Question

Whywould researchers believe that an ordinary least squares regression is superior to a partial correlation coefficient?

Answer 1

partial correlation coefficient allows us to measure the degree of relationship between two variables by adjusting the effect of others so it just merely tells us how strong/weak a relationship is. On the other hand OLS(oedinary least squares) is a mathematical procedure which allows you to get the BLUE(Best Linear unbiased estimators) of the coefficient of the model.

so what does it means?

It means OLS will allow you to form a model and will give you best estimates for the parameters of that model.

Once you get those parameters you will be able to forecast values (extrapolate) in future. So we can clearly see that OLS is a better technique becuase it allows one to form a model with best estimates which can then be used for predicting values of dependent variable for some given set of independent variables.

correlation is just a measure of how strong a association between two variables is.