Marta García Granero: Yes.

King Douglas: About how long have been using SPSS?

Marta García Granero: I began using SPSS, the PC+ version, in 1989 or 1990…I don't remember exactly…then I used SPSS 5 for a while, then 6 and 6.1—the first versions the University installed—then the University upgraded directly to SPSS 9, after that to SPSS 11.

My students at the University work with SPSS 11, although I also use SPSS 12. I haven't tried SPSS 13 yet, although Jon Peck [an employee of SPSS and an excellent programmer] keeps on tantalizing me with private messages about its great graphics capabilities. I don't like SPSS 12 graphics at all—I have complained a lot in the SPSSX-L list. I'm also really interested in SPSS 14 news—about Python language and all that, and I've already taken a look at www.python.org. I keep a copy of every message related to SPSS 14 that SPSS people send to the list.

King Douglas: What other statistics and data management applications do you use?

Marta García Granero: For data management, I use Excel sometimes. People at the University often come to me with databases in Excel, because they are not SPSS users at all. They usually have messy data that have to be cleaned before importing them to SPSS. I have found an excellent Excel add-in called Poptools really useful for that data cleaning, for instance, to detect numerical data in string format scattered along the dataset and that would be lost when importing to SPSS.

For statistics, I sometimes use freeware programs like WinPepi, EpiCalc 2000 and PQRS, which gives exact p values for a lot of distributions and is a good solution if you don't have the SPSS Exact Tests module. They do things that SPSS doesn't, although lately I've begun programming these methods with SPSS instead.

King Douglas: What SPSS modules do you use on a regular basis?

Marta García Granero: Base, Advanced, Regression Models and Exact Tests.

King Douglas: Do you consider yourself a specialist in any area of SPSS programming?

Marta García Granero: I like to program with MATRIX. I'm convinced that a lot of statistical methods not implemented by SPSS can be very easily programmed with MATRIX—it’s an excellent and underused tool. Some complex data handling, impossible to perform otherwise, can be done easily with MATRIX.