Exam 4

Question 1

As degrees of freedom moves away from infinity, t distribution:

a. flattens
b. becomes more normal
c. more closely resembles the z distribution
d. remains the same

Answer 1

A: flattens

Question 2

what is not known when performing single sample t-tests?

a. Sample standard deviation
b. Population mean
c. Sample size
d. All of the above are known

Answer 2

D: all the above

Question 3

the degrees of freedom for single sample t is equal to

a. N
b. N – 1
c. N – 2
d. the population mean

Answer 3

N-1

Question 4

the 99% confidence interval will be __ the 95% confidence interval

a. wider than
b. narrower than
c. the same as
d. none of the above

Answer 4

wider than

Question 5

Which is not true of the sampling distribution of t?
a. It gives all possible different t values for samples of size N
b. It gives the probability of getting each value if sampling is random from the Null
Hypothesis Population
c. It is identical to the z distribution
d. All of the above are true

Answer 5

C: identical to z distribution

Question 6

With other factors held constant, as the sample size decreases, the width of the confidence
interval will
a. decrease
b. increase
c. remain the same
d. all of the above are possible

Answer 6

B: increase

Question 7

Which of the following are properties of the t distribution?

a. it is a family of many curves that varies with degrees of freedom
b. it is symmetrical
c. it has a mean of 0
d. all of the above are true
e. a and b

Answer 7

D: all the above

Question 8

The independent group design involves analysis of the ______ scores and the repeated
measures design involves analysis of the ______ scores
a. raw; difference
b. raw; raw
c. difference; raw
d. difference; difference

Answer 8

A: raw; difference

Question 9

Which of these is (are) true when using the t-test for correlated groups?

a. The t-test uses information on the direction of the difference scores
b. The t-test uses information on the magnitude difference scores
c. The t-test will be more powerful than the sign test
d. All of the above are true

Answer 9

D: all the above

Question 10

In the t-test for independent samples, there are _____ degrees of freedom

a. n1 – 1
b. n1 + n2 – 2
c. n1 – n2
d. n1 + n2

Answer 10

B: n1 + n2 - 2

Question 11

When conducting a t-test for independent groups, an appropriate non-directional alternative hypothesis would assert that the sample mean difference is due to sampling from a population where

a. µ1 = µ2
b. µ1 = 0
c. µ1≠ 0
d. µ1≠ µ2

Answer 11

D: µ1≠ µ2

Question 12

Which is NOT true regarding assumptions for use of the independent groups t-test?

a. populations from which samples are taken from are normal
b. homogeneity of variance is assumed
c. homoscedasticity is assumed
d. all of the above are true

Answer 12

c. homoscedasticity is assumed

Question 13

In a correlated groups design, the null hypothesis would state that the mean of the sample difference scores are from a population of differences scores where

a. µD= 0
b. µD≠ 0
c. µD= 1
d. None of the above

Answer 13

a. µD= 0

Question 14

What is true about the sampling distribution of F?

a. There are negative values
b. It is negatively skewed
c. It is normal
d. None of the above

Answer 14

d. None of the above

Question 15

What do you need to know about a population when conducting an F test?

a. µ
b. standard dev
c. Neither µ nor standard dev
d. Both

Answer 15

C: neither

Question 16

The alternative hypothesis used in the ANOVA is

a. always non-directional
b. always directional
c. sometimes directional and sometimes non-directional
d. µ1 = µ2 = … = µk

Answer 16

a. always non-directional

Question 17

If Fobt is less than Fcrit, what would you decide?

a. Retain the null hypothesis
b. Accept the alternative hypothesis
c. Reject the null hypothesis
d. Depends on df

Answer 17

a. Retain the null hypothesis

Question 18

SSB + SSW =

a. 1
b. s2T
c. SST
d. Depends on df

Answer 18

C. SST

Question 19

What is the problem with doing multiple t tests between each group of the experiment when there are more than two groups?

a. It is impossible to calculate t in that manner
b. The probability of making a Type I error increases
c. The probability of making a Type II error increases
d. There is no problem

Answer 19

b. The probability of making a Type I error increases

Question 20

What does “a priori comparison” mean?

a. Comparisons planned after the experiment has been completed
b. Comparisons planned before the experiment has been completed
c. Multiple comparisons are made
d. The same as post hoc comparison

Answer 20

b. Comparisons planned before the experiment has been completed

Question 21

The probability of making one or more Type I errors for the full set of possible multiple comparisons in an experiment is called

a. beta
b. comparison-wise error rate
c. alpha
d. experiment-wise error rate

Answer 21

d. experiment-wise error rate

Question 22

One employs a correlated groups design

a. specify calculations eliminate random error
b. eliminate random error
c. reduce effects of individual differences among subjects
d. for prediction

Answer 22

C.

