statistical vs anecdotal evidence

statistical vs anecdotal evidence

Posted Dec 21, 2007 17:32 UTC (Fri) by giraffedata (subscriber, #1954)
In reply to: Insufficiently free? by felixfix
Parent article: Insufficiently free?

The thing is: there's no anecdote here. An anecdote would be, "my neighbor switched to OSX because it can run free software."

What TRS-80 started with is an argument in identical form to RMS's. RMS said "people usally don't ..." and TRS-80 said, "a lot of people ..."

Statistically speaking, comparing one statistical statement to another, TRS-80's claim doesn't disprove RMS's because the number of people interested in OSX may be too small fraction of everyone. BUT: it's an example, and one more than RMS gave.

Since no statistical evidence has been presented on RMS's side, I think refuting with examples is a perfectly valid approach.

"People won't usually buy unripe fruit."

"What about bananas? Those continue to ripen after you buy them, so people who don't want overripe bananas a week later buy unripe bananas."

"OK, I didn't think of that. So it's not unusual to buy unripe fruit"

---

(Log in to post comments)

You are oh so wrong.

Here is what he said that I was responding to.

"This can be shown false by example - a lot of people desire ..."

An example does not disprove a statistical assertion.  His example may say " a lot of ..." and
be a statistical assertion itself, but it does not disprove RMS's statistical assertion.