2
$\begingroup$

I have a correlation matrix for dependent var Vs independent variable as below:

Year 1 Sales Sales -0.5453 1.0000 PriceIndex -0.6089 Income -0.5033 Interest -0.3842

As we can see that all independent var have negative correlation with the dependent variable. However as one of the requirement that i have for my project is that based on these values can i predict correlation between my independent variables.

Kindly suggest if we can predict the same. Thanks, Shivi

$\endgroup$

1 Answer 1

0
$\begingroup$

Here's a quick analysis of your situation:

  1. Correlation among Independent Variables: The correlation between independent variables (PriceIndex, Income, and Interest) is not explicitly provided in your data. To determine the correlation between these independent variables, you would need to calculate their correlation coefficients separately. This can be done using the same method you used for the correlation between Sales and the independent variables.

  2. Prediction of Correlation among Independent Variables: It's important to note that the correlation between independent variables (multicollinearity) is distinct from their relationship with the dependent variable. The correlation between independent variables indicates whether they are linearly related to each other. You can predict or calculate this correlation by performing a correlation analysis between the independent variables themselves.

Here's how you can proceed:

  1. Calculate the correlation coefficients between each pair of independent variables (PriceIndex, Income, and Interest).
  2. Analyze the results to understand if there is a significant linear relationship between the independent variables.
  3. If the correlation between independent variables is substantial, it suggests multicollinearity, which can affect regression analysis. In such cases, you may need to consider techniques like Principal Component Analysis (PCA) or Variable Selection to address multicollinearity.
$\endgroup$

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.