2024 NSG 6001 EXAM PREP WITH
CORRECT ANSWERS
Which of the following changes will decrease the length of a confidence
interval (make it narrower), holding all else constant? - CORRECT-
ANSWERSsmaller standard deviation
Which of the following changes will increase the length of a confidence
interval (make it wider), holding all else constant? - CORRECT-
ANSWERSlarger confidence level
Which of the following changes will decrease the length of a confidence
interval (make it narrower), holding all else constant? - CORRECT-
ANSWERSlarger sample size
A surgeon wonders what the average time it takes her to perform a
particular type of surgery is. She takes a random sample of past surgeries
and records the length of the surgery in minutes. If she wants to test
whether the average time it takes him is 120 minutes, what is the most
appropriate set of null and alternative hypotheses? - CORRECT-ANSWERSH0:
u = 120 versus Ha: u not equal to 120
In a sample of 250 people, 11.6% of them have diabetes. A one-sample z-
test (Wald test) of whether the true proportion of people in the population
who have diabetes is equal to 9% yields a p-value of 0.15. What is the
appropriate conclusion to draw? (Assume the usual p-value cutpoint of
α=0.05.) - CORRECT-ANSWERSThe percentage of people with diabetes in the
population is not significantly different from 9%.
When we conduct a one-sample z-test (Wald test) for a proportion, we are
making all of the following assumptions except: - CORRECT-ANSWERSThe
data values come from a normal distribution
We might want to use the Exact (Binomial) test instead of the z-test (Wald
test) for a proportion when: - CORRECT-ANSWERSthe sample size is too small
The one-sample t-test relies on all the following quantities except: -
CORRECT-ANSWERScritical value from the normal distribution
A large study was conducted to assess cardiovascular health of cancer
survivors. A sample of 180 survivors had their HDL cholesterol ("good
, cholesterol") measured, and a 95% CI for the mean HDL in the population
was calculated to be (51.8, 56.2). If the same data were used to test whether
the mean HDL was equal to 50 (with a t-test), the resulting p-value and
conclusion would be: - CORRECT-ANSWERSp < 0.05, reject the null
A Type II error is when: - CORRECT-ANSWERSThe null hypothesis is false, and
you fail to reject it.
Suppose the true proportion of dogs who have symptoms from seasonal
allergies is 40%. I gathered a small random sample of dogs and assessed
them for allergy symptoms. Using my sample data, I tested the following
hypothesis: H0: p = 0.40Ha: p ≠ 0.40(p = proportion of dogs who have
seasonal allergy symptoms) I got a p-value of 0.004 and therefore rejected
the null. Which of the following has happened? - CORRECT-ANSWERSA Type I
error
Which of the following changes to a study design will result in lower power
for a one-sample t-test? - CORRECT-ANSWERSIncreased standard deviation
of the variable being tested
Jaylen is planning a new research study and did a sample size calculation.
They concluded that they needed to enroll 50 subjects to have 90% power to
detect the hypothesized effect size (with 5% type 1 error). Jaylen realized
they are actually interested in a smaller effect size. That would require: -
CORRECT-ANSWERSEnrolling more than 50 subjects
In a sample of 250 people, 11.6% of them have diabetes. A one-sample z-
test (Wald test) of whether the true proportion of people in the population
who have diabetes is equal to 9% yields a p-value of 0.15. What is the
appropriate conclusion to draw? (Assume the usual p-value cutpoint of
α=0.05.) - CORRECT-ANSWERSThe percentage of people with diabetes in the
population is not significantly different from 9%.
A one-sample t-test is conducted on blood glucose measurements in a
sample of 15 cats. When testing whether the mean glucose is greater than
200 mg/dL, the appropriate reference distribution to use for calculating the
p-value is: - CORRECT-ANSWERSt distribution with 14 degrees of freedom
A small study measured the body mass index (BMI) of 30 yoga practitioners.
A one-sample t-test was conducted to test whether the mean BMI (in the
population) is equal to 25 with a resulting p-value of 0.18. If we constructed
a 95% confidence interval for the mean BMI using these data: - CORRECT-
ANSWERSThe CI would contain the value 25
Statistical power is... - CORRECT-ANSWERSThe probability you (correctly)
reject the null when the alternative is true.
