This is a draft at the moment but thinking about it for a long time. I want to see if local populations' perceived threat, stigmatising attitudes and xenophobia (IVs) is related to immigrants emotional exhaustion and isolation (DVs). In this model, IV's will be collected from local people where they live with immigrant community and DVs will be from immigrants from the same neighborhood. I know it is problematic but how should we design and test such a model in a proper way.
I agree with the problematic part and I really don't think it's a good idea to build a regression model using data from different groups of people. You could just run a correlation though. And do not over-generalize your conclusion.
Thank you for your answer Liangzi, we are actually thinking about supporting quant results with qual findings but concluding a model seems not possible when unit of analysis is individuals.
Such a multilevel research design can be employed. However, the issue I notice is relevant to your research problem. Local people data (IV) and immigrants data (DV) should be linked in such an application, and survey data should be linked accordingly, however this seems challenging in terms of your research context.. I hope you can handle easily..
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning