Analyzing Data for Good Outputs

Data analysis has evolved from expensive, incomplete, and time-consuming censuses in the ancient world, to the real-time analysis of minutely detailed business process information. In the future, I expect this iterative loop to turn faster and faster, further incorporating #ML, automated task analysis, and big-picture strategic overviews. I remember my first real exposure to data analysis was in Grad school (Case Western Reserve University, many moons ago). Back then, we used Lotus 1-2-3 and Quattro Pro for data analysis and visualization. It was time consuming, tedious, and error-prone but served our purpose. With so many data analysis tools and techniques available, the real question is “what are the qualities of a good analytical output?”. Here is what I have learned from my experiences –

• It is built off of good, reliable data
• It has clear labels and legends
• It is simple and intuitive to understand
• It provokes constructive questions
• It provides actionable insights
• It tells a story (visual outputs are usually more impactful)
• Most important of them all, it is communicated in a timely manner, to the right audience

If you are into data analysis AND movies, check out Moneyball and Big Short. You will see some excellent examples of how to use data analysis to gain competitive advantage. BTW, the picture is from the data analysis work done by Abraham Wald, a Hungarian mathematician who worked for the United States (U.S.) during World War II. His contribution to this list of great data analysis? Not falling for what we now call “survivorship bias”. Wald is credited with “saving the lives of countless flight crews who likely would have been shot down. Google him! #dataanalysis #visualization #insightsanddata #worldwar2 #dod #usarmy #usairforce #usnavy #analytics #gooddata