What is wrong with using Excel for conducting fire department data analysis? Nothing is wrong with it, but there are better ways. Fire Department data analysis often involves many, many steps that can be time consuming and difficult to double-check using Excel. Here is an example of conducting a response time analysis:

- Step 1: Download raw data from your emergency dispatch center and read that data into CAD
- Step 2: Data includes 100 different dispatch codes that do not translate directly to what we consider major incident type categories such as fires, EMS, et cetera … so you code the 100 dispatch types and use excel to create another column of incident types.
- Step 3: You only want to include emergency responses in your response time analysis, so you eliminate all public service calls.
- Step 4: Finally, you run the response time analysis by incident type using the 90th percentile times.

Imagine working for several hours to prepare a report that includes several different types of analysis. Then, after delivering the report, your chief ask whether or not the analysis included mutual aid responses and that they wanted b0th percentile times instead of 90th percentile times. You probably don’t remember if where you did or did not include mutual aid responses and you realize that you will have to redo significant portions of the analysis to compute 80th percentile times. Scenarios like this are exactly why Excel is not well-suited for complex analysis or analysis that you expect to repeat on a regular basis (such as monthly reports).

So what is the answer? The answer is statistical programming languages that do the exact same thing as Excel, but in the form of computer scripts (basically text documents that contain computer code). Imagine writing a word document that outlines exactly how you did your Excel analysis and with what data and having that word document actually do the work and spit out the results. The beauty of this is that you can go back and update your work and rerun the analysis in seconds rather than hours. Furthermore, you can use that computer script to re-run the exact same analysis month after month without having to do any work. Enter statistical programming languages …

There are several different statistical programming languages out there … SAS, SPSS, S, S-Plus and R are the most popular. R is, however, the only free program and it allows users to develop their own additions for industry-specific analysis. I have been working on an R add-on package called FireTools that will allow you to read in CAD and NFIRS data and automatically run several different types of fire department analysis. Once this package is released, I will use this blog to provide tutorials on exactly how to use this add-on package. In the meantime, you can already get started on learning the basics of R. The following tutorial provides a good introduction using the analysis of baseball statistics.

http://www.oreillynet.com/pub/a/network/2004/10/27/baseball.html

After completing this tutorial you should begin to understand how analysis with R is similar to what can be done with Excel, but lends itself well to updating and repeating analysis. And, because R allows users to develop and publish add-on packages, R is a much more powerful analysis tool than Excel. Good luck with the tutorial.