Quiz (Chapter 16)
Bivariate, Multiple, and Logistic
Regression
Introduction
 (T/F) Though regression is related to correlation, the former has
a different purpose than the latter.
 In some studies, regression is used to "predict"; in other studies,
regression is used to "_____."
 What are 3 synonyms for "dependent variable"? What are 2 synonyms
for "independent variable"?
 Excerpt 16.6 has how many predictor variables? How many criterion
variables?
 Which kind of analysis allows for more inferential tests, a correlational
analysis or a regression analysis?
Bivariate Regression
 (T/F) Bivariate regression is used more often to "predict" than to
"explain."
 If a scatter diagram were to be created for Excerpt 16.3, which
variable's name would be used to label the ordinate?
 If bivariate regression is used to predict, predictions are made
of the ____ (independent/dependent) variable.
 In bivariate regression, the "regression line" and the "line of best
fit" ___ (do/don't) refer to the same line.
 In a regression equation, the regression coefficient is represented
by the letter __.
 Using the regression equation on page 374, what would Sam's predicted
"down time" be if his age was 32?
 In Excerpt 16.4, what is the criterion variable?
 (T/F) In standardized regression equations, the beta weight is very
similar to a regression coefficient.
 In a regression equation, the term "a" indicates the regression line's
_____ while "b" indicates its _____.
 A regression equation's ability to predict well is given by r,
for it equals the correlation between Y ' and __.
 r2 indicates the proportion of variability in the ______
that's "explained" by variability in the ______.
 (T/F) In most regression analyses, a, b, and r are viewed as sample
statistics, not population parameters.
 (T/F) If, in Excerpt 16.8, the inferential test had focused on
r (rather than b), the outcome would have been the
same.
 If, in Excerpt 16.7, r had been equal to .80 (rather than
.93), the predictor would have explained __% of the variance in running
performance.
Multiple Regression
 In multiple regression, how many independent variables will there
be? How many dependent variables?
 In Excerpt 16.9, what is/are the numerical value(s) of the regression
coefficient(s)?
 Excerpt 16.11 shows that beta weights ____ (can/cannot)
be compared.
 A standardized regression equation has _____ rather than regression
coefficients, each symbolized as ___.
 If used as an independent variable, _____ would be a dummy variable?
 height
 gender
 income
 What 3 terms are used to distinguish the 3 main types of multiple
regression from one another?
 Does a researcher determine the "entry order" of independent variables
in stepwise multiple regression?
 How does a researcher represent the multiple correlation coefficient?
Is it usually squared or left as is?
 Is it possible to compute DR2
for a simultaneous multiple regression? For a hierarchical multiple
regression?
 Adjusted R2 will be _____ (smaller/larger) than R2
because of anticipated ______.
 In Excerpt 16.16, what was the value of R2 at
the end of Step 2?
 In Excerpt 16.18, how many models were there in the hierarchical
multiple regression?
Logistic Regression
 (T/F) Logistic regression is used exclusively within the fields of
medicine, health, and epidemiology.
 A subset of the independent variables, each called a ______, is typically
included for "______" purposes.
 (T/F) Logistic regression is like bivariate or multiple regression
in that there is only 1 dependent variable.
 Would "mother's age" (measured in years), "length of labor" (measured
in hours), or "baby's sex" (male or female) be the dependent variable
in a logistic regression?
 What are the odds against getting two heads if you flip a
fair coin twice?
 (T/F) The concept of an "odds ratio" is involved in logistic regression
but not in multiple regression.
 If the number 200 in Figure 16.3 had been 400 instead, the OR for
females watching Program B would be __.
 (T/F) The lowest possible OR is 1.00.
 If a logistic regression involves 4 independent variables, how many
odds ratios will there be?
 If an odds ratio is computed for a particular independent variable,
and if that particular OR takes into consideration the other independent
variables, then it is called an "_____" OR.
 In Excerpt 16.25, stroke severity was a significant predictor
because the OR of 1.24 was quite far away, statisticall speaking,
from
the null hypothesis number that was equal to ___ .
 0
 .01
 1.0
 10
 In logistic regression, the main tests of significance focus on _____.
 individual ORs
 the full regression equation
 the slope of the regression line
 An OR equal to __ (what number?) would mean that membership in the
two categories of the dependent variable is unrelated to the independent
variable under consideration.
 What is the name of the test procedure that determines whether a
particular OR is statistically significant?
Final Comments
 The issue of "multicollinearity" arises when ___ (low/high) correlations
exist among independent variables.
 (T/F) In hierarchical multiple regression and logistic regression,
efforts at "control" don't work perfectly.
 (T/F) If important variables are overlooked in a regression analysis,
or if irrelevant variables are included, the regression model is said
to be "imploded."
Click here for answers.
