University of Washington

BIOL 180

. [14 pts] Last January Mustafa Chopan & Benjamin Littenberg published a study (doi: 10.1371/ journal.pone.
0169876) for which they had tracked 16,000 Americans for an average of 17 years. Each participant reported at
enrollment whether they ate hot chili peppers at least once a month. The death rate among participants who ate
hot peppers was 21.6%; the death rate among participants who did not eat hot peppers was 33.6% (P = 0.01).
a) What hypothesis does this study suggest?
b) What is the corresponding null hypothesis?
c) Identify a key shortcoming in the study that prevents us from drawing definitive conclusions about these
hypotheses, and explain:
Design a better study to test the hypothesis you stated in (a):
d) What is your independent variable?
e) Describe your treatment groups:
f) What is your dependent (response) variable?
g) Name one (and only one) variable you will need to control for:
h) How will you control for the variable you named in (g)?
i) Draw graphs predicting the results of
your study if…

The hypothesis is correct: The null hypothesis is correct:
Eating hot chili peppers increases longevity.
Eating hot chili peppers has no effect on longevity.
Possibilites include:
* Other uncontrolled differences between subjects who ate vs. did not eat hot chili peppers.
* The subjects knew which group they were in.
Consumption of hot chili peppers
EXPERIMENTAL: Swallow daily pill containing ground hot chilies; CONTROL: Swallow daily sham pill.
Possibilities: Death rate; Longevity; physiological indicators of health.
Possibilities: Age, baseline health, exercise, smoking, knowledge of which group they’re in, etc.
Best answers:
Most variables can be controlled by RANDOM assignment of a LARGE NUMBER of subjects.
Knowledge of treatment group can be controlled with sham pills.
Partial-credit answers:
Careful matching of treatments and controls for age, health, exercise, smoking…
Death rate
Chilies Sham
Exptl. Cntrl
D