Practice Quiz Chapter 15

* not completed

. What is logistic regression?

a) A type of regression used to make predictions about a dichotomous dependent variable correct incorrect

b) A type of regression used to make predictions about a dichotomous independent variable correct incorrect

c) A type of regression used to make predictions about a tri-chotomous dependent variable correct incorrect

d) A type of regression used to make predictions about a tri-chotomous independent variable correct incorrect

e) A type of regression used to make predictions about a right-skewed dependent variable correct incorrect

* not completed

. How do researchers work around the issues that arise when they use linear regression to find the straight line that best fits the relationship between a ratio-level independent variable and a dichotomous dependent variable?

a) They transform dichotomous dependent variables after using them in regression correct incorrect

b) They transform dichotomous dependent variables before using them in regression correct incorrect

c) They transform dichotomous dependent variables during regression correct incorrect

d) They transform dichotomous independent variables after using them in regression correct incorrect

e) They transform dichotomous independent variables during regression correct incorrect

* not completed

. Which of the following statements about the mathematical transformation of the dichotomous dependent variable is not valid?

a) The mathematical transformation of the dichotomous dependent variable in a logistic regression has three steps correct incorrect

b) Instead of trying to predict the value on the dependent variable, researchers try to predict the probability that a case will have the value 1 on the dependent variable correct incorrect

c) Move from probabilities to odds correct incorrect

d) Move from odds to probabilities correct incorrect

e) Using log transformations to transform a dichotomous dependent variable so it can be used in regression correct incorrect

* not completed

. What are odds?

a) The number of times something occurs correct incorrect

b) The number of times something does not occur correct incorrect

c) The number of times something occurs relative to the number of times it does not occur correct incorrect

d) The number of times something does not occur relative to the number of times it does occur correct incorrect

e) The probability that something occurs correct incorrect

* not completed

. Calculate the odds of attending a statistics class when there are 12 classes per term and you attend 8 of them.

a) 2 to 1 correct incorrect

b) 4 to 1 correct incorrect

c) 20 to 1 correct incorrect

d) 2 to 2 correct incorrect

e) 1 to 1 correct incorrect

* not completed

. Calculate the odds of attending a statistics class when there are 12 classes per term and you attend 6 of them.

a) 2 correct incorrect

b) 1 correct incorrect

c) 10 correct incorrect

d) 12 correct incorrect

e) 6 correct incorrect

* not completed

. Calculate the probability of a dichotomous variable being equal to 1 if the probability of it equalling 0 is 0.2.

a) 8.1 correct incorrect

b) 0.08 correct incorrect

c) 8 to 1 correct incorrect

d) 8 correct incorrect

e) 0.8 correct incorrect

* not completed

. Calculate the probability of a dichotomous variable being equal to 0 if the probability of it being equal to 1 is 0.7.

a) 7 correct incorrect

b) 3.1 correct incorrect

c) 0.03 correct incorrect

d) 0.3 correct incorrect

e) 3 correct incorrect

* not completed

. What is Euler's constant?

a) 2.71828 correct incorrect

b) 2.78128 correct incorrect

c) 2.78182 correct incorrect

d) 2.71882 correct incorrect

e) 2.71228 correct incorrect

* not completed

. Which of the following statements about log-transformed, dichotomous dependent variables is not valid?

a) Log transformations are the third and final step in transforming a dichotomous dependent variable for use in a regression correct incorrect

b) When a variable is log-transformed, values less than 1 in the original variable become negative values in the transformed variables correct incorrect

c) When a variable is log-transformed, values less than 1 in the original variable become positive values in the transformed variables correct incorrect

d) When a variable is log-transformed, the value 0 in the original variable becomes the value 1 in the transformed variable correct incorrect

e) When a variable is log-transformed, values greater than 1 in the original variable become positive values in the transformed variable correct incorrect

* not completed

. One of the following statements about log odds is not true - which one is it?

a) When odds are transformed into the natural log of the odds, they are called log odds correct incorrect

b) When there is less than a 0.5 probability of something occurring, the log odds are negative values correct incorrect

c) When there is exactly a 0.5 probability of something occurring, the log odds are 0 correct incorrect

d) When there is more than a 0.5 probability of something occurring, the log odds are positive values correct incorrect

e) When there is exactly a 0.5 probability of something occurring, the log odds are 1 correct incorrect

* not completed

. If the probability of something occurring is 0.7, then calculate the odds of something occurring.

a) 2.33 correct incorrect

b) 3.22 correct incorrect

c) 0.0233 correct incorrect

d) 0.233 correct incorrect

e) 33.2 correct incorrect

* not completed

. If the probability of something occurring is 0.3, then calculate the odds of something occurring.

a) 0.43 correct incorrect

b) 4.3 correct incorrect

c) 0.043 correct incorrect

d) 43.3 correct incorrect

e) 0.403 correct incorrect

@See page 48 correct incorrect

* not completed

. Calculate the natural log odds of something occurring if the probability of something occurring is 0.7.

a) 85 correct incorrect

b) 0.0085 correct incorrect

c) 0.085 correct incorrect

d) 8.5 correct incorrect

e) 0.85 correct incorrect

* not completed

. Calculate the natural log odds of something occurring if the probability of something occurring is 0.3.

a) 0.0085 correct incorrect

b) 0.085 correct incorrect

c) -0.085 correct incorrect

d) 0.85 correct incorrect

e) -0.85 correct incorrect

* not completed

. Although logistic regression uses an approach that is like linear regression, there are several key differences–which one of the following is not one of those differences?

a) Logistic regression is a type of non-linear regression correct incorrect

b) Because logistic regression incorporates a log transformation of the dependent variable, it does not predict a straight-line relationship between variables correct incorrect

c) Because of the transformation of the dependent variable, it's not possible to calculate the slope and constant coefficients of a logistic regression in the same way as a linear regression correct incorrect

d) Because of the transformation of the dependent variable, the interpretation of the regression coefficient changes correct incorrect

e) Because of the transformation of the dependent variable, it's possible to calculate the slope and constant coefficients of a logistic regression in the same way as a linear regression correct incorrect

* not completed

. What are odds ratios?

a) Shows how the odds of something occurring in two different groups compares to each other correct incorrect

b) Shows how the odds of something occurring in three different groups compares to each other correct incorrect

c) The odds of something occurring in the second group expressed as a ratio of something occurring in the first group correct incorrect

d) Shows how the odds of something occurring in four different groups compares to each other correct incorrect

e) Not a good way to make fair comparisons between groups correct incorrect

* not completed

. If the odds of lower-income adults going hungry is equal to 0.176 and the odds of higher-income adults going hungry is 0.053, then calculate the odds ratio - what is it?

a) 0.032 correct incorrect

b) 3.32 correct incorrect

c) 33.2 correct incorrect

d) 3.03 correct incorrect

e) 0.0033 correct incorrect

* not completed

. One of the following statements about odds ratios is not true–which one is it?

a) An odds ratio greater than 1 indicates that the odds of something occurring for the group shown in the numerator of the ratio are higher than the odds of that thing occurring for the group shown in the denominator correct incorrect

b) An odds ratio less than 1 indicates that the odds of something occurring for the group shown in the numerator of ratio are lower than the odds of that thing occurring for the group represented in the denominator correct incorrect

c) An odds ratio of 1 indicates that the two groups have the same odds of something occurring correct incorrect

d) An odds ratio of 1 indicates that the two groups don't have the same odds of something occurring correct incorrect

e) Researchers usually report how much higher (or lower) the odds of something occurring are for a specific group using percentages correct incorrect

* not completed

. It's easiest to interpret odds ratios using a two-step process–what are the two steps?

a) Subtract 1 from the odds ratio and divide the remaining number by 100 to show the percentage difference in the odds correct incorrect

b) Add 1 to the odds ratio and divide the remaining number by 10 to show the percentage difference in the odds correct incorrect

c) Add 1 to the odds ratio and multiply the remaining number by 10 to show the percentage difference in the odds correct incorrect

d) Add 1 to the odds ratio and multiply the remaining number by 100 to show the percentage difference in the odds correct incorrect

e) Subtract 1 from the odds ratio and multiply the remaining number by 100 to show the percentage difference in the odds correct incorrect

* not completed

. Since many things that social science researchers want to make predictions about cannot be captured in ratio-level variables, logistic regression is widely used.

a) True correct incorrect

b) False correct incorrect

* not completed

. Transforming dichotomous dependent variables so they range from negative infinity to positive infinity allows researchers to continue using the same general approach they use in linear regression to make predictions about a dichotomous dependent variable.

a) True correct incorrect

b) False correct incorrect

* not completed

. When dichotomous dependent variables are transformed from negative infinity to positive infinity the interpretation of the regression coefficients remains the same.

a) True correct incorrect

b) False correct incorrect

* not completed

. When transforming a dichotomous dependent variable in a logistic regression, the value on an independent variable is used to predict the probability that the dependent variable will be equal to 1.

a) True correct incorrect

b) False correct incorrect

* not completed

. When transforming a dichotomous dependent variable in a logistic regression, the value on the dependent variable is used to predict the probability that the independent variable will be equal to 1.

a) True correct incorrect

b) False correct incorrect

* not completed

. When transforming a dichotomous dependent variable in a logistic regression, transforming probabilities into odds is useful because odds range from 0 to positive infinity.

a) True correct incorrect

b) False correct incorrect

* not completed

. For a dichotomous dependent variable, the probability that something will not occur is never equal to 1 minus the probability that something will occur.

a) True correct incorrect

b) False correct incorrect

* not completed

. There are only two possible values in a dichotomous variable, and the probability of having either value is equal to 1, or 100%.

a) True correct incorrect

b) False correct incorrect

* not completed

.

Researchers often use p_i to denote the probability that y will be equal to 1, given x.

a) True correct incorrect

b) False correct incorrect

* not completed

. When the probability that a dichotomous dependent variable will be equal to 1 is transformed into odds, the values on the dependent variable can potentially range from 0 to negative infinity.

a) True correct incorrect

b) False correct incorrect

* not completed

. In the natural sciences and mathematics, researchers often transform values of some variable using the natural log, a transformation which relies on Euler's constant.

a) True correct incorrect

b) False correct incorrect

* not completed

. The decimals of Euler's constant are limited–in other words they don't continue forever.

a) True correct incorrect

b) False correct incorrect

* not completed

. Log odds refer to the natural log of the odds of something not occurring.

a) True correct incorrect

b) False correct incorrect

* not completed

.

ln is notation used to indicate the process of finding the natural log of a number.

a) True correct incorrect

b) False correct incorrect

* not completed

. The natural log of the odds of something occurring is 4.60 if the probability of something occurring is 0.99.

a) True correct incorrect

b) False correct incorrect

* not completed

. Since the log odds range from negative infinity to positive infinity, the log odds of a dichotomous dependent variable can be substituted into a linear regression prediction equation without any problems.

a) True correct incorrect

b) False correct incorrect

* not completed

. With odds ratios, the odds of something occurring in the first group are expressed as a ratio of the odds of that same thing occurring in the second group.

a) True correct incorrect

b) False correct incorrect

* not completed

. Using percentages to report differences in odds doesn't make sense because odds are non-linear.

a) True correct incorrect

b) False correct incorrect

* not completed

. An odds ratio of 0.25 indicates that the group shown in the ratio's numerator has 75% lower odds of something occurring than the group shown in the ratio's denominator.

a) True correct incorrect

b) False correct incorrect

* not completed

. If the odds ratio of a dummy variable is greater than 1, then the group captured in the dummy variable is predicted to be more likely than the reference group to have something occur.

a) True correct incorrect

b) False correct incorrect