Makeup work: For any problem (#1 thru #11) on the exam where
you lost 3 or more points, you can earn back up to half the
lost points by doing this: Study the solutions, determine
what
I was testing for, write a new problem that tests that, and
solve
it. Give me your old exam and your new problems and solutions
by class Wednesday, April 17, Day 32. You may ask for help
from
anyone including me, but must give credit on your paper to those
people.
(If you only missed a chunk of a several part problem, you don't need
to redo the parts you did successfully.)
If you want to redo the Placebo
effect problem, answer these:
1)
According to the article, what do scientists think is the mechanism that
causes the Placebo Effect?
2)
Give one example of the Placebo effect from the article.
Comments:
To do well, you had to have done the Activstats, the reading on
reserve, the text, and know what was added in class.
You also had to do careful close reading.
#4 Placebo effect. This was given as it was specifically
because a large portion of the class hadn't read the article on reserve,
on the two days I asked about it and urged the reading of it.
The placebo effect is a genuine physiological effect--people do get
well, or at least better, from placebos. I took off if you implied
that people only thought they felt better.
#5 Late bedtime causes bad stress-hormone response?
To test this you have to impose a treatment, consisting of a forced
bedtime, and see what the resulting hormone patterns are are after stress
is applied.
explanatory variable: bedtime
response variable: hormone pattern in response to stress
A complexity is natural bedtime--it may not be easy to go to sleep
early if you're a night owl, or to stay up late if you're not.
I accepted any design that randomly assigned individuals to different
bedtimes, and afterward administered stressful tasks and measured cortisol
levels.
#6 a) Literary Digest Poll .
I've given this problem before, when it was not discussed in
class, and not available on Activstats. I was disappointed to find
that people this year didn't seem to do better on it, even though I specifically
asked you to do this activity.
b) Here Gallup is not interested in predicting the election.
He's interested in predicting the Digest poll result, with a tiny subsample
of their huge sample. (This was not in Activstats. It was just
an exercise in careful reading.)
Full solutions on reserve + outside my door.
If you still have questions after looking at the solutions, please come see me! And come and see me anyway, if you got lower than 70.
| Sievers home | Math151-Sp02/Exam2comments.htm | 4/8/02 |