STATISTICA Blogs

I sat down with Roy Brooks (RB) Wiley  and he took a moment to discuss his use of STATISTICA.

1.  What is your background (career, education, etc.)?
I am currently a Principal Investigator for the North American rail industry. I have been working in this industry for more than eleven years and have also held positions such as IT Manager and Software Development Manager.

Data collection has always been part of my day-to-day life, even before my first career out of college as a ranger for the Forest Service and then the Bureau of Land Management. At the time, with a BS in Range and Forest Management, the experience of data compilation meant days and weeks with tape calculators and typewriters. On the other hand, it was clearly the best job in the world. Months on end were spent hiking, jeeping, snowshoeing, helicoptering, and horseback riding all over the mountains and foothills of Colorado and Wyoming just to collect data.

While finding that early desktop computers held promise for greatly increasing data analysis productivity, I followed a circuitous path to an IT career. Along with C, Cobal, and Fortran, it was Visicalc on an Osborne, to Excel on a Macintosh, to programming data crunching routines out of desperation with no alternatives. That led to more school and an MS in Industrial and Systems Engineering. My varied background met with the happy accident of crossing paths with THE world class railway research facility needing an engineer who could program well.

2.  What challenges did you face that drove you to use STATISTICA?
A graduate student project turned into a business venture of analyzing and solving capacity problems on telephone switches, data networks, and early ISP modem banks. Excel and a custom-built database with switch log processing and data reduction coded in C were the primary tools. But it meant weeks of effort and was somewhat error prone, so a lot of time was spent re-checking data and outcomes from intermediate steps. STATISTICA 4.0 happened to beat these challenges by nearly eliminating the need for programming while providing excellent and accurate outputs for reports.

3. How do you use STATISTICA? Which modules?
The Basic Statistics support at least 50% of our work. I'm a big proponent of doing thorough EDA (exploratory data analysis) where most of the day-to-day answers for customer needs are found. These tend to involve troubleshooting a rail vehicle performance problem. There is also heavy reliance on the Industrial Statistics/Quality tools, Nonparametrics, Distributions, Multiple Regression, and Advanced Linear/Nonlinear Modeling. I hope to get into tapping the R integration capabilities soon.

4. Why did you choose to use STATISTICA vs. the competitors, such as SAS, SPSS, etc.?
Originally, two criteria were key: an integrated programming language plus having the most accurate calculations and statistics available. STATISTICA 4.0 was the winner versus JMP, Systat, and SAS, amongst others if I remember correctly. When researching various reviews, I found that STATISTICA was always the best in accuracy tests, the interface was judged the easiest to use, and the SCL (STATISTICA Command Language) could control every feature and function. In the first month of use, the usual ten-day data processing and analysis effort collapsed to a single day, including error checking. As a result, there was never any rework required. Ever since, I've been introducing and proving that STATISTICA is a necessary tool for subsequent employers' needs.

5.  Did STATISTICA give you the solution you were seeking?
Without a doubt. The integrated macro functions for importing, merging, recoding, and batch transforming data can save weeks of writing and verifying C code or Basic, etc. Furthermore, at any data manipulation step, it is incredibly easy to run some breakdowns or cross tabs or frequencies and know instantly that the processing worked correctly. The statistical powers were what I expected, but the longer term benefit has been in learning to use and become proficient with so many more powerful and sophisticated statistical analysis techniques. STATISTICA is always ahead of my needs.

6. STATISTICA strengths?
Besides the data handling and augmentation capabilities previously mentioned, several capabilities also stand out. The extremely wide range of graphical output and analysis functions coupled with customization of any detail imaginable are unmatched. Custom styles are a plotting diamond mine. Raw horsepower or the capacity to handle millions of rows of data efficiently and to produce results quickly is always amazing. For innumerable tasks, being able to achieve an accurate result within minutes vs. hours or days has brought many sole source projects our way. It also always impresses me that with incremental upgrades, Statsoft will include so many new and useful features.

7.  Would you recommend STATISTICA to a colleague or friend? Why?
I always do but don't really have to. The STATISTICA analyses outputs and plot products speak for themselves at seminars, conferences, and project meetings. Customers ask what software I'm using, and at least half-a-dozen have purchased multiple licenses. I provide periodic training to customer groups in use of STATISTICA for analyzing data that we provide to them and/or for their own data.