In scientific experiments or tests that aim to establish something about living humans, statistical analyses are typically done at the 95% confidence level, which means that there’s a 5% chance (or one in twenty) that the results are spurious — i.e., that a false positive or a false negative was obtained (so-called

*Type I* or

*Type II* errors). For example, a single clinical trial of a new medication might show that it is effective above placebo when in fact it isn’t, or show that it is not effective when in fact it is. When such experiments are repeated independently by others in different settings using other test subjects and they show similar results, the overall confidence in that result increases. If different results are obtained, this indicates that more work is needed. (All of this leaves aside the thorny issue of publication bias, where scientists tend preferentially to publish results that show positive findings.)

So much for 95%.

Expressed as the simplest possible fraction, 96% is equal to 24/25. 25 = 5

^{2} is a perfect square. 24 = 2

^{3}×3, and all integers that are relatively prime to 24 (i.e. all integers that don’t have 2 or 3 as a divisor, the smallest of which is 5) leave a quadratic residue of 1 modulo 24. For example, 97

^{2} = 9,409, which leaves a remainder of 1 upon division by 24.

To find out whether the above is in any way relevant, you’ll have to consult your nearest numerologist.

No, I strongly suspect it’s mostly marketing hype and bluster. “96 per cent” sounds like a nice impressive number but not as blatantly contrived and self-important as “100 per cent.” Translated, it says, “Most people prefer our product. There is a small minority that has a mind of its own and chooses differently.” Part of the hype is that it obscures the foot-in-the-door effect: If you first offer someone a sample of your product and then ask them if they would buy it elicits a much higher positive response rate than if you just ask them out of the blue. Also,

*saying* that you will or might buy the offered product (whether you actually buy it or not) becomes a “preference” just as soon as the marketing magicians sprinkle their fairy dust on your statements. With such semantic sleight of mind, it’s hardly a wonder that competing products A and B are both preferred by 96% of consumers. It also means that they’ll be able to “justify” such preference claims if challenged.

Consider also that many of the products

*not* preferred by that 96% would not remain viable for very long because sheer weight of the allegedly preferred product would drive them out: Why would any business-savvy floor manager waste precious shelf space stocking a dog of a competing product when the preferred product really does turn over 24 times as fast?

'Luthon64