Search 75,768 tutors

Statistics Calculation Packages

In the past few days I've helped several students deal with statistic "packages" that are available to "help" people deal with those sort of calculations. The most famous of these is probably SPSS, but there are others such as STATA and R. The problem with all of these is that, while they are directed at those who are not necessarily well-versed in math, engineering, and computing, they demand expertise in all of those areas just to get on with getting a simple average. This is despite having newer graphical user interfaces available. In the case of SPSS, there's a tendency to throw jargon around, then have outputs that are phrased in a confusing manner. An example is showing "significance" by showing the value of p, which at high significance approaches zero---thus the question "is this statistic significant?" is answered "yes, this statistic with a ZERO in the "sig" column is significant". That's very counterintuitive for the new user. The R package is more difficult, since it works by "command lines", in a throwback to programming in BASIC or similar non-compiled languages, with only a few graphical functions available to help.

As this is written, I'm about to start teaching an MBA-level course in managerial statistics, for which the statistics package students will be required to use will be plain, old, "kind to its mother" Microsoft Excel. In Excel, once the data are set up, getting an average, a standard deviation, a regression, or other statistics are all very straightforward. You just enter the pre-set function or a formula in a cell in the spreadsheet, and you get the result. No muss, no fuss. If you need a fancy graph, you can generate that. If you need a nice table, you can copy the results into cells and copy those into Word. Virtually everybody has Microsoft Office for either a PC or a Mac nowadays, and by doing stats in Excel, you don't need to have and learn some "special" package to accomplish your goals.

Having said that, though, if your class or thesis advisor insists on using a package like SPSS (or if your data are already set up for such a package), I stand at the ready to help you through it. If you have your choice, though, consider using plain old Excel. I think you'll be glad you did.