Chapter 15 Test Items

1. To conduct a chi-square test, what is the appropriate combination of variables?

a. nominal, ratio

b. nominal, interval

c. interval, ratio

d. nominal, nominal

2. Which of the following is NOT a basic assumption that must be met prior to the utilization of the chi-square test?

a. Both variables being analyzed must be nominal in nature.

b. Participants contributing data should represent a random sample drawn from the population of interest.

c. If you have a 2 × 2 table, there should be no fewer than three cases in every cell.

d. One participant’s appearance in a category should not affect the probability of another

participant’s appearance in another cell.

3. Which is the correct way to present a chi-square finding?

a. χ (34 = 5.5), p < .05

b. χ (3, N = 34) = 5.5 , p < .05

c. χ2 (3, N = 34) = 5.5 , P < .05

d. χ2 (3, N = 34) = 5.5 , p < .05

4. If you receive the finding phi = .40, what can you say about your finding?

a. 40% of the variance is accounted for

b. 16% of the variance is accounted for

c. 20% of the variance is accounted for

d. .40% of the variance is accounted for

5. Which of the following should be the first step in computing the chi-square test of independence?

a. add the totals from each row separately

b. calculate the degrees of freedom

c. keep careful track of where you are getting your numbers from

d. calculate the sampling error

6. Which of the following is defined as the number of participant scores in a sample that can or are free to vary?

a. calculated value

b. critical value

c. degrees of freedom

d. expected frequency

7. Chi-square equals which of the following?

a. sum of (expected frequency – observed frequency)2/expected frequency

b. sum of (expected frequency – observed frequency)2/observed frequency

c. sum of (observed frequency – expected frequency)2/expected frequency

d. sum of (observed frequency – expected frequency)2/observed frequency

8. Which statistic tells a researcher how much of the difference in one variable can be accounted for by the variance in another variable?

a. effect size

b. Cronbach’s alpha

c. Cramer’s phi

d. Chi-square

9. The chi-square is what type of test?

a. a difference test

b. a relationship test

c. a power test

d. an effect size test

10. One way to prevent Type I error is to lower the significance level. Which of the following tests may be used to help a researcher lower the significance level?

a. Cross tab

b. Cronbach’s alpha

c. Cramer’s phi

d. Dunn-Sidak test

For questions 11 to 18, you will use the chi-square results below. A researcher wanted to determine if female and male shoppers differed in the type of dress they wear at a large superstore. The researcher sits outside and categorizes shoppers as conservative, preppy, or punk. Here are the data used for calculating the chi-square.

11. In the chi-square presented, what were the degrees of freedom?

a. 2

b. 1

c. 100

d. 2.93

12. In the chi-square presented, what is the calculated value?

a. 2.93

b. 2.95

c. 2.39

d. 1.50

13. In the chi-square presented, what was N?

a. 100

b. 50

c. 2.93

d. .005

14. In the chi-square presented, what was the p value?

a. .23

b. 2.93

c. 2

d. .17

15. In the chi-square presented, which of the following statements is correct?

a. The chi-square was not statistically significant.

b. The chi-square was statistically significant.

c. The chi-square reported a large effect size.

d. The chi-square reported a small effect size.

16. In the chi-square presented, how would you categorize the p value?

a. p > .05

b. p < .05

c. p < .01

d. p < .005

17. In the chi-square presented, which of the following is the correct APA write-up?

a. χ (2) = 2.93, p = .23

b. χ² (1) = 2.95, p = .23

c. χ (2) = 2.93, p < .05

d. χ² (1) = 2.95, p < .05

18. In the chi-square presented in this quiz, what was the effect size?

a. .17

b. .23

c. 2

d. 2.93

19. Bob is conducting a chi-square and has one group that has three nominal variables and another that as four. What would his df be?

a. 6

b. 12

c. 4

d. 2

20. When using SPSS to calculate chi-square, which functions should you use?

a. Crosstabs

b. Crossword

c. calculated value

d. Cramer’s phi

21. The one-sample chi-square test is more likely to yield significance if the sample proportions

for the categories differ greatly from the hypothesized proportions and if the sample size is small.

a. True

b. False

22. Before a researcher ever begins calculating the statistical test, he or she must set the significance level of the test.

a. True

b. False

23. The purpose of a chi-square test is to examine the balance between observed frequencies and expected frequencies.

a. True

b. False

24. If the observed and expected frequencies are equal, the chi-square statistic will always equal one.

a. True

b. False

25. Jake wants to analyze the difference between men and women (nominal variable) and self- disclosure (interval variable). Jeff should use the chi-square test.

a. True

b. False

26. When conducting a chi-square, it is recommended that both variables being analyzed are nominal in nature but not necessary.

a. True

b. False

27. In a 2 × 2 chi-square table, there should be no fewer than five cases in every cell.

a. True

b. False

Essays:

28. Explain the purposes of a chi-square statistical test.

29. Explain why a researcher may need to adjust his or her alpha level using a Dunn-Sidak test.

30. Explain the purpose of degrees of freedom within statistics.

31. Explain how to conduct a chi-square test, step by step.

32. Write one research question and one hypothesis that can be addressed using a chi-square.

Matching:

33. Match each of the following terms with the correct statement.

a. Critical Value Table = A table that has been previously created to determine where specific chi-square calculated values are statistically significant.

b. chi-square = Statistical difference test used when a researcher has two nominal variables.

c. Degrees of Freedom = The number of participant scores in a sample that can or are free to vary.

d. Phi or Cramer’s V = Statistical test that demonstrates how much of the difference in one variable can be accounted for by the variance in another variable.

e. Dunn-Sidak Test = Statistical test that readjusts the probability level to correct for possible compounded error resulting from multiple pairwise comparisons to help prevent type I error.