CORRECT ANSWERS
Which of the following changes will decrease the length of a confidence
interval (make it narrower), holding all else constant? - CORRECT-
ANSWERSsmaller standard deviation
Which of the following changes will increase the length of a confidence
interval (make it wider), holding all else constant? - CORRECT-
ANSWERSlarger confidence level
Which of the following changes will decrease the length of a confidence
interval (make it narrower), holding all else constant? - CORRECT-
ANSWERSlarger sample size
A surgeon wonders what the average time it takes her to perform a
particular type of surgery is. She takes a random sample of past surgeries
and records the length of the surgery in minutes. If she wants to test
whether the average time it takes him is 120 minutes, what is the most
appropriate set of null and alternative hypotheses? - CORRECT-ANSWERSH0:
u = 120 versus Ha: u not equal to 120
In a sample of 250 people, 11.6% of them have diabetes. A one-sample z-
test (Wald test) of whether the true proportion of people in the population
who have diabetes is equal to 9% yields a p-value of 0.15. What is the
appropriate conclusion to draw? (Assume the usual p-value cutpoint of
α=0.05.) - CORRECT-ANSWERSThe percentage of people with diabetes in the
population is not significantly different from 9%.
When we conduct a one-sample z-test (Wald test) for a proportion, we are
making all of the following assumptions except: - CORRECT-ANSWERSThe
data values come from a normal distribution
We might want to use the Exact (Binomial) test instead of the z-test (Wald
test) for a proportion when: - CORRECT-ANSWERSthe sample size is too small
The one-sample t-test relies on all the following quantities except: -
CORRECT-ANSWERScritical value from the normal distribution
A large study was conducted to assess cardiovascular health of cancer
survivors. A sample of 180 survivors had their HDL cholesterol ("good
, cholesterol") measured, and a 95% CI for the mean HDL in the population
was calculated to be (51.8, 56.2). If the same data were used to test whether
the mean HDL was equal to 50 (with a t-test), the resulting p-value and
conclusion would be: - CORRECT-ANSWERSp < 0.05, reject the null
A Type II error is when: - CORRECT-ANSWERSThe null hypothesis is false, and
you fail to reject it.
Suppose the true proportion of dogs who have symptoms from seasonal
allergies is 40%. I gathered a small random sample of dogs and assessed
them for allergy symptoms. Using my sample data, I tested the following
hypothesis: H0: p = 0.40Ha: p ≠ 0.40(p = proportion of dogs who have
seasonal allergy symptoms) I got a p-value of 0.004 and therefore rejected
the null. Which of the following has happened? - CORRECT-ANSWERSA Type I
error
Which of the following changes to a study design will result in lower power
for a one-sample t-test? - CORRECT-ANSWERSIncreased standard deviation
of the variable being tested
Jaylen is planning a new research study and did a sample size calculation.
They concluded that they needed to enroll 50 subjects to have 90% power to
detect the hypothesized effect size (with 5% type 1 error). Jaylen realized
they are actually interested in a smaller effect size. That would require: -
CORRECT-ANSWERSEnrolling more than 50 subjects
In a sample of 250 people, 11.6% of them have diabetes. A one-sample z-
test (Wald test) of whether the true proportion of people in the population
who have diabetes is equal to 9% yields a p-value of 0.15. What is the
appropriate conclusion to draw? (Assume the usual p-value cutpoint of
α=0.05.) - CORRECT-ANSWERSThe percentage of people with diabetes in the
population is not significantly different from 9%.
A one-sample t-test is conducted on blood glucose measurements in a
sample of 15 cats. When testing whether the mean glucose is greater than
200 mg/dL, the appropriate reference distribution to use for calculating the
p-value is: - CORRECT-ANSWERSt distribution with 14 degrees of freedom
A small study measured the body mass index (BMI) of 30 yoga practitioners.
A one-sample t-test was conducted to test whether the mean BMI (in the
population) is equal to 25 with a resulting p-value of 0.18. If we constructed
a 95% confidence interval for the mean BMI using these data: - CORRECT-
ANSWERSThe CI would contain the value 25
Statistical power is... - CORRECT-ANSWERSThe probability you (correctly)
reject the null when the alternative is true.
