We have conducted three representative surveys with the same questioner on high school students from three different cities. The cities have different characteristics from ethnic, economic and other perspectives.
I would like to analyze them together, and to see if the city where they socialize (more exactly some of the characteristics, that they face - ex. ethnic composition, inter-ethnic relations, etc.) influence or not the issues that were asked in the questioner
It sound like you can do what you want using dummy variable regression, where you create an ID variable that uses dummy variable coding (i.e., the three cities are coded 0, 1, and 2 as their IDs). For some of your analyses, you may also want to examine interaction effects (e.g., does the effect of economic perspective differ between cities).
A good introduction to dummy variables and interaction effects that involve dummy variables can be found in Hardy's "little green book" from Sage, "Regression with Dummy Variables."
thank you very much for your answer. and for the book suggestion, I will definitely look into it.
If I understand how logistic regression is using categorical variables, correctly, if I code the dummy variable this way, then it will compare them only to the base-city. I can add more variables to have all of the possibilities covered. This would be no problem in the case if I have the 3 cities, but hopefully we will have data-sets from 3 other cities...
You may refer to inferential statistics using SPSS. That will help.
This link maybe useful:
https://www.surveymonkey.com/mp/how-to-analyze-survey-data/
- Badges
- Science topic