Hello,
For a project i want to do a regression analysis with 1 indepentent variable (amount of airbnbs per municipality) and multiple dependent variables that together make up "housing demand". I want to know how the number of airbnbs in an area influences the different indicators of housing demand. I have come to find that a Multivariate regression analysis is the right way. However, an assumption of this analysis is that the independent variabale is categorical, which, at the moment, it is not. My question therefore is twofold. Firstly, am i correct in stating that a multivariate regression analysis is the proper anaylsis for this case? Secondly, am i correct in stating that i need to transform my indepentent variable from ratio to categorical?
Thanks in advance!
Kind regards,
Willem van de Corput
You can run a regression analysis with multiple dependent variables within the framework of path analysis. There is no need to make your independent variable categorical. Categorizing could be detrimental because of a potential loss of information and statistical power.
In case your dependent variables could be seen as interchangeable indicators of a single latent factor (latent variable), you could use structural equation modeling with multiple indicators.
Christian Geiser Thank you very much for your response! I am very glad that I can use my independent variable as ratio data, cause I was worried about loss of information as well. I am not very familiar with SEM and path analysis so I will definitely have to look into that, thanks for bringing it up.
Hello,
It seems to me that you are talking about a multivariate analysis of variance rather than a multiple regression.
In this case, the dependent variables must be categorical.
Can you show an extract from your elementary table?
Best regards,
AP
Hello Antonin, what you are saying is indeed what I found on the internet. But since I dont want to lose information by categorizing I asked this question. I noticed in SPSS that it is possible to use a ratio variable but I am not sure if the results are still meaningful that way.
Regarding your question about the elementary table, we haven’t created the full database since we are still setting up the project. I will upload it here once we have it.
Anyway, thank you very much for your response!
For multiple dependent variable you can use MANOVA.
This would give you an independent relation of each Y with your X.
Alternatively, try multiple correlation. Exclude some week depends. Execute SEM hope to get significant with some independent..
Anuraj Nayarisseri, thanks for your reaction. Can you tell me what the risk is if my independent variable is not categorical but ratio?
Willem van de Corput , manova takes categorical. If it’s ratio go for regression. Make each dependent y with your x. Alternatively, change the ratio independent into factor using histogram bins and use that as categorical
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress