Lesson 12 : Bivariate Analysis  II – Testing Relations

Each question has only one best answer. Circle clearly the letter of the best answer. If you make a mistake, cross out the circle, and write the letter in capitals next to the question. If a question has both a capital letter and is circled, the letter will be considered to be the answer.

 

1.     Tammy wants to know whether state levels of Internet usage covary with state education levels. She represents all 50 states as dots on an usage-by-education graph. Then she draws a line through the dots that minimizes the distances from each state’s position in the plot to the line. What is this line called?

a.     The least squares line

b.     The regression line

c.      The best-fitting line

d.     b and c

e.     a, b and c

 

2.     The formula for the regresssion line is y = a + bx. What statement(s) is(are) true about the variable a?

a.     a  is a constant.

b.     a is the dependent variable

c.      a represents the point where the regression line crosses the y axis.

d.     a  is the independent variable

e.     a and c.

 

3.     Saskia correlated respondents’ ages with their incomes and got a Pearson’s r  of 0.30 What percentage of the variance did she explain in her data?

a.     6%

b.     9%

c.      15%

d.     30%

e.     60%

 

4.     What is the PRE measure that tells you how much better you could do if you predicted the separate means for chunks of your data than if you predicted the mean for all your data?

a.     eta

b.     eta-squared

c.      Pearson’s r.

d.     r 2

e.     gamma.

 

5.     Under what conditions is it acceptable to treat ordinal-level variables as if they were interval?

a.     When there are large sample sizes

b.     When there are fewer than 5 ordinal categories

c.      When there are at least 11 ordinal categories

d.     When there are 5 or more ordinal categories

e.     Whenever you want.

 

6.     What is another word for extreme values?

a.     Weights

b.     Outliers

c.      Gemeinschaft

d.     Gesellschaft

e.     Bonferroni correction.

 

7.     What is the best thing to do with outliers?

a.     Eliminate them

b.     Keep them in

c.      Report the results of your analysis with and without them

d.     Divide them by the sample size

e.     Throw out the highest and lowest scores.

 

8.     What does the Bonferroni correction do?

a.     It corrects for the probability of getting a correlation significant at some known level by chance alone.

b.     It increases the size of the p value you can accept to reject your null hypothesis.

c.      It tells you how big a sample you need

d.     It tells you how many times to repeat an experiment

e.     None of the above.

 

9.     What do you need to calculate statistical power?

a.     The minimum size of the difference between two outcomes that you will accept as a real difference.

b.     The sample size

c.      The variance

d.     a and b

e.     a, b and c.

 

10.            John is doing some exploratory research. After collecting data on 20 different variables, he creates a correlation matrix and looks for relationships that are statistically significant. What is John doing?

a.     Conducting an outlier analysis

b.     Shotgunning

c.      Conducting a confirmatory analysis

d.     Conducting a power analysis

e.     Calculating the Bonferroni correction.

 

11.            T F To test for the possibility that one mean is higher than another, you need to do a one-tailed t-test.

a.     True

b.     False

 

12.            T F Analysis of variance requires that the Ns  from each group must be equal.

a.     True

b.     False

 

13.            T F Systematic patterns in data are always linear.

a.     True

b.     False

 

14.            T F A lambda of zero means that there is clearly no association between two variables.

a.     True

b.     False

 

15.            T F Sample size affects the value of chi-square

a.     True

b.     False

 

16.            T F Cramer’s V  and phi are measures of association among two nominal variables and are based on chi-square.

a.     True

b.     False

 

17.            T F Values for gamma range from 0 to 1 .

a.     True

b.     False

 

18.            T F it is possible to test whether a value of a Pearson’s r  is the result of sampling error or reflects a real covariation in the larger population.

a.     True

b.     False.