Deck 20: Data Analysis and Interpretation: Basic Difference Questions

Question 1

Which of the following could NOT be a single-factor between-groups design?

A) Randomized experimental approach
B) Comparative approach
C) Associational approach
D) Quasi-experimental approach

Answer 1

Associational approach

Question 2

The single-factor between-groups design may have an independent variable with:

A) Two levels
B) Three levels
C) Four levels
D) All of the above

Answer 2

All of the above

Question 3

Nonparametric statistics are used with single-factor between-groups designs when:

A) Assumptions of normality are violated
B) The dependent variable is normally distributed
C) The dependent variable is at least interval level
D) All of the above

Answer 3

Assumptions of normality are violated

Question 4

There are three major assumptions underlying the use of the t test for independent samples. The violation of one assumption is not well tolerated. This assumption is:

A) The independent variable should be normally distributed
B) The dependent variable should be normally distributed
C) Groups, and the data from them, must be independent
D) Variances of the dependent variables should be equal

Question 5

Joan Bliss wants to find out if there are gender differences in marital satisfaction. She has 40 sets of spouses complete the 10-item Dyadic Adjustment Scale, which is an interval measure. She used an independent samples t test to compare wives to husbands, and found husbands to be significantly happier t(78) = 6.23, p < .001. What did Dr. Bliss do wrong here?

A) One shouldn't test for gender differences; it's politically incorrect
B) Dr. Bliss should have used an F test because she had a nondirectional hypothesis
C) She should have used a Wilcoxon matched-pairs test because the dependent variable was an interval measure
D) Dr. Bliss used an independent samples t test with groups who are dependent
E) She should have done a randomized experiment to examine gender differences

Question 6

A researcher conducts a study and rejects the null hypothesis at an alpha level of .02. Which of the following is true?

A) The difference in population means could not have happened by chance
B) There is a 98% chance that the true population mean difference is within the confidence interval
C) The true population mean difference would be in 98 of the confidence intervals but not in two of them
D) b and c

Question 7

The single-factor ANOVA should be used when:

A) The hypothesis being tested is nondirectional and the dependent variable is nominal
B) The hypothesis being tested is directional with at least 2 levels to the IV
C) The hypothesis being tested is nondirectional with at least 2 levels to the IV
D) The hypothesis being tested is directional with at least 3 levels to the IV

Question 8

Multiple comparisons between groups using t tests is discouraged because:

A) The probability of as Type I error is increased
B) The probability of a Type II error is increased
C) The degrees of freedom are substantially increased
D) a and c

Question 9

A researcher analyzes her data using a one-way ANOVA. When would she use post-hoc comparisons?

A) The F value is significant and there are more than two groups being compared
B) The F value is significant and there are two groups being compared
C) The F value is nonsignificant and there are 3 or more groups being compared
D) a and b

Question 10

A researcher decides to use non-parametric statistics for her study's analyses because:

A) The independent variable is at an ordinal level
B) The dependent variable is at an ordinal level
C) The independent variable is not normally distributed
D) b and c

Question 11

Compared to parametric analyses, nonparametric tests

A) Have greater power
B) Increase the probability of rejecting the null hypothesis
C) Are more often based on rankings
D) Are most appropriate when one has interval or ratio data.
E) None of the above

Question 12

The Mann-Whitney U test can be used to compare differences between groups when:

A) There are a small number of participants
B) Sample sizes are not equal
C) Data are ranked or interval level
D) All of the above

Question 13

Dr. Dread teaches one group of students with a standard lecture format whereas the other class is taught using a group discussion and cooperative learning approach. He gives a cumulative final exam at the end of the term that is scored pass of fail. Which of the following statements is true of Dr. Dread's approach?

A) His use of nominal outcome data represents a loss of information in comparison to an alternative method that uses ordinal level data
B) There are multiple levels of the IV
C) The IV and DV are both categorical
D) a and c

Question 14

A researcher does a series of studies to see what affects motor development. In Study 1, preterm infants are compared to full-term infants on their motor development at one year of age. In Study 2, the motor development of two groups is compared: one-year-olds whose parents
Had enrolled them in "baby gymnastics" versus one-year-olds without this experience. In
Study 3, six-month-olds are randomly assigned to a group that receives extra large-motor
Stimulation (in "walkers") versus no enrichment; the two groups' large-motor skills are
Compared at twelve months.
Assuming that the test of motor development is at the interval level, which type of statistical
Test is most appropriate for each study?
Study 1:
Study 2:
Study 3:

A) correlation coefficient
B) t test
C) Wilcoxon matched-pairs
D) F test
E) Mann-Whitney U test

Question 15

A researcher does a series of studies to see what affects motor development. In Study 1, preterm infants are compared to full-term infants on their motor development at one year of age.
In Study 2, the motor development of two groups is compared: one-year-olds whose parents
had enrolled them in "baby gymnastics" versus one-year-olds without this experience. In
Study 3, six-month-olds are randomly assigned to a group that receives extra large-motor
stimulation (in "walkers") versus no enrichment; the two groups' large-motor skills are
compared at twelve months.
Assuming that the test of motor development is at the interval level, which type of statistical
test is most appropriate for each study?
Study 1:
Study 2:
Study 3:
t test^{(for all)}
What type of approach is used in each study?

