What is statistics and why should we care?
2013 is the International Year of Statistics. SAS co-founder John Sall explains why we’re celebrating.
Statistics are numbers, but the practice of statistics is the craft of measuring imperfect knowledge. The field of statistics is about uncertainty. Although we are comfortable with uncertainty in ordinary daily life, we are not used to embracing it in our knowledge of the world. We think of knowledge in terms of hard facts and solid logic, though much of our most valuable real knowledge is far from solid.
The business of statistics
Why should we care? Because business, world economies, scientific discoveries and medical breakthroughs are all based, at their core, on statistics. The discipline of statistics provides the framework of balance sheets and income statements for both business and scientific knowledge. Statistics is an accounting discipline, but instead of only accounting for money, it also accounts for scientific credibility. It is the craft of weighing and balancing observational evidence.
Scientific conclusions are based on experimental data in the presence of uncertainty, and statistics is the mechanism to judge the merit of those conclusions. The statistical tests are like credibility audits. Of course, you can juggle the books and sometimes make poor science look better than it is. However, there are important phenomena that you just can't uncover without statistics.
The joy of discovery
Why are we celebrating? One reason is the special joy in using statistics as a tool for discovering new phenomena.
There are many views of your data — the more perspectives you have on your data, the more likely you are to find out something new. Statistics as a discovery tool is the auditing process that unveils phenomena that are not anticipated by a scientific model and are unseen with a straightforward analysis.
Statistics fits models, weighs evidence, helps identify patterns in data and then finds data points that don't fit the patterns. Statistics is induction from experience; it is there to keep score on the evidence that supports scientific models.
Statistics is the science of uncertainty, credibility accounting, measurement, truth-craft, the stain you apply to your data to reveal the hidden structure, the sleuthing tool of a scientific detective.
Statistics improves our world
A hundred and fifty years ago, most of us were farmers, and it was a struggle to feed the world. Statistics was one of the keys to agricultural research that led to improvements in agricultural productivity that changed that. Intensive productive crops on less land feeds more people using much less labor, enabling the world to be urban.
More recently, our lives have changed due to the enabling effects of the microelectronics revolution. But in an engineering field measured in nanometers, getting an economic yield of reliable components took the power of statistics used in designed experiments to surmount the challenges.
Very recently, we have begun to discover the complex workings of life itself, through molecular biology and genomics. Sorting through this complexity will only be possible through statistical methods.
Not long ago we saw a bump on the side of a hill. The hill was data from the Large Hadron Collider that cost $6 billion to pile up. The bump was seeing, for the first time, the Higgs particle, the biggest discovery of the 21st century so far. Statistics made seeing that bump possible.
Bio: John Sall is a co-founder and Executive Vice President of SAS. He leads the JMP business division, which creates software that links robust statistics with interactive graphics, in memory and on the desktop.