SDCAR2010【逻辑入门】(十)Flaw (part 2)
Common flaws (continued)
9. Overstating the evidence
• No evidence for a claim does not mean the claim is false.
• No evidence against a claim does not mean the claim is true.
• Some evidence for a claim does not mean it is true.
• Some evidence against a claim does not mean it is false.
10. Survey errors
A) Biased or unrepresentative samples
• When studying a certain groups, researchers must survey a random sample of that group or the survey results will be flawed.
• Researchers can err by choosing people who are likely to represent only a subset of the entire group or by letting people voluntarily respond. Either way, the sample is not random and thus unrepresentative.
• Example: Randomly polling subway riders all across the country to determine whom Americans will likely elect as President will probably not account for rural voters, who tend to vote differently than urban voters.
B) Biased questions
• Tempting: “Do you support reasonable reform?” encourages respondents to say they do because to say otherwise would be “unreasonable.”
• Misleading: “Do you think we should kill Osama bin Laden or stop searching for him?” suggests we have only two choices. (Well, now we do not need to worry about this question any more!)
• Confusing: “Do you have a higher elasticity for bread or cheese?” is really asking, “Are you more sensitive to a change in price for bread or cheese?” or, better yet, “Which food prices do you monitor more closely?”
C) Biased answers
• Example: In the 2008 Presidential primaries, polling critics questioned polls that showed Barack Obama leading Hillary Clinton. They argued that the respondents were likely overstating their support for Obama because they did not want to appear racist.
11. Generalizations
• The passage makes an unwarranted generalization based on only one or two examples.
• Flawed example: “Poor people will go from rags to riches. Felix Dennis did.”
• Sound example: “Poor people can go from rags to riches. Felix Dennis did.”
