Question

Question5. Please answer the following questions as either "TRUE" or "FALSE" (Write out either “True” or “False”)

A. A normal distribution is characterized by its mean and its degrees of freedom. ______________

B. A paired t-test is considered a nonparametric test. ______________

C. A 93% confidence interval for µ is that numerical interval constructed to contain 93% of the values of µ. ____________

D. A 94% confidence interval for a mean is narrower than a 93% confidence interval. ______________

E. A regression analysis involving 93 patients would have 92 degrees of freedom. ______________

F. A negative correlation coefficient indicates a poor linear relationship. ______________

G. For a two-tailed hypothesis test comparing two group means, assuming a normal distribution and standard deviation σ known, a p-value < 0.05 is analogous to | z-value | > 1.96. ______________

H. An increase in the Type I error will decrease Power, everything else held constant. ______________

I. If the Type II error of a hypothesis test is equal to 0.06, the Power of the test is 0.94. __________

Multiple Choice: (Please circle the correct answer!)

J. The normal approximation may be used as an approximation to the binomial distribution when: (n=sample size)

a) the sample size is at least 30

b) np and nq are both less than 5 [p=P(success), q = 1-p]

c) np and nq are both greater or equal to 5 [p=P(success), q = 1-p]

K. For the normal distribution, with continuous data and the population variance known, the standard error of the mean is determined by the:

a) The population mean and the sample size

b) The population standard deviation and the sample size

c) The population mean and the population standard deviation

d) Sample mean and the population standard deviation

d) np is greater than or equal to 30

L. A 95% confidence interval for a population mean birth weight was determined to be 2100 to 2700 grams. Assume birth weights are normally distributed. The p-value for a test of the null hypothesis H0: Uo= 3000 grams versus Ha: Uo # 3000 grams is approximately.

a) 0.10

b) 0.06 e) 0.25

c) < 0.05

d) > 0.05

Answer 1

A. A normal distribution is characterized by its mean and its degrees of freedom. ___False___________

A normal distribution is characterized by its mean and standard deviation.

B. A paired t-test is considered a nonparametric test. ___False___________

A paired t-test is considered a parametric test.

C. A 93% confidence interval for µ is that numerical interval constructed to contain 93% of the values of µ. _________False____

If we construct 93% confidence interval many times, then about 93% of the intervals produced will capture µ

D. A 94% confidence interval for a mean is narrower than a 93% confidence interval. _________False____

94% confidence interval for a mean is wider than a 93% confidence interval

E. A regression analysis involving 93 patients would have 92 degrees of freedom. _________False____

For one predictor, degree of freedom = n - 2 = 93 - 2 = 91

F. A negative correlation coefficient indicates a poor linear relationship. _________False____

A negative correlation coefficient of -1 indicates a strong linear relationship.

G. For a two-tailed hypothesis test comparing two group means, assuming a normal distribution and standard deviation σ known, a p-value < 0.05 is analogous to | z-value | > 1.96. True

H. An increase in the Type I error will decrease Power, everything else held constant. _________False____

An increase in the Type I error will decrease Type II error and thus increases Power,

I. If the Type II error of a hypothesis test is equal to 0.06, the Power of the test is 0.94. True

Power = 1 - Type II error

Multiple Choice: (Please circle the correct answer!)

J. The normal approximation may be used as an approximation to the binomial distribution when: (n=sample size)

c) np and nq are both greater or equal to 5 [p=P(success), q = 1-p]

K. For the normal distribution, with continuous data and the population variance known, the standard error of the mean is determined by the:

b) The population standard deviation and the sample size

L. A 95% confidence interval for a population mean birth weight was determined to be 2100 to 2700 grams. Assume birth weights are normally distributed. The p-value for a test of the null hypothesis H0: Uo= 3000 grams versus Ha: Uo # 3000 grams is approximately.

Since 3000 grams is greater than upper limit of confidence interval, p-value should be lower than 0.05

c) < 0.05