I’m doing a research right now and I’m trying ti find how and to what extent corruption impacts the interest of retail investors in a particular country. After a brief lit review, I tought to correlate and do a regression between the corruption perceptions index and equity shares ownership by households in a given country. Now, I think that If I pop in other control variables it will dilute the significance of this one and I won’t be able to distinguish how high of significance is just corruption. Should I be adding things such as GDP, interest rates, inflation or some other even?
I believe you may add variable, such as, GDP, interest rates, inflation. You may use multiple regression analysis to analyze the relationship between a single dependent variable and several independent variables, which can be used when you want to predict the value of a variable based on the value of two or more independent variables.
Dear Dr. Mahfuz Judeh , do you think I should include some other variables? Although literature argues that they should be strongly correlated, my r square is only 0,52 and durbin watson index is 0,7. I don’t understand why
David L Morgan I have data available from 2013-2021 from the Central Bank on the amount and type of net financial assets and their change by year. My hypothesis is that corruption will impact the change of these throughout years. Although it doesn’t really decrease every time there is a slight change in the corruption index it is on a general downward trend as is the corruption index since 2013. Do you have any suggestions? How can I shows this and what would I need to account for?
One important reason for including control variables is to "eliminate threats to validity." In particular, you want to show that your key independent variables have effects that are "above and beyond" the effects of other relevant variables. If you don't do this, then you are open to counter-arguments that effects you present were "really due to..."
So what do I do then? Just plot 2 separate regressions? One including the control variables as independent and one just including corruption? Then I check the change in R square? Or I can just plot a multiple regression add GDP, inflation and interest rates with other ones and check the significance level of each one and if its’ under 0.05?
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning