Reference no: EM13806203
1. Can we use gamma as a measure of association for nominal level variables?
A) No, because gamma implies order and nominal categories cannot be ordered.
B) Yes, because level of measurement does not matter with regard to measures of association.
C) Yes, because gamma is appropriate for any level variable.
D) No, because gamma is not a measure of association.
2. What is the numerical range of gamma?
A) 0 to +1
B) -1 to +1
C) -1 to 0
D) 0 to infinity
3. A "proportional reduction error" (PRE) statistic is a measure of association that...
A) indicates how much error is reduced when the appropriate test of statistical significance is used.
B) indicates how much knowledge of one variable reduces the error involved in predicting a second variable.
C) indicates how much knowledge of the mean reduces the error involved in calculating the standard deviation.
D) indicates how much error is reduced when testing the relationship between two related variables.
4. Let's say I have a 4x4 table of nominal level variables, and Chi-square is significant. Which measure of association would I select?
A) I would not choose to do a measure of association.
B) Phi
C) The chi-square measure of association
D) The contingency coefficient
5. Chi-square may be used even if the variables are not normally distributed.
A) True
B) False
6. Chi-square is a sample statistic with an unknown sampling distribution, and, because so, it can be used to test a null hypothesis.
A) True
B) False
7. In a chi-square it is assumed that the data are collected from a/an:
A) quota sample
B) typical-case or purposive sample
C) random sample
D) accidental sample
8. Chi-square is a statistic used only with:
A) relatively small, random samples
B) relatively small, nonrandom samples
C) relatively large, nonrandom samples
D) relatively large, random samples
9. Researchers often state the hypothesis of a chi-square test in terms of the ________ of the variables.
A) incompatibility
B) independence
C) dependence
D) compatibility
10. Chi-square is most appropriate when the variables are measured at the:
A) ratio level
B) interval level
C) nominal level
D) all choices given
E) ordinal level
11. All of the measures of association for nominal and ordinal variables share a ______ scale, meaning that a value of 0 indicated no relationship and a value of 1 or -1 indicates a perfect relationship.
A) concordant
B) standardized
C) proportional
D) normal
12. Lambda can only be used with 2x2 tables
A) True
B) False
13. What is the numerical range for lambda?
A) -1 to +1
B) 0 to +1
C) 0 to infinity
D) -1 to 0
14. Yule's Q is the same as gamma for
A) a 2x3 table
B) a 2x2 table
C) a 3x3 table
D) a 4x4 table
15. Through a chi-square calculation, we can determine if empirically obtained cell frequencies differ significantly from those we would have expected if no relationship existed between the variables.
A) True
B) False
16. The magnitude of chi-square is related to the number of categories.
A) True
B) False
17. When is phi an appropriate measure of association?
A) For a 2x2 table
B) Nominal level variables
C) When chi-square is significant
D) All of the above
E) A and B only
F) A and C only
18. The _______ frequencies are determined by the ________ hypothesis.
A) expected, research
B) observed, null
C) expected, null
D) observed, research