I conduct research on how different health factors, income and education affect people's savings behaviour. And the difference of this between countries in europe.
My regerssion is as follows:
Savings = General Health + longterm illness + BMI + Smoke + alcohol + Income + (control variables are: age, household size, gender)
I also want to control for countries so I added Factor(countries) to my regression. to add the dummy variable of countries. However, I am not sure how to interpret these results and whether this is the right way to examine the difference in countries.
can anyone help me with this?
Hi Julia, is the difference between countries of substantive interest to your research question or are you trying to eliminate the effect of between-country differences? How do variables relate to each other?
If you are actually interested in the differences between countries, a multilevel mixed effects model might be more appropriate (you seem to have enough number of countries). But also you seem to be adjusting for a lot of variables without a clear idea on how these variables affect each other. I recommend using a directed acyclic graph (DAG) for this.
This paper is a useful primer: https://academic.oup.com/esr/article/32/1/3/2404247
Hope this is helpful, Sebastián
- Badges
- Science topic