I want to test 2 ( High Crowding * Low Crowding) * 2 (Deception* Not Deception) on emotions of consumers (experimental survey study design). I have four vignettes representing (Crowding and Deception, Crowding Non-Deception, Non-Crowding and Deception, Non-Crowding No Deception) vignettes. If I give one vignette for one respondent rather than all, is it ok...
If I want to test all four from one respondent how should I include all four vignette to test emotions of respondents. What is the way that I should include them to test dependent variable. Appreciate early help
Hi,
you need to have four groups of participants. Each group receives 1 vignette. You have to have in each group a minimu of 25 people, but better 30+. There is a special software to find out the optimal sample size (http://www.gpower.hhu.de/en.html). It is free to use, but you have to know which analysis you are going to apply (probably ANOVA F-Test, group comparison, 4 groups).
You are not allowed to give to each person two or more vignettes - the results of the second (thrid, fourth) answer might be biased through their previous answer and the overall understanding on what is your research about. Moreover, the induced emotions might change or "overlap".
Remeber, that all your vignettes should be almost identical, apart of the manipulation (1-2 words). This is the only way to exclude all other influencing factors! Do not forget to test for confouning variables like age, current emotional state, and so on.
It is also recommendable to make a pretest of your vignettes on a small sample of participants. If you test for crowding/deception, you can do it with students as well as your main study. You are not able to merge the rpetest and the test, however, since other factors (time of the day, weather, light) might have influence on the answers of participants.
Regards,
Eugene
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA