I'll confess up front: I had to use a slide rule when I was an undergraduate engineer, but when the scientific calculators came out (Hewlett-Packard HP-35 and Texas Instruments SR-50) I switched in a flash. And why not? The slide rule was a pain for any sort of complex calculation, it didn't do addition and subtraction, and it had resolution to only three figures. Despite my personal interest in scientific instruments, I don't feel any actual nostalgia for the slide rule at all.

But the article points out that the slide rule's limitation was also its virtue: it forced engineers to estimate the answer in advance, to determine where the decimal point would be. Using a slide rule gave me good training at doing quick estimates that would be within 20 or 30% of the final answer. It was a good sanity check on the credibility of a calculation: was that resistor we needed 1.23 ohms, 1.23 kohms, or a 1.23-Mohms?

Unfortunately, modern calculating devices take away from users the need to stop and think, especially lazy or complacent users. I see too many analyses and press releases with numbers that are way out of line, or with excess precision. Is it meaningful to sample several hundred respondents about the size of a market in 5 years, and then analyze the results to four places?

No, for two reasons. First, the data comes from a limited sample size. Second, the very act of predicting the future is a blurry one, and using statistics to analyze the data puts a false veneer of credibility on the numbers. Saying that 47.35% of respondents said this or that is misleading.

I always keep in mind a quote from Alan Greenspan, venerated (now former) head of the Federal Reserve System for many years. When commenting on some economic data and trends, he said, "I'd rather be roughly right than precisely wrong." That's a message we should all keep in mind as we toss those high-precision numbers around so effortlessly. Take the time, make the effort to first do an estimate or even a SWAG (scientific wild-ass guess) and you'll avoid big, embarrassing mistakes that are nonetheless precise.

Bill Schweber, Site Editor, Planet Analog
bschweber@cmp.com