I would like help on what falls into the category of mathematica for dummies, running a multivariate regression on excel-imported data. I have spent many more hours than I would like to admit searching Mathematica’s website resources, watching demonstrations and tutorials, googling and binging, downloading and reading 20 or so pdfs, checking Wolfram community q&as, all in an attempt to find one realistic example. In my teaching and research activities, I routinely use SAS (and limdep and, occasionally, STATA – statistical software programs that enable users to easily input data, deal with missing values, transform data, estimate models with much much less suffering than I have found with Mathematica. I very much like Mathematica’s integrated approach and have used it, along with MathStatica, in teaching graduate courses for the past couple of years. Yet, during this period, I have been very frustrated in not being able to easily integrate ‘real world’ data sets into the courses and into some research related activities. And it is remarkable that, with all of my searching, I have yet to find one straightforward example that accomplishes all of the following (there are gobs of separate examples and toy problems illustrating aspects of each): 1) imports an excel file with headers, missing values (e.g. SAS uses a period (.)), large number of variables where the dependent variable is not the last variable; 2) transforms the data into a form that Mathematica requires; 3) calculates descriptive statistics for all variables, accounting for missing values across several varibles 3) uses LinearModelFit to estimate a multivariate regression model. If anyone can point me to such an example or a resource that includes such detailed examples (Varian's book including Belsley's contribution are not all that helpful for this purpose), I would be most appreciative. If this is not straightforward in Mathematica, then that also would be very useful to know. By the way, an excellent resource for SAS is a publication ‘The Little SAS Book’. Given Mathematica's power and complexity, such a resource for its statistical database users would be seriously welcome. I feel like I’m sitting in a Ferrari and have no idea how to access the gas tank without which knowledge all of the car's power and accessories go untapped! Pat