0
$\begingroup$

If I have a model of level-percentage such as:

Harvest(Y) = β0 + β1 GrowthPrice%(X1) + β2 GrowthCostIndex%(X2)

Regression Result: Y = 43946.11 + 7498.16(X1) - 30585.34(X2)

The dependent variable (Y) is the harvest in ton (level variable), then is it right to interpret the model like level-log model? "One unit increase in X1 will increase the harvest by (7498.16/100) unit of Y"

or how should I interpret it?

Thank you!

$\endgroup$

1 Answer 1

1
$\begingroup$

Since this is a linear model, an increase by 1 in one of the independent variables will increase the dependent variable by its corresponding beta. If your independent variables are percentages, than it would be correct to say "An increase in X1 of one percentage point is associated with a harvest increase of β1" (mind the "associated", in your statement you are implicitly assuming a causal effect).

$\endgroup$
2
  • $\begingroup$ Also one should add "expected" or "on average", as there is variation, even though assumed to be random. $\endgroup$ Jun 1, 2022 at 13:12
  • $\begingroup$ Thank you for your help! $\endgroup$ Jun 4, 2022 at 2:39

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.