Deck 13: Putting Your Knowledge to Work: 20 Methodology Problems

Question 1

I'm Speechless

Answer 1

The first problem with this claim is that the researcher concluded that,unlike people,chimpanzees do not have the ability to learn and use language.It is possible that the researcher's data showed this,but before drawing this conclusion,I'd want to know not only whether people learned the language better than chimps but whether the chimps learned it at all.If the people were constructing the artificial language equivalents of sonnets and encyclopedias but the chimps were constructing the equivalent of some pretty good Haikus or animal rights pamphlets,I'd want to give the chimps a little more credit than the researchers did.

Question 2

The Early Bird Gets the Win?

Answer 2

My coach's conclusion about take-downs suffered from an interesting - and pretty dramatic - example of the third variable problem.Specifically,I assume that certain kinds of wrestlers (wrestlers with superior skills across the board,wrestlers of exceptional quickness,agility,or physical strength)are exactly the kind of wrestlers who are likely to get the take-down at the beginning of a match.(In wrestling circles,they have a phrase for these wrestlers.They're called "good wrestlers". )On the other hand,it may be disproportionately important to get the take-down in a wrestling match (e.g. ,it may set the tone for the match or "psych out" one's opponent).To figure this out scientifically,we would need to keep careful records of all points scored during a series of many different wrestling matches,also noting the winner and loser of each match.We could then find out (perhaps by conducting a multiple regression analysis)whether successfully executing some kinds of moves (a take-down,an escape,a reversal,a near-pin,etc. )rather than others better predicted success (winning versus losing a match).

Question 3

He May Be Small But He's Slow

Answer 3

The major problem in this study has to do with the fact that the researcher conducted 50 different statistical tests,and adopted an alpha level of .05 for all 50 tests.If there were absolutely no relation between similarity and marital happiness,and we set alpha at .05,we'd expect to see that about 2½ out of every 50 (i.e. ,about 5%)statistical tests we ran proved to be "significant" - because that's what an alpha of .05 means.

Question 4

Let's Get Supernatural.

Question 5

To Thine Own Selves Be True

Question 6

In Search of a Delicious,Low-fat TV Show.

Question 7

Clever Who?

Question 8

Working Your Fingers to the Dean's List

Question 9

Impressive Pickup Lines.

Question 10

Old Geniuses Never Die Young?

Question 11

Fly Away Home.

Question 12

Sampling Student Opinion

Question 13

Life Sucks and So You Die

Question 14

A Rosy Mood by Any Other Name?

Question 15

On the Drawbacks of Liking Yourself

Question 16

Testosterone Makes Better Dive Bombers