This document analyzes the relationship between GDP (pib00), investment (i00), and consumption (cu) in a country over 35 years using regression analysis. It finds that:
1) The initial regression shows that i00 and cu can explain over 93% of the variation in pib00.
2) Adding the lag of i00 (i001) slightly improves the model fit but i001 is not a statistically significant factor.
3) Diagnostic tests reveal that the errors in the regression are not independent, violating an assumption of the model. Specifically, both the Durbin's alternative and Breusch-Godfrey tests detect the presence of serial correlation.
This document analyzes the relationship between GDP (pib00), investment (i00), and consumption (cu) in a country over 35 years using regression analysis. It finds that:
1) The initial regression shows that i00 and cu can explain over 93% of the variation in pib00.
2) Adding the lag of i00 (i001) slightly improves the model fit but i001 is not a statistically significant factor.
3) Diagnostic tests reveal that the errors in the regression are not independent, violating an assumption of the model. Specifically, both the Durbin's alternative and Breusch-Godfrey tests detect the presence of serial correlation.
This document analyzes the relationship between GDP (pib00), investment (i00), and consumption (cu) in a country over 35 years using regression analysis. It finds that:
1) The initial regression shows that i00 and cu can explain over 93% of the variation in pib00.
2) Adding the lag of i00 (i001) slightly improves the model fit but i001 is not a statistically significant factor.
3) Diagnostic tests reveal that the errors in the regression are not independent, violating an assumption of the model. Specifically, both the Durbin's alternative and Breusch-Godfrey tests detect the presence of serial correlation.
(Analytical Methods for Social Research) Janet M. Box-Steffensmeier_ John R. Freeman_ Jon C. Pevehouse_ Matthew Perry Hitt - Time Series Analysis for the Social Sciences-Cambridge University Press (20.pdf