3 ms·

Title is silly. The wonderful thing about statistics is that if you have a truly random sample of 1176 people you can extrapolate from 136 people to seven milli

Title is silly. The wonderful thing about statistics is that if you have a truly random sample of 1176 people you can extrapolate from 136 people to seven million (plus or minus a certain error bar which I'm too lazy to figure out right now). The other points are somewhat valid, but by the usual standards of political misuse of statistics this is pretty small beer.

ajross 17y ago
Yeah, I was immediately screaming the same thing. Arguing that "only" a tiny subset of a population is involved with a survey as a basis for rejecting its results is just plain ignorance. The other stuff in the article does seem dodgy though, like arbitrarily tacking on 50% to reflect an assumed-but-unmeasured bias in the input sample. Ridiculous. Still, the title is innumerate, sensationalist junk, and the poster should be ashamed.
- dkokelley 17y ago
  The title was actually copied directly from the article. I completely agree that the title appears to be linkbait, but the blame should be placed on the editor and/or author of the post, not the poster.
  - pbhjpbhj 17y ago
    The editor/author created the title to be sensationlist on purpose. HN is, I hope, not here to offer sensational headlines merely to grab extra viewers but as a way to share knowledge, information and wisdom. We should have higher standards for titles here IMO. What I'd like to see is a subtitle (or tagging) system with a descriptive subtitle that can't be applied by the poster only by someone else. Poor subtitles would be downmodded too.
sharpn 17y ago
I agree a small truly random sample can often be reliably extrapolated - but in this case the 136>7m figure was based on research paid for by the BMI (who lobbied heavily against file-sharing) in a report still not fully published, based on two numbers of which one differs from the official ONS number without explanation. The fact that the lobbyist organisation paid for the original report was obscured by misattributing the source, which also seems dubious to me. I don't doubt that political misuse of statistics is common, but it should always be called out.

ajross 17y ago

Yeah, I was immediately screaming the same thing. Arguing that "only" a tiny subset of a population is involved with a survey as a basis for rejecting its results is just plain ignorance. The other stuff in the article does seem dodgy though, like arbitrarily tacking on 50% to reflect an assumed-but-unmeasured bias in the input sample. Ridiculous. Still, the title is innumerate, sensationalist junk, and the poster should be ashamed.

dkokelley 17y ago
The title was actually copied directly from the article. I completely agree that the title appears to be linkbait, but the blame should be placed on the editor and/or author of the post, not the poster.
- pbhjpbhj 17y ago
  The editor/author created the title to be sensationlist on purpose. HN is, I hope, not here to offer sensational headlines merely to grab extra viewers but as a way to share knowledge, information and wisdom. We should have higher standards for titles here IMO. What I'd like to see is a subtitle (or tagging) system with a descriptive subtitle that can't be applied by the poster only by someone else. Poor subtitles would be downmodded too.

sharpn 17y ago

I agree a small truly random sample can often be reliably extrapolated - but in this case the 136>7m figure was based on research paid for by the BMI (who lobbied heavily against file-sharing) in a report still not fully published, based on two numbers of which one differs from the official ONS number without explanation. The fact that the lobbyist organisation paid for the original report was obscured by misattributing the source, which also seems dubious to me. I don't doubt that political misuse of statistics is common, but it should always be called out.