Timeline for Word for "drawing a decisive conclusion about a phenomenon according to specific personal experience"
Current License: CC BY-SA 3.0
14 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 9, 2012 at 17:38 | comment | added | J.R. | @FumbleFingers: Exactly – and I think your latter example is a classic hasty generalization, although I don't claim to be an expert in logical fallacies (my initial introduction to the subject came from Shulman's Love is a Fallacy: "Cool was I and logical" – LOL). | |
| Dec 9, 2012 at 13:51 | comment | added | FumbleFingers | @J.R.: Apparently it can also be called the fallacy of converse accident, reverse accident, destroying the exception, or a dicto secundum quid ad dictum simpliciter. I've no idea which if any of those terms are considered "standard" by students of formal logic. But there's a difference between the kind of generalisation involved in saying "I don't like Mondays" (because on average I don't, over a lifetime) and "I don't like people from Timbuktu" (because I didn't like the only one I ever met). | |
| Dec 9, 2012 at 11:58 | comment | added | J.R. | @FumbleFingers: I thought "Hasty generalization" was the formal name of a logical fallacy, much like Post hoc ergo propter hoc, Argumentum ad populum, or Poisoning the well. Inclusion of the word hasty implies that the generalization occurs as part of a logical fallacy, as opposed to simple generalization. I think you can generalize without committing a fallacy, such as I don't like rainy days, which might be generally true, albeit with exceptions (such as on the day after I happened to plant five new trees in my yard). | |
| Dec 9, 2012 at 6:31 | comment | added | user21497 | @S: Yes, the second guy's looking for consistency. Reproducibility is perhaps possible (not necessarily probable, though) with two study populations of 10,000 patients, but not with two athletes or even 100 athletes who start out with the same stats in their first 79 at-bats: each is a sample of only 1. Take 1000 players after 1000 at-bats, & you can generally predict how many will consistently have a high or a low OPS, but you can't tell about any one specific player. | |
| Dec 9, 2012 at 6:20 | comment | added | StoneyB on hiatus | @BillFranke "I think Kozma's our starter. .952 OPS!" "SSS, dude; that's in 79 plate appearances. Talk to me when he's got 600 under his belt." | |
| Dec 9, 2012 at 6:18 | comment | added | FumbleFingers | ...and the coin-flipping is a terrible example. If you point a gun at two people one after the other, pulling the trigger each time, what do you conclude about the dangers of playing with guns? More accurately, if an alien you've never seen before suddenly appears and eats two of your friends, what do you conclude about its intentions? Examples where you already know a lot about the situation aren't really relevant to what we're talking about here. | |
| Dec 9, 2012 at 6:15 | comment | added | FumbleFingers | @Bill: You're being somewhat literal-minded there. My bias (as I thought I'd laboriously explained in that comment) is against the word being used indiscriminately for both positive and negative contexts. Obviously I have much the same cognitive faculties as all human beings. Well, if I'm honest, I truly believe I have more ability than most to revise my opinions as more evidence becomes available, but that's not really the point. | |
| Dec 9, 2012 at 6:10 | comment | added | user21497 | @StoneyB: SSS is universally condemned in statistical analysis because statistical power comes, in part, from a sufficiently large sample, which reduces the possibility that the data are merely flukes. If you flip a coin twice & it comes up heads, what can you conclude about coin flipping? Nothing. Flip it a million times, and the chances are significantly better that half will be heads & half tails. However, that's got nil to do with how athletes should be rated: some are flashes in the pan & others are finally consistent, even if their debut performances are spectacularly good or bad. | |
| Dec 9, 2012 at 6:00 | comment | added | user21497 | Your bias against generalizing is shortsighted, I think. We do that all the time. Without generalizations, We'd need to take the time to analyze & reanalyze the same or similar conditions over & over & then arrive at the same inferences over & over. If you see a cobra bite a man who then dies, you immediately generalize that cobras are dangerous & you infer that you must avoid them: next time you see a cobra, you'll avoid it: evolutionary survival value. Overgeneralization is a different thing, but "first impressions are lasting impressions": that's human nature. | |
| Dec 9, 2012 at 4:10 | comment | added | FumbleFingers | @J.R.: Personally, I don't normally even bother with the "hasty" bit. I tend to use words like "universalise/generic" when I have positive connotations in mind. Leaving "generalise" to be almost exclusively a negative term - not that there's really any scope for confusion if I say "You can't generalise", or "You're just generalising". It ain't never a good thing to me. | |
| Dec 9, 2012 at 3:45 | comment | added | J.R. | Hasty generalization was the first thing that popped into my mind. | |
| Dec 9, 2012 at 3:31 | comment | added | FumbleFingers | @StoneyB: Well, I don't know exactly how common it is across the world at large, but "based on a sample size of one" is far from unknown in my neck of the woods. I notice several biology/human origins/SETI contexts in that link - they're certainly typical contexts where I use/encounter it. | |
| Dec 9, 2012 at 2:13 | comment | added | StoneyB on hiatus | +1 On the baseball site I frequent the catchphrase is SSS, for Small Sample Size, employed to condemn evaluating a player based on his first few weeks in the majors. | |
| Dec 8, 2012 at 23:43 | history | answered | FumbleFingers | CC BY-SA 3.0 |