I am studying the satisfaction of students being taught through studio classes. I have used the standard scale developed by Biner, called the Telecourse Evaluation Questionnaire. This comprises of three constructs - Instructor, Technology and Course management.
I collected the data using google survey and received 141 valid responses. The questions are all likert.
One of the objectives of the research is to measure the students' satisfaction on each construct and overall. The question in this is to find if there is a significant difference in students' satisfaction in the different constructs.
The second objective is to check if the Program type (Business - Core, Business- Elective or Non-business) or Number of courses ( 1, 1-2 and More than 2) affect the satisfaction of students (construct-wise and overall). Which test should be applied?
For the comparison of the answers on the three constructs, you could use a set of paired-comparison t-Tests.
For your second objective, you could use a set of ANOVAs, and possibly a MANOVA to include all three dependent variables in one analysis.
For the comparison of the answers on the three constructs, you could use a set of paired-comparison t-Tests.
For your second objective, you could use a set of ANOVAs, and possibly a MANOVA to include all three dependent variables in one analysis.
You are studying students’ satisfaction being taught through studio classes. You still don’t have independent variable!! Are you going to compare “studio classes” with other teaching methods? I did not see this clear in your question! If yes, and you have 2 teaching methods, or you want to compare based on demographics of the students,, {MANOVA can be used, however, it is a little difficult to interpret} you may do separate ANOVAS for each DV.
@David - Thanks. If I use ANOVA or MANOVA, will this allow me to check the satisfaction on all constructs and overall? Or I need to check individually for effect of Program Type on Construct 1 of satisfaction, Construct 2, Construct 3 and Overall. Is this a case to propose a model and apply SEM?
@Muayyad - There are two independent variables - Program Type (Business - Core, Business - Elective and Non Business) and Number of courses (1, 2-3 and More than 3 courses). I need to see if these two variables affect the satisfaction construct-wise and overall.
MANOVA will test all three at once.
If you are uncertain about such basic tests as ANOVA, I do not recommend that you take on more advanced methods such as SEM.
Try the ANOVA and MANOVA and get results. From there, you can then ask further questions on the outputs and how they relate to your study.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire