I'm gonna use SPSS for the analysis. For my survey, the same respondents have seen pictures of two different products (3 categories: total 6 products, the difference is the design), and rated their preference in the same question.
For example, given a products A, B of the same categories, respondent were asked to rate their relative preference: 7 Likert scale (anchored with definitely option A, definitely option B)
How can I test the main effect of the design on their preferences in SPSS.
Thank you
This is not entirely clear. Is the design the following?
Rate Product A vs. Product B (7-choice scale) for Category 1
Rate Product C vs. Product D (7-choice scale) for Category 2
Rate Product E vs. Product F (7-choice scale) for Category 3
If this is correct, you could do a one-way ANOVA to compare the three categories. This will tell you if choice is significantly different across categories. If you have 6 different products compared in pairs across categories, you have confounded product with category, and it will not be certain if differences are due to category or product. That limitation should be noted.
If you want to know if there is a preference for one product over the other, you will need to compare the mean you have for a category against the middle response. With a 7 point scale I assume 4 = no preference, so you could compute a one-group t-test to see if your mean for each category is significantly different from 4. You would have 3 of these, one for each category.
Dear Mr. Paul E. Spector ,
The same question is repeated for each category (products in the choice set are form the same category with the difference of their appearance)
My measurement question for my dependent variables was:
For category 1, which product would you be more likely to choose?
1=Definitely option A, 4=Indifferent, 7=Definitely Option B
Same with category 2 and 3.
For the one way ANOVA, I need to have a factor (IV) which I don't know in my case what to input since my independent variable was pictures of the products paired in categories.
I did try the repeated measure Anova since the same respondents were measured for all the products. And I got the following with Wilks’ Lambda=0.879, F(5,434)=2.000, p=0.006.
But when it comes to the mediation effect, I couldn't find a proper way to analyze that.
Could you please advise. Thank you in advance
Wilks' Lambbda is for the multivariate approach. I think you need to concentrated on the univariate approach.
see: http://www.statsoft.com/Textbook/ANOVA-MANOVA
Having several repeated-measures DVs one can apply a univariate approach (also called Repeated Measures sensu stricto or split-plot approach) or multivariate approach (or MANOVA). In univariate approach, RM levels are treated as deviations from one variable, their average level. In multivariate approach, RM levels are treated as covariates of each other. Univariate approach requires sphericity assumption while multivariate approach does not, and because of this it is becoming more popular indeed. However, it spends more df and thus needs larger sample size. Also, univariate approach retains its popularity because it generalizes to Mixed models. When sphericity assumption (and beyond expectation more general compound symmetry assumption too) holds results by both approaches are very similar, as far as I know.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS