I have two sample groups of same respondents in a pre and post intervention. Let me be clear, a group of youths to be asked about unemployment before a civil conflict and unemployment after the conflict. The aim is to determine whether unemployment has been reduced or has gone up by comparing the result between the two periods. same group of 40 participants to answer 20 statements related to youth unemployment. I intend to use the Linkert-Scale ranking where respondents could either strongly disagree, disagree, neither agree nor disagree, agree, or strongly agree with a number of issues related to youth unemployment. Please advice.
I am not sure what you intend to do with your outcome variable, but it seems to me that you would calculate a mean of the 20 statements. With this, you can calculate whether the mean is different pre and post intervention. This is done with a paired samples t-test (paired, as you have one sample).
In addition to calculating and comparing means in the pre and post test, you would need to do more advanced statistical analysis for Ph. D. purposes. In that regard, you can also utilize and compare Scatterplots, Pearson's R and Regression Analysis coefficients for both the pre and post test
If normality are present you can use Student's t, if you don't find normality is better use the Welch's t. And I agree with Russell Foote, you'll need more powerfull statistical treatment.
How many variables have in your research? And, your questionnaire is unifactorial? If not, the mean comparison will not sufficient to your research.
- Badges
- Science topic
- Similar topics
- Econ
- Manufacturing
- Industry