2,002
edits
Changes
Jump to navigation
Jump to search
Line 523:
Line 523: + + +
Talk:Examples of Bias in Wikipedia (view source)
Revision as of 10:39, March 11, 2007
, 10:39, March 11, 2007→Wikipedia six times more liberal than public?
:: Then what you should be saying is that Wikipedia is six times less conservative, not six times more liberal. As you pointed out yesterday, there is a major difference. Plus, if you're going to make extraordinary claims, you need to back them up with actual numbers, not a hand-waving argument. If you do the appropriate substitutions into the equation above, you'll find that C1/C2 = 1.573/(1-M2). For this to be equal to 6 would require that M2=0.73, i.e., about 3/4ths of Wikipedians would have to be moderates.--[[User:Manchot|Manchot]] 22:21, 10 March 2007 (EST)
:: Then what you should be saying is that Wikipedia is six times less conservative, not six times more liberal. As you pointed out yesterday, there is a major difference. Plus, if you're going to make extraordinary claims, you need to back them up with actual numbers, not a hand-waving argument. If you do the appropriate substitutions into the equation above, you'll find that C1/C2 = 1.573/(1-M2). For this to be equal to 6 would require that M2=0.73, i.e., about 3/4ths of Wikipedians would have to be moderates.--[[User:Manchot|Manchot]] 22:21, 10 March 2007 (EST)
This calculation is indeed deeply flawed. If you keep it like this you run the risk that liberal will view it as confirmation that conservatives can't do math. You compare a representative poll with the areas of interests of Wikipedians. The results of the poll are that 18% of Americans are liberal, 34 % conservatives, and 40% moderates. The Wikipedia numbers are that from 3.7 million Wikipedians, 200 are interested in liberal themes and 75 in conservative themes. This means that less than 0.0075 % of all Wikipedians are either conservative or liberal (or both). And the chances that this sample is representative is next to nothing. You can't do statistics with this sample. You simply compare apples with next to nothing. I'd suggest to remove this entry altogether. Otherwise you just confirm the liberal idea that conservatives fudge numbers as they please. [[User:Order|Order]]