que--epidem - targt250

[ArchivalUser]

Investigators in a multisite trial are attempting
to determine how many patients are needed
before they break the blinding on their study
and analyze their data. They want to ensure
that there are enough patients to detect a difference
between treatment and placebo if one
is actually there. It is a very costly trial, and the
study sponsors do not want to pay for more
work than is needed to prove a difference.
What calculation do the researchers need to
perform to determine how much funding to request
from their sponsoring agency?
(A) Confi dence interval
(B) P value
© Power
(D) Type I error
(E) Type II error

[ArchivalUser]

e?

[ArchivalUser]

C)Power as power depends on sample size

[ArchivalUser]

The correct answer is C. Power is the probability
that a study will fi nd a statistically signifi -
cant difference when one is actually there. It
relates directly to the number of subjects. A
power calculation is required in many grant
applications.