Means Testing:
Generally this type of analysis is used in hypothesis testing when a client or researcher has a specific hypothesis in mind when testing a theory or method of treatment/programming. Depending on the analysis used this can be as simple as a t-test for significance or multiple factor Anova. Generally reserved for the research lab this type of analysis can be used when needed providing certain statistical assumptions can be made.
Regression:
Regression modeling allows for prediction of an unknown variable based on observations/measurement of another variable(s). This is a fairly straight forward process unless and until multiple factors come in to play. Where multiple factors exist it is possible for the predictor factors to ‘interact’ with each other and enhance or attenuate the dependant variable. Through rigorous statistical modeling I ‘control’ for those factors in constructing statistical models of prediction. I can provide this analysis when necessary and appropriate.
Correlation:
Correlation, also known as covariance, measures how much two things are related and the strength/direction of the relationship. While not necessarily cause and effect, we can measure the degree to which two things are related and how strongly. Correlations can be positive (increasing together) or negative (decreasing as the other increases) and having knowledge of this relationship can be advantageous to businesses or service providers.