I would support all that Art Kendall says; except that I do not have the public health experience to judge whether one package is in fact "best" for you Master's candidates. But I would add one thing: Make more than one package available, and provide enough resource material so that the interested student can at least invoke, say, SAS or MINITAB, and thus begin to acquire a feel for two rather general qualities: (1) what the several statistical packages all tend to have in common (like, variables being represented in columns and cases in rows of a rectangular data matrix, ability to read data in usefully formatted plain text files, and what a generally useful file format might be); (2) some of the particular differences between packages (in, say, ease of operation, steepness of learning curve, style of output presentation, etc.).
Not all of your students will stay in the public health field forever, or so I imagine by analogy with fields with which I am familiar, and even if it is arguable that SPSS be the preferred or default package that you should teach (for reasons like those Art invokes) in that field, it may not be the most commonly encountered package in the field where a particular student eventually ends up. For such a student it will be advantageous if s/he can fairly readily switch to another package favored by the employer, or by the professional environment. And some things are more elegantly, or more easily (or both!), done in one package than in another. (I can supply examples, if you like.) That there are differences of opinion among your faculty about which package is "best" sounds to me like a very good thing. Students need to hear about such things, and not suppose that everything comes with a unanimous imprimatur... Cheers! -- Don. On Tue, 25 May 2004, jj diamond wrote: > i am on the faculty at a medical school in philadelphia. i also have > an appointment in the graduate school wherein sits our masters in > public health program. a group has been putting together a lab course > that will complement the statistics classes. in the lab, we want to > teach the students how to use software that will - among other things > - allow them to do calculations. we have had lots of discussions > about which package to use, as well as larger discussions about which > package would give the students something they can put on their CV > when they look for jobs. there are differences of opinion in our > group about which package should be used. part of the problem is > trying to weigh ease of use, cost, CV enhancement, etc. > > so, to my question, and thanks for any feedback. among SAS, SPSS and > STATA (alphabetic order), which one do you all think is called for the > most among graduates with masters degrees. as i said, the students > are in public health. they are not stat majors but many are likely to > get jobs where data are collected and analyzed. none will even > identify themselves as "power users." but we know that some are > looking at jobs that ask for software capabilities. > > sorry to ramble, and i've probably left off something key. > nonetheless, if you have ideas to share, i'd really love to hear them. > thanks very much to all. ------------------------------------------------------------ Donald F. Burrill [EMAIL PROTECTED] 56 Sebbins Pond Drive, Bedford, NH 03110 (603) 626-0816 . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
