For example, if I want to study the red-light running behavior of drivers. I might ask them questions using a Likert scale such as:
How likely are you to run a red light if you are in a hurry?
How likely are you to run a red light at midnight? etc.
(Please do not take these questions seriously. I just wrote them down off the top of my head.)
After preparing the questionnaire, I can distribute it to drivers in "several countries". Can I combine data from all those countries into one data set and analyze it as a whole?
Logic: My logic here is that even when we talk about red-light running behavior within a country/region, there is so much variation within the country/region. For example , such behavior might be different in urban, suburban and rural areas. Similarly, such behavior might vary even within a city but we still perform these studies and report our results. So, can we do it at a larger level i.e. at several countries level?
I think it is directly related to your sampling method and your target sample that you defined in your study. If all these population groups in different countries could be in the same sample groups, it would be fine to analyse them as a whole. However, it is more common to analyse data based on a comparison between different case studies and contexts. For example, young generation care about traffic rules less than older in case A, while it is opposite in case B. Then, you can discuss the reasons behind the possible differences that you observe in different countries/contexts. Good Luck.
Yes, you can. And it would be very interesting to relate this question with context questions like "why", "where", "when" and social variables like sex, age, and so on. Good luck
Muhammad Abdullah, I think you have partially answered your own question. If there are noticeable differences from one region to another within countries, combining all data from one country and all data from another country is likely to result in a bland collection of information. Some interesting, and perhaps important, features of your topic will be obscured. However, I guess that much depends on how you code your data and the kind of analyses you perform in relation to demographic variables.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire