Category Archives: Analytics

Stephen Few’s latest book, “Now you see it”

Stephen's latest book is a brilliant interplay of simple to advanced statistical concepts with powerful visual equivalents and methods. It offers wonderful guidance with great examples around visualizing time series, correlations and multivariate analysis problems. It will have a major impact for many years to come in changing how business analysts work, think and improve their companies!

A friend of mine recently asked "... are statisticians becoming irrelevant" (hey, that's me!) with the advent of advanced visualization tools like Tableau. I think the answer is yes and no. New tools like Tableau combined with methods expounded by Stephen will allow daily business questions to be answered rapidly. This should allow the statisticians to focus on tackling very high value problems where even a 1% improvement in the outcome is a huge win.

I have personally found that combining my work in data mining with beautiful stories about the results (created with Tableau) is truly impressive and informative for guiding decision-makers. Business decision-makers gain the advantages offered by advanced data mining algorithms such as decision trees, logistic regression and neural networks to identify primary predictive patterns and variables. The analyst can then clearly explain the results with compelling visuals to easily convey the findings and recommendations.

I think Stephen's latest book will become the "gold" standard for data exploration books to follow. If you need the next step past "Show Me the Numbers", this is it! BTW- it is even better than "Show Me the Numbers", which is indeed very impressive. I plan to incorporate some of this book into my advanced Tableau course, "Data Exploration and Elegant Dashboards with Tableau".

S&P Composite Stock Market Returns Dashboard

Finance-SnP-stock-market-returns-100-yearsEconomic Indicators and S&P Composite Returns
from 1901-2008
A Dashboard to Explore Past, Present and Future Results

When creating a dashboard, two critical questions to ask include, “what are the right metrics” and “what collection of graphs will optimize exploration and understanding from a complex relationship between multiple metrics”? This example uses historic economic and stock market data to explore these concepts and to provide powerful insights into stock market patterns relative to economic conditions. This interactive dashboard is powered by tableau public, a new service of Tableau Software. One means of interacting with this dashboard is data point brushing, highlight data in a part of the dashboard by single-clicking and dragging Continue reading

Visual Analytics Best Practices—avoid flash, keep it simple!

Why don’t you just show me the numbers?

  • Poor presentation of data can make informative interpretation difficult or even impossible.
  • We spend years diagramming sentences & solving equations but have minimal exposure to telling stories with data!
  • The ability to explore, understand and utilize real-world data is a rare skill that can dramatically improve your organization and career.

Sample revenue data from a university

College Quarter Total Revenue Tuition Research Plan Revenue
Arts 2009-01 $5,510,268 $5,230,198 $280,070 $6,000,000
Design 2009-01 $703,770 $408,762 $295,008 $680,000
Engineering 2009-01 $25,139,083 $6,201,932 $18,937,151 $26,800,000
Sciences 2009-01 $4,439,991 $2,159,010 $2,280,981 $4,400,000

 

A prickly example

While it is indeed looking sharp, this graph is uninformative.
Continue reading