Have you ever heard of the terms, Poisson, Logistic, Probit, Logit, Panel Data, Principal Component Analysis, Spatial Econometrics, Quantile Regression, Propensity Score Matching, or Seemingly Unrelated Regression?
Didn't think so. Yet, these newer linear models have been developed in recent years to deal with the data as it is. No longer is it is assumed you must create a "Randomized, Balanced, Factorialized Design" in order to extract meaningful information from your data. The real world does not allow the time nor provide the money to conduct Design of Experiments by the book. You need analysis that can handle your data as it is, in a meaningful turnaround time. Developing models to predict or explain is crucial if a business can expect to succeed in today's economy. It's also imperative that the results can be communicated in a simple and meaningful way. Where non-technical types can immediately grasp the significance of the data. This usually means charts.
Enter the "R" programming language. A powerful analytical and graphical open source program that is continually being expanded in its features. Universities have nearly all turn to this program to teach their undergraduate and graduate level applied math classes. The most redeeming feature of "R"? It's FREE! Yes, a program as powerful as SAS ,SPSS, or MATLAB is absolutely free and available to all who want to learn it without cost, except for, the time needed to invest in its somewhat prickly and demanding language.
This blog will take you through examples and uses of these various modelling analytics with practical examples and with a minimum of math jargon.
Problem Definition and Project Outcomes Defined
This is usually accomplished with a sitdown/phone call/skype where you describe the problem and scope o...