Question 16

A single-factor within-subjects design would be used when there is one independent variable with two or more levels and:

A) Participants undergo all conditions of a study
B) The two groups are related or matched on some variable related to the dependent variable
C) A one group pretest-posttest design is used
D) All of the above

Question 17

A single -actor within-subjects design can be used with the following research approaches:

A) Associational
B) Comparative
C) Quasi-experimental
D) b and c
E) All of the above

Question 18

There are some advantages to using a repeated-measures design, such as:

A) Variability among participants is increased
B) Type I error variance is greater, increasing the probability of a significant difference
C) Fewer participants are needed
D) b and c

Question 19

Disadvantages to using a repeated-measures design include:

A) Increased design flexibility
B) Statistical power due to n is reduced
C) Carryover effects
D) b and c
E) All of the above

Question 20

To use the matching procedure:

A) Participants are grouped in dyads or triads based on a characteristic unrelated to the DV
B) Participants are randomly assigned to groups and then matched on a characteristic
C) Participants are grouped in dyads or triads based on a characteristic related to the DV
D) b and c

Question 21

Conditions for using the t test for related samples include:

A) The independent variable has two or more levels
B) Participants have been matched on a relevant variable
C) Normality of the dependent variable has not been violated
D) b and c
E) All of the above

Question 22

Conditions for using the single-factor repeated-measures ANOVA include:

A) The independent variable has two or more levels
B) There is a directional hypothesis
C) The dependent variable is at least ordinal in level of measurement
D) a and b

Question 23

A researcher performs a single-factor repeated-measures ANOVA and finds a significant difference among the study's three treatment conditions. The researcher should:

A) Perform a Tukey HSD post hoc test
B) Assume that the group differences were due to chance alone
C) Establish at least a 95% confidence interval
D) All of the above

Question 24

Professor U. R. Okay wanted to find a way to decrease bullying and increase peer acceptance among 4^th graders. He matched 10 triads of children in terms of their peer popularity, and then randomly assigned one member of each triad to a social skills training program, another to a group discussion that focused on changing peer reputation, and the third member of each triad was in the control group. After 3 months, he measured whether the children were popular or rejected. The statistical approach he should use is:

A) McNemar test
B) an ANOVA (F test)
C) Wilcoxon Sign test
D) Cochran Q test
E) Friedman two-way analysis of variance

Question 25

The Wilcoxon matched-pairs test is used when:

A) There is one IV with two levels
B) Participants undergo both conditions
C) There have been violations of assumptions of the t test for correlated samples
D) All of the above

Question 26

The Friedman two-way analysis of variance by ranks is used in a repeated-measures designs when:

A) The independent variable has only two levels
B) There are two independent variables
C) The dependent variable is an ordinal level of measurement
D) a and c

Question 27

Single-factor between-groups designs use the same statistical analyses.

Question 28

Single-factor between-groups designs are interpreted differently due to their relative internal validity.

Question 29

The numbers of independent pieces of information from a study's data are used to figure the degrees of freedom.

Question 30

Rejecting the null hypothesis implies that a difference between population means could not have happened by chance.

Question 31

The t test provides more power than an ANOVA when using a directional hypothesis.

Question 32

Data should be at least ordinal level when using the single-factor ANOVA.

Question 33

The ANOVA is used to compare the variability between groups to the variability within groups.

Question 34

The F statistic is the summary statistic for all the analyses of variance.

Question 35

The F value tells you which groups or conditions are significantly different when performing an ANOVA.

Question 36

When an F statistic is significant, a post hoc test must be performed when the independent variable has more than two levels.

Question 37

Post-hoc comparisons are necessary when the overall F is significant and the independent variable has two levels.

Question 38

Nonparametric tests are typically not as powerful as parametric tests.

Question 39

Most nonparametric tests use a ranking procedure.

Question 40

Levene's test should be used to determine if the assumption of homogeneity of variance has been violated when doing comparative statistics.

Question 41

Post hoc comparisons are not necessary to do when using nonparametric statistics with an independent variable that has more than two levels.

Question 42

Researchers should always try to reduce their data into a smaller number of categories for easier statistical management.

Question 43

The level of measurement for the dependent variable does not affect the amount of information gleaned from the data.

Question 44

With repeated-measures designs, error variance is reduced because the same participant is experiencing each of the conditions.

Question 45

Carryover effects reduce the design flexibility of repeated-measures designs.

Question 46

The matching procedure can be used to circumvent carryover effects with repeated-measures designs.

Question 47

The matching procedure can be used to increase error variance with repeated-measures designs.

Question 48

A repeated-measures design will decrease sample size and decrease error variance.

Question 49

The t test for correlated samples utilizes change scores.

Question 50

With repeated-measures analysis of variance statistics, the treatment effect comes from the within-subjects component.

Question 51

The Wilcoxon matched-pairs test utilizes ranks.

Question 52

Under what conditions would you use a single-factor ANOVA?

Question 53

How do you determine the degrees of freedom for a dependent samples t test?

Question 54

Explain what the numbers mean in the following: t₂₈₀= 6.21, p = .032.