The analysis is crying for multivariate regression analysis. Do the conclusions change if we add other (at least one) explanatory variables? Even partial correlations would be enough at a preliminary stage: controlling for the level of development, what would the graphs look like?
The analysis is crying for multivariate regression analysis. Do the conclusions change if we add other (at least one) explanatory variables? Even partial correlations would be enough at a preliminary stage: controlling for the level of development, what would the graphs look like?
ReplyDelete