I am doing a quick study on availability of agricultural insurance and food security. I am using data for 3 years from 20 countries. I am using the Food Security Index (independent variable) to the availability of agricultural insurance (dependent variable).

Availability of agricutlural is a dependent variable which ahs 3 independent variables: FDI, Government spending on agriculture (%) and Insurance penetration ratio.

I'm just a bit confused as to which regression i should use.

Thanks alot

  • $\begingroup$ Just to add, i am trying to compare the Food security Index to the availability of agricultural insurance. I want to assess whether there is any affect $\endgroup$ Jul 26, 2018 at 3:34

1 Answer 1


You will use simple multiple linear regression.

Your $y$ variable (dependent) is: Availability of Agricultural Insurance (AAI)

Your $x$ variables (independent) are: FDI, Government Spending on Agriculture (GSI), Food Security Index (FSI), and Insurance Penetration Ratio (IPR).

Your final model will look something like this, where you are estimating the $\beta_i$ values:

$$ AAI = \alpha + \beta_1 \cdot FDI + \beta_2 \cdot GSI + \beta_3 \cdot IPR + \beta_4 \cdot FSI + \epsilon $$

  • $\begingroup$ and then would i use linear regression again to compare AAI to Food Security Index? $\endgroup$ Jul 26, 2018 at 16:39
  • $\begingroup$ Your question is confusing. You may want to add FSI as another independent variable in your regression, if you want to see the effect another independent variable has on your results. Just add in $\beta_4 \cdot FSI$ $\endgroup$
    – ERT
    Jul 26, 2018 at 16:43
  • $\begingroup$ basically what i want to do is compare availability of agricultural insurance to Food Security index. The Food Security index is an independent variable because the data i get is from the institute. The AAI is a dependent variable. So i should basically do a multiple linear regression on all the variables? Also do i have to care about heteroskedasticity and auto correlation like in multivariate. $\endgroup$ Jul 26, 2018 at 18:35
  • $\begingroup$ You shouldn't be worrying about heteroskedasticity until you understand the basics of regression, and what it should be used for. Do some reading on the basics of regression first. $\endgroup$
    – ERT
    Jul 26, 2018 at 18:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.