In: Statistics and Probability
Why can we not rely on predictive methods for alternative explanations?
Predictive methods is the practice of extracting information from existing data sets in order to determine patterns and predict future outcomes and trends. Each model is made up of a number of predictors, which are variables that are likely to influence future results. Once data has been collected for relevant predictors, a statistical model is formulated.
When we use Predictive methods, we don't just systematically manipulate the variables of interest we just measure them. Which means even though we may observe a relationship between variables but we couldn't conclude it is a casual relationship since there could be other alterntaive explanations for this relationship. An alterntaive explanation is the idea that it is possible that some other, uncontrolled , extraneous variable may be responcible for the observed relationship. In conclusion because these methods leave the possibility for alternative explanations, we cannot use them to establish cause-effect relationship using predictive methods.