Systematic Data Analysis and Reporting

I wish that there had been a course in college or graduate school called "Data Analysis 101".  I have taken a number of statistics and computer programming courses that have taught me specific tools.  But there was no course available to help prepare me for organizing my work as a programmer/analyst.  I understood analytical issues and I knew how to extract information from data, but the missing component was a systematic work process that would structure the analytical activity in a way that would increase my efficiency, document the results, and ensure that I could replicate the results, should that be required.  Therefore, I found it necessary to develop an approach and set of standards that would organize my work so that the following questions could be answered:

• Who - needs this information?

• What - questions do they want answered?

• When - do they need the answers?

• Where - are the data?

• Why - is this important?

• How - did I get the answer?

It is the “How” question above that will be the focus of most of this book.  Chapters 1 through 9 provide an overview of my vision for a systematic approach to analytic work.  The Appendix, which contains three comprehensive (although relatively simple) case studies is the real "heart and soul" of the book.  The case studies apply the process using data representative of three different functional areas.  It is my belief that case studies bring concepts alive and thereby stimulate the imagination of the reader to take those concepts and implement/tailor them in the unique context of their own work environment.

Throughout this book you will see many references to the SAS^® System.  Please note that this is not a book about SAS programming.  There are many other excellent resources devoted to that topic.  The SAS System is used to demonstrate concepts and tools that are discussed in this book.  Other software tools could have been used.   I chose the SAS System because I believe it is an excellent tool.  However, you can apply these concepts using any tool set.

Not all analysts will choose to use a systematic approach to data analysis.  For those that do, the decision is the easy part.  Following through on that decision and developing the discipline to apply the process is actually very challenging, particularly when under the pressure of tight project deadlines.  With that reality in mind, I have attempted to develop a process approach to analytic work that is simple yet comprehensive in scope, flexible in how it can be implemented, and practical, i.e., the benefit of using a systematic approach is greater than the costs of doing so.  If you choose to adopt a systematic approach to your work, please remember that it is truly is a craft, and therefore takes time and practice to master.  The first time you attempt to apply these concepts you will be frustrated.  Persevering through that initial experience becomes a matter of faith that in the end, the effort will have been worthwhile.  Over the past twenty years I have become increasingly convinced that a systematic approach to analytic work has a significant payback, to me professionally as well as for my client.