean squares, explaining the terms, and showing correctly formed F-ratios.

7. Perform and interpret any post-hoc comparisons that are necessary.

8. State your conclusions based upon the results of this analysis.

UCLA Department of Education