A questionnaire is used to gather data on women's experiences in an antenatal clinic. Each question within the questionnaire has four options for an answer and there are 20 questions. The questionnaires are given to 20 respondents. What statistical method can be used to study the differences in responses obtained from the respondents?
Dear Wilbert,
If you are planning to analyse the response to the questions individually then you would need some kind of grouping variable and use chi-squared or Fisher's exact. However, you may need to raise your significance threshold (e.g. to 99%) to take into account the number of tests you are running.
If you wish to analyse all the responses together the best test might be Friedman's ANOVA. However, you may need to transpose your raw data matrix and look for differences between respondents rather than individual questions.
Thanks Peter. That helps! The intention is to analyse all the responses at the same time and thus determine if there are differences between survey participants!
The main problem I see iswith having only 20 respondents. In general, this will give you too little power to detect effects. I suggest you search the internet for more information on sample size calculations.
I think you should use the non paramatric statistics method due to the sample size you have..
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire