Raynald Levesque: Solving complex or unusual problems posted to the list is a fun way to learn and share knowledge. I only answer questions posted to the list in the evening. I read all the titles…I don’t necessarily read all the questions. When they ask statistical questions, I don’t bother to read them unless I’m interested in learning more about that particular technique. When it’s Marta, I look at it more carefully because very, often she has the right syntax and the example and I can use that. She’s aware of that, and a few days later she sometimes sends me an improved version.

King Douglas: Say we are coming to you for help with an SPSS problem. What's the best way for us to frame our question?

Raynald Levesque: The best way is to supply dummy data, any syntax you attempted, and an actual example of results you are looking for. An example is worth a 1,000 words. Words do not mean the same thing to everybody. An example does mean the same thing to everybody.

The top deterrents that will stop me from answering a question posted to the SPSSX list are, first, if the question is not clear and there is no example. I like to answer difficult syntax questions but hate to guess at questions. And, second, if the question is too trivial.

King Douglas: Has there been an SPSS problem that you couldn't solve without help?

Raynald Levesque: Yes, my first post on the SPSS newsgroup, back in 1998, was about a macro problem. I believe I have asked less than 5 questions. You see, the actuarial exams are very difficult—the toughest thing I have done in my life—and it’s all self-study. You take the material, you study it and you go, you pass the exam. This was a helluva training to learn by yourself. You have nobody else but yourself, and that was the approach I took to learning SPSS. I think I know enough SPSS that I can do whatever I need with it.

King Douglas: I imagine that your mastery of SPSS could give you a certain amount of satisfaction.

Raynald Levesque: It does.

King Douglas: Do you have a preferred method of displaying SPSS output in a report?

Raynald Levesque: I like to use graphics when appropriate.

King Douglas: What application(s) other than SPSS do you use when reporting the results of SPSS analysis?

Raynald Levesque: Results are displayed in Microsoft Word if a formal document is required. Otherwise, I display them in Excel.

King Douglas: What makes a problem difficult?

Raynald Levesque: When the question is not well defined.

King Douglas: Many problems can be solved in a variety of ways. Do you value parsimonious code and work to attain it? Or is parsimony irrelevant given that a computer is going to do the work?

Raynald Levesque: I value compact code because I find shorter code to be easier to maintain and modify—macros are helping a lot here. I value fast code because most of the code I write is to be executed on different client data and some of these files can be huge…sometimes over 10 million cases. Some runs took 15 to18 hours on a fast PC. And I value automation…that is, code that self-adjusts to varying data conditions. For instance, code that does not create an error when a table is empty. Automation is a big thing.

I also try avoiding unnecessary data passes and I am careful about memory hungry commands…like CREATE when I’m working with especially large files.