A few years ago I interviewed for a position that included some pretty significant data analysis. “Do you know how to analyze data?” asked the interviewer. Of course I did. I knew how to read all the latest data visualizations, I knew the most popular and many, many lesser-known KPIs and metrics and how they are derived. I knew the difference between cause and correlation and how to look for them. “Yes, I do.” “OK, when you first access a database with millions of records, what do you do?”
There are lots of “correct” answers to this question—lots of different ways to familiarize yourself with the data. You could look at the fields and figure out what each one is. You could look at data sources to get a handle on the accuracy and consistency of the data. You could look at the database rules to understand most trusted sources, what determines a duplicate record, and how often records are imported, exported, and deleted. But none of these was the answer my interviewer was looking for. She thought you should do all of these things before you ever entered your password. What she wanted to know was exactly how I would start looking at the data.
Recently I came across an article by Stephen Few, visualization guru, consultant, and author of the most well-worn book on the shelf in my office, "Information Dashboard Design", as well as a number of other books. He suggests you get acquainted by viewing distributions, interconnections among variables, parts of the whole, rankings and correlations, changes through time, and locations. In his article, he recommends specific graphic formats for viewing each of the data set characteristics above—for instance, he recommends using histograms for viewing distributions and line graphs for getting an overview of changes through time. Nothing new there, but he also suggests using newer visualizations such as table lens displays for looking at rankings and correlations and treemaps for viewing parts of the whole. Heck, when I went on that interview, I didn’t even know what a table lens display was. I wish I had read Mr. Few’s article before I went on that interview. But then again, if I did, I wouldn’t be where I am today, so maybe it was for the best that I find it now so I can recommend it to you.
BIO business intelligence makes it easy for you to manipulate your data and look at it exactly how you want to see it. BIO includes a powerful back-end that does all the heavy lifting and a sophisticated but easy-to-use viewer that enables you to look at the data the way you need to see it and change your view at nearly the speed of thought. And BIO includes more “content” than other BI systems—predefined views, hierarchies, cubes, reports, graphs, calculations, and more that will save you hundreds and even thousands of hours of research and development when you’re looking to implement business intelligence in your organization. With BIO, you can get started right away and spend your time analyzing your data, not figuring out how to look at it.
Read Stephen Few's 24-page how-to article, “
Join BIO for a
By Sandi Forman of BIO Analytics, Corp.,