Bus 3059 business analysis week 4 assignment 3: multiple regression
Before running a regression, it is important to make predictions about the expected direction of each variable in order to better understand how they may influence the overall model. For instance, if one variable is positively correlated with another – then this would imply that as one increases, so will the other – resulting in a positive relationship. Conversely, if two variables are negatively associated then any increase in one should be accompanied by a decrease in the other (negative sign)
Additionally, certain patterns can also be observed when looking at different types of relationships between variables which can further inform expectations for signs prior to running a regression. For example – factors related to economic growth or consumer spending often tend to have positive correlations while those tied to risk aversion typically show negative associations.
Ultimately, by making educated guesses about which way signs should go based on prior knowledge and existing data – researchers can create more accurate models that provide meaningful insights into underlying trends and their potential impacts.