PAGES: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Fallacies from Chapter 5 - 5469
|
|
|
Fallacy
Questionable Statistics
Short Definition
doubtful statistics
Long Definition
Employing statistics that are false, misleading, biased, unknowable, or irrelevant without further support.
Form
none
Keywords
numbers, percents
How to Explain
State why statistics are doubtful
Example
Some studies show that humans only utilize about 10% of their brain
|
1. Statistics are quantities or percentages; for example, the U.S. minimum wage is $5.15 per hour or 60% of the students attending SRJC are female. Statistics are derived from surveys, polls, studies, experiments, and various other means. They are calculated by government, corporations, academic institutions, and various other groups and individuals. They are important facts which dictate the making of many significant decisions.
2. The fallacy of questionable statistics is employing statistics that are questionable without further evidence. Many statistics are knowable in theory, but just inaccurate. They may be educated approximations or exaggerations and may be used to support a hidden agenda. Many government statistics are inaccurate at the time of posting as they are continually revised later; for example, the unemployment rate or the number of jobs created each month. For example in 1992, a Government jobs survey reported that California gained 15,000 jobs. This figure later turned out to be very inaccurate. After tax receipts were analyzed, it turned out that California lost 300,000 jobs.
3. It is important to check the source of the statistic. Is this source an expert on the issue? Are they reliable? Do they have a reason to fudge the numbers? What is their track record on such matters?
4. A type of questionable statistics is unknowable statistics. These statistics are not even knowable in theory. Some of them apply to the distant past or the future; for example the average person will live to 138 by the year 2104. Unknowable statistics are usually too precise and may seem to be overly authoritative. They are generally guessed at or made up, with insufficient or questionable evidence.
5. You should avoid the tendency to blindly accept statistics that reinforce your beliefs while questioning statistics that do just the opposite.
6. When you think that a statistic is doubtful then you should do research to try to verify or refute the statistic. Use of the Internet can greatly help in this regard.
|
1. Some past psychological studies have shown that humans only utilize about 10% of their brain.
Analysis: 10% is a questionable statistic. How do they know it's 10%? Who conducted the studies? Supporting evidence is needed here.
2. A new study shows that kids watch an average 4,286 acts of violence by age 18.
Analysis: These statistics are questionable, in fact unknowable given how exact they are. Who conducted the study? Supporting evidence is needed here.
3. A new sociology study show that the average person lies 3 times a week.
Analysis: Questionable statistic since this is most probably unknowable. Who conducted the study? Supporting evidence is needed here.
4. A seemingly reputable Bay Area poll taken before the first Gulf War stated that 91% of the people in the Bay Area favored this war.
Analysis: This turned out to be questionable after I discovered that the sample was biased. The poll's sample greatly over-represented the East Bay and under-represented San Francisco.
5. State Dept. Spokesperson: "99% of Somalians want the U.S. to be there."
Analysis: This was a very questionable statistic and perhaps unknowable. What happened to the U.S. military in Somalia seems to have shown that far less of a percentage of Somalians wanted us there.
6. An earthquake study by the U.S. Geological Department predicted that if there was a 7.5 earthquake on the Hayward Fault at 2:30 pm on a weekday then approximately 4,000 people would die.
Analysis: Questionable statistics. That approximately 4,000 people would die in that circumstance seems to be an unknowable statistic. No evidence was given by the U.S. Geological Department on how they arrived at their findings.
7. Center for Disease Control: "Conditions at hospitals kill nearly 90,000 Americans a year, mostly from infections they got in the hospital.
Analysis: No fallacy - the statistic is accurate. Approximately two million people catch infections from going to hospitals. The main reason is that hospital personnel don't wash their hands regularly, especially with alcohol hand rubs. If they did then they say it would cut infections by 50%. Note that hospitals are not required to report their infection rates.
8. New England Journal of Medicine study: "Guns in the home for protection are 43 times more likely to kill a family member than an assailant.
Analysis: No fallacy - the statistic is correct and the source of the study is legitimate. Nevertheless, I don't believe many gun owners will be dissuaded from having a gun in their house for protection. They may argue that the threat of having a gun the house may prevent some assailants from even trying to enter a house. But how would an assailant even surmise that a particular house has a gun owner inside?
9. Scientist: "I am 80% sure that global warming is a problem."
Analysis: Questionable statistic of 80%. The scientist admitted later that he had created a fuzzy statistic and was hoping no one would ask him to support the exact choice of 80%.
CATE: Computer-Assisted Teaching Environment
Distance Education office at Santa Rosa Junior College, Santa Rosa, CA USA
Last updated: 15:23 on 30 January 2013
Copyright © Steve RubinContact Steve Rubin
Email Webmaster