Stats Made Easy

Practical Tools for Effective Experimentation

Monday, July 31, 2006

Is something always better than nothing?

In last week's Parade magazine column, Marilyn vos Savant fields a question from a fellow whose friend invests effort into anything at all that provides the tiniest advantage.* Marilyn sums things up nicely with her observation that "Too many people spend time and money pursuing lots of goals just a little." I see this happening from time to time to clients of Stat-Ease who pore over the statistical analyses of their experiments and get hung up on tiny, but potentially signicant effects. The advent of robotic chemistry with systems such as Sagian AAO make it ever more likely that significant effects of no practical importance will be uncovered via well-designed experiments. The problem is that by focusing attention on these trivial things one loses sight of the really important opportunities that merit a full investment of their time and money. The question then is one of statistical significance versus practical importance.

*Ask Marilyn


Post a Comment

<< Home