Question 23

if a test is a robust test

a. sensitive to its underlying mathematical assumptions
b. doesn’t estimate any population parameters
c. insensitive to its underlying mathematical assumptions
d. may be used despite violations of its underlying mathematical assumptions
e. c and d

Answer 23

E

Question 24

independent groups experiment, if we must estimate the population standard dev. to determine the significance of the sample results, the appropriate inference test is the

a. sign test
b. z test
c. t test
d. none of above

Answer 24

c

Question 25

students t test for correlated groups really reduces to a. sign test b. students t test for single samples using differences scores c. students t test for independent groups d. none of above

Answer 25

b

Question 26

when analyzing data from experiments that involve more than 2 groups a. doing t tests on all possible pairs of means increases probability of making type II errors b. doing t tests on all possible pairs of means increases probability of making type I errors c. generally permissible to do t tests between all possible pairs of means and use students t distribution d. doing t tests on all possible pairs of means decreases probability of making type I errors

Answer 26

b

Question 27

a priori comparisons a. are planned in advance of experiment b. often arise out of theory and prior research c. may be done w/o obtaining a significant F in the one-way ANOVA d. all the above

Answer 27

d

Question 28

t-test for independent groups, the unbiased estimate of the population variance is a. weighted avg of s1^2 and s2^2 b. s2^2 alone c. s1^2 alone d. none

Answer 28

a

Question 29

the one-way ANOVA partitions the total variability into a. SSw and SSb b. SSw and SSt c. Sw^2/MSWithin and SSw d. Sb^2/MSBetween and SSb

Answer 29

a

Question 30

which would cause Fobt to increase? a. increase in difference between the means b. increase in the within-groups variability c. increase in magnitude of the independent variable's effect d. a and b e. a and c

Answer 30

e

Question 31

when N is less than infinity, t crit for any alpha level is __ z crit a. same b. less than c. greater than d. dependent on N

Answer 31

c

Question 32

how many t distributions are there? a. 1 b. 2 c. 30 d. one for each value of df

Answer 32

d

Question 33

for a given value of alpha, the critical value of t is __than the critical value of z a. no different b. less c. greater d. none

Answer 33

c

Question 34

as N gets __, the critical value of t gets __ a. larger, larger b. smaller, smaller c. larger, smaller d. smaller, larger e. both c and d

Answer 34

e

Question 35

as N gets infinitely large, the critical value of t equals the critical value of z a. true b. false

Answer 35

a

Question 36

if the population parameters are known, the t test is __ powerful than the z test a. more b. less

Answer 36

b

Question 37

s(x bar) is an estimate of __ a. u b. population standard dev c. mean of standard dev [o(xbar)] d. s

Answer 37

c

Question 38

in a correlated t test, if the ind. variable has no effect, the sample difference scores are a random sample from a population where the mean difference score (uD)= a. 0 b. 1 c. N d. cannot be determined

Answer 38

a

Question 39

in analyzing the result of the correlated t test, the __ scores are analyzed a. standardized b. normalized c. raw d. difference

Answer 39

d

Question 40

in the t test for independent groups, a. we estimate u1 - u2 b. we estimate o^2 c. we estimate s^2 d. df = N-1

Answer 40

d

Question 41

as the differences between the group mean increases (rejecting null) a. Sb^2 increases b. Fobt decreases c. Tobt increases d. none

Answer 41

a

Question 42

a major limitation of a 2 group design is that a. we cant run enough subjects b. very often, 2 groups are insufficient for a clear interpretation c. population parameters may not be applicable d. means between groups may not differ

Answer 42

b

Question 43

when analyzing data from experiments that involve more than 2 groups, a. doing t tests on all possible pairs of means decreases the probability of making type I errors b. doing t tests on all possible pairs of means increases the probability of making type I errors c. is it generally permissible to do t tests between all possible pairs of means and use students t distribution d. doing t tests on all possible pairs of mean increases the probability of making type II errors

Answer 43

b

Question 44

which of the following is (are) illegal values for Fobt? a. 1.00 b. 0.96 c. -2.97 d. b and c

Answer 44

c

Question 45

by doing multiple t tests when there are more than 2 experimental groups we increase the risk of making what kind of mistake? a. accepting the null b. type I error c. type II error d. all the above

Answer 45

b

Question 46

type I error is

Answer 46

rejecting null when it is true

Question 47

type II error is

Answer 47

not rejecting null when the alt hypothesis is true