Documente Academic
Documente Profesional
Documente Cultură
A. Unjustified Assumptions
B. Causation Errors
When an argument jumps to a causal relationship, question that relationship. Couldn’t it be the other
way around?
Correlation/Causation errors often have to do with changes or events that occur at the same time or in
sequence. Don’t assume that simultaneous events are necessarily connected. Likewise, if X happened
shortly after Y, you cannot necessarily conclude that X was caused by Y.
C. Comparison Errors
Unrepresentative Sample
Marketers, pollsters, and social scientists of all stripes use samples. Its impossible to ask everyone in
the entire population for their opinions on single- versus double-ply toilet paper, so instead you ask 100
people. You have to ensure that the sample is representative, though. In particular, you have to be wary
of volunteers, as noted above.
Survivor Bias
It is not logical to judge an entire group by concentrating only on who or what survived a process or
time period, while ignoring the non-survivors. Its easy to fall into this trap, though; after all, its often
hard to find out much about the people or things that didn’t make it!
Ever-Changing Pool
Many groups of people have a rotating cast of members. If a civic club voted in favor of something
yesterday
and against it 20 years ago, you wouldn’t automatically conclude that people in the club changed
their minds over time; its pretty likely that the club includes different people than it did back then.
D. Math Errors