MANCOVA and MANOVA Discussion > ANCOVA Analysis for Nested Design
Uzma,
I'm afraid the level of assistance that you are seeking is not realistic to be fully provided on a discussion forum. If you are looking for in-depth help, consider contacting me for consulting services (link below). The discussion forum is intended for troubleshooting and general questions. It isn't a good venue to provide the type of step-by-step help you are seeking.
http://www.statsmakemecry.com/stats-make-me-cry-project-info/
To answer your general questions, I think you would be fine with doing a general linear model or multiple regression, though you could use a nested model because of the different sections. With that said, I don't think you have enough different teachers/sections to really analyze between-section effects. I would probably just use a General Linear Model (which would allow you to examine interaction effects also), unless I see reason to believe (in the data) that there is substantial variability between teachers/sections. I hope that helps!
Jeremy Taylor
In an experimental study, 210 students were randomly assigned into 6 sections, out of which 2 sections were treated as experimental group 1 subjected to instructional method A, other 2 sections were treated as experimental group 2 subjected to instructional method B & the remaining 2 sections were treated as control group subjected to instructional method C. Two teachers were also randomly assigned to teach these 3 groups so as to minimize teacher differences. Each section of each group was taught separately.
This study was conducted to study the effects of instructional methods on students' achievement and attitude.
Independent Variables:
1. Instructional method (3 levels)
2. Gender (2 levels)
Dependent Variables
1. Achievement
2. Attitude
Query 1: Is this a nested design?
Query 2: How to do ANCOVA analysis for nested design in SPSS?
Query 3: Can interaction effect (Instructional method x Gender) on Achievement and Attitude be tested for nested designs in SPSS? I read in some book that it can't be done.
Moreover, the following variables are to be used as predictors in a regression model to predict students’ learning outcomes (including achievement and attitude):
1. pretest score
2. instructional methods
3. PLE score (Preferred Learning Environment score)
4. instructional methods × PLE
How to do this in SPSS? How to enter this interaction term: instructional methods × PLE in regressional model in SPSS?