Statistics

I really enjoy math, always have. Something that intrigues me is quantifying and drawing out conclusions from sophisticated statistical models. The key flaw is that, yes Virginia, you can lie with numbers. It's not the math to blame of course. It's the selective attention of the person crunching the numbers. What one selects to measure is much more important than how. And the power of the conclusions drawn is only as good as the understanding of the models, measures, and data used.

It's too easy to slip through invalid data for the simple fact that too many people do not understand the math and the reality that most people do not have the time to go back and thoroughly check another person's work too closely. Moreover, the data does not exist in a vacuum. Data arises from a date in time, at a place, etc. The environment around the sample is as important as the samples themselves. Quantitative analysis looks backwards and struggles with looking forwards for these reasons.

I suppose the prompt for this is an article I read about a robust statistically based system to help eliminate wasteful practices in the workplace. A sinking ship can as smoothly as possible, but if it does not see the iceberg ahead or faces a devastating hurricane on its present course, all the grunt work is for nothing. A good business needs to be more than looking down, looking for errors and flaws. The more important goal is to appeal more than the competitor. Ignoring competition, ignoring the future, ignoring the marketplace trends, these are keys to a former success gone awry.