Hello! can anyone help?
I'm currently working on a research paper, where I created a questionnaire containing 15 items related to challenges imposed by climate change. Respondents are supposed to choose 1 out of 3 options (whether to adapt, mitigate or build resilience), which signifies their most preferred choice of action for each imposed challenge.
Now, I would like to understand the most preferred choice amongst respondent.
I already used analysis of variance (ANOVA), but one of the reviewers stated the following;
"The 15 climate action questions are not continuous, so that we could not employ variance analysis. Instead, you could see the difference in the number of respondents for each category--meaning count measures-- by employing techniques such as cross-tab/chi-square(x2) analysis
The questionnaire sample is attached
I'm wondering whose approach is right, mine or the reviewer's
More information is needed. What did you wrong the ANOVA on? I assume you used something like a latent trait model and used ANOVA on the resulting trait(s) to compare groups?
Did you consult a statistician before conducting this study? Did they recommend an ANOVA?
In attachment an R script for the chi-square test as proposed by a reviewer. The Word doc is for the results.
I analysed the data for Table S3. April 2017 (0 month).
The Fisher's Exact Test for Count Data with simulated p-value (based on 2000 replicates) is very significant: p-value = 0.0004998.
It appears from you questionnaire that you want to compare the answers from three different countries. If so, I agree with your reviewer that you should not use ANOVA because that requires a continuous dependent variable. Instead, you should indeed use a 3 x 3 chi-square for each of your 15 questions.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Modeling
- Analyses of Variance