What are some of the pitfalls and limitations of regression and correlation analysis and how might they impact a business decision?

Multicollinearity is the most common problem. This occurs when there are a lot of explanatory variables that have an extremely linear relationship. It can be hard to determine exactly what causal effects one variable has on another. Additionally, regression models are only as good as the data that they use – meaning any errors in the data set (such as outliers or missing values) could potentially lead towards inaccurate results if not addressed properly.

Finaly, regression models are based on linearity of variables. However, real world scenarios involve systems that have non-linear relationships and can change over time. This means that while regression/correlation analysis can provide useful insights into certain types of problems, relying solely on them without taking into account other factors may not always produce accurate results especially when dealing with complex decision making situations. Businesses should be cautious when they interpret the results of such analyses to prevent making bad decisions that are based more on assumption than fact.