I. a. Imagine that you designed an experiment with 4 conditions. How many pairwise comparisons are there in your desi
Posted: Thu May 05, 2022 8:25 pm
I.
a. Imagine that you designed an experiment
with 4 conditions. How many pairwise comparisons are there in your
design? (1 point)
b. Imagine that you will test all of the
pairwise comparisons in this design. What is the probability that
your experiment will contain at least one type I error if you set
the alpha criterion to .05 and you do not perform any corrections?
(2 points)
c. Now imagine that you decide to use the Dunn
correction (a.k.a. Bonferroni correction) to control the
experiment-wise error rate. You decide that you want to hold the
experiment-wise error rate to .05. What should you set your alpha
criterion to? (2 points)
d. Here are 3 of the p-values that you obtain
from your pairwise comparisons: .0063, .0415, .0123 (this is not
all of the p-values, that would give away the answer above!). Based
on the alpha criterion that you calculated in (c), which of these
p-values would be statistically significant? (3 points)
a. Imagine that you designed an experiment
with 4 conditions. How many pairwise comparisons are there in your
design? (1 point)
b. Imagine that you will test all of the
pairwise comparisons in this design. What is the probability that
your experiment will contain at least one type I error if you set
the alpha criterion to .05 and you do not perform any corrections?
(2 points)
c. Now imagine that you decide to use the Dunn
correction (a.k.a. Bonferroni correction) to control the
experiment-wise error rate. You decide that you want to hold the
experiment-wise error rate to .05. What should you set your alpha
criterion to? (2 points)
d. Here are 3 of the p-values that you obtain
from your pairwise comparisons: .0063, .0415, .0123 (this is not
all of the p-values, that would give away the answer above!). Based
on the alpha criterion that you calculated in (c), which of these
p-values would be statistically significant? (3 points)