Name:________Ashley Ross________
PSYCH 201 Study Guide for Exam 3
On the exam you will be asked to answer multiple choice, matching, and short answer questions. Multiple
choice questions will require application of the material. Short answers will consist of critical analysis
questions. If you have read chapters 9, 10, 11 & 14 of the text, reviewed the slides, participated in class
activities, and can answer all of the questions on this study guide, you should be well prepared for the exam. To
earn extra credit, turn in this completed study guide before 4:00PM on May 6th. Study guides may be turned in
as a hardcopy or via email.
Chapter 9: Multivariate Research
1. What are longitudinal designs?
They measure the same variables in the same people at different time points
a. Which of the three causal criteria do longitudinal designs help establish?
Temporal precedence
2. List and describe the three correlations related to longitudinal designs:
a. Cross-sectional correlation: test whether two variables, measured at the same time point,
are correlated
b. Autocorrelations: determine the correlation of one variable with itself, measured on two
different occasions
c. Cross-lag correlations: determine whether the earlier measure of one variable is associated
with the later measure of the other variable
3. What are three benefits of longitudinal designs?
a. Examine changes over time
b. Rich information
c. Especially beneficial in developmental research
4. What are three challenges of longitudinal designs?
a. Time consuming
b. Attrition (when people drop out of a study)
c. Costly
5. What is multiple regression?
It is measuring more than two variables.
a. Which of the three causal criteria does multiple regression help establish?
Covariance
6. Define these key terms:
a. Predictor: (Independent variable) The rest of the variables measured in a regression
analysis.
1
, b. Criterion: (Dependent variable) The variable that the researcher is most interested in
understanding or predicting.
7. What does it mean to “control for” a variable?
Controlling for a variable means to focus on one potential third variable.
a. Give an example: Age
b. What other words or phrases are used in the popular press to indicate that multiple regression
was used?
“Taking into account” or “Considering” and “Correcting for” or “Adjusting for”
8. What are betas?
They are similar to r, betas can tell us the strength and direction of the relationship (standardized)
and are used to test for third variables.
9. Why can’t regression establish causation?
Because multiple regression is not a foolproof way to rule out all kinds of third variables.
10. What is pattern and parsimony?
Parsimony is the degree to which a scientific theory provides the simplest explanation of some
phenomenon. To find the simplest explanation for a pattern of data.
Getting a causal claim through measuring variables, using multiple studies, instead of doing an
experiment, look at the weight of the evidence.
11. What is a mediator?
It explains the relation between two variables and why the relation exists.
12. How is a mediator different from a moderator? Be sure you can recognize the difference in an example.
A moderator’s third variables are external to the bivariate correlation while a mediator’s are internal to
the causal variable. The moderators ask “is there a relation?” and a mediator asks “why is there a
relation”. The moderator can change the relationship while the mediator explains the relationship. An
example is, when children view violent content of TV they express more aggressive behavior when there
is no parental discussion about it while with parental discussion there is less aggressive behavior being
expressed (moderator); children express more aggressive behavior when they watch violent content on
TV because they become desensitized to violence (mediator).
13. Discuss multivariate designs and the four big validities:
a. Construct: How well is each variable measured?
b. Internal: What are the third variables? Is there temporal precedence?
c. External: How were the participants sampled (randomly?) and from what population? Is
the sample generalizable?
2
PSYCH 201 Study Guide for Exam 3
On the exam you will be asked to answer multiple choice, matching, and short answer questions. Multiple
choice questions will require application of the material. Short answers will consist of critical analysis
questions. If you have read chapters 9, 10, 11 & 14 of the text, reviewed the slides, participated in class
activities, and can answer all of the questions on this study guide, you should be well prepared for the exam. To
earn extra credit, turn in this completed study guide before 4:00PM on May 6th. Study guides may be turned in
as a hardcopy or via email.
Chapter 9: Multivariate Research
1. What are longitudinal designs?
They measure the same variables in the same people at different time points
a. Which of the three causal criteria do longitudinal designs help establish?
Temporal precedence
2. List and describe the three correlations related to longitudinal designs:
a. Cross-sectional correlation: test whether two variables, measured at the same time point,
are correlated
b. Autocorrelations: determine the correlation of one variable with itself, measured on two
different occasions
c. Cross-lag correlations: determine whether the earlier measure of one variable is associated
with the later measure of the other variable
3. What are three benefits of longitudinal designs?
a. Examine changes over time
b. Rich information
c. Especially beneficial in developmental research
4. What are three challenges of longitudinal designs?
a. Time consuming
b. Attrition (when people drop out of a study)
c. Costly
5. What is multiple regression?
It is measuring more than two variables.
a. Which of the three causal criteria does multiple regression help establish?
Covariance
6. Define these key terms:
a. Predictor: (Independent variable) The rest of the variables measured in a regression
analysis.
1
, b. Criterion: (Dependent variable) The variable that the researcher is most interested in
understanding or predicting.
7. What does it mean to “control for” a variable?
Controlling for a variable means to focus on one potential third variable.
a. Give an example: Age
b. What other words or phrases are used in the popular press to indicate that multiple regression
was used?
“Taking into account” or “Considering” and “Correcting for” or “Adjusting for”
8. What are betas?
They are similar to r, betas can tell us the strength and direction of the relationship (standardized)
and are used to test for third variables.
9. Why can’t regression establish causation?
Because multiple regression is not a foolproof way to rule out all kinds of third variables.
10. What is pattern and parsimony?
Parsimony is the degree to which a scientific theory provides the simplest explanation of some
phenomenon. To find the simplest explanation for a pattern of data.
Getting a causal claim through measuring variables, using multiple studies, instead of doing an
experiment, look at the weight of the evidence.
11. What is a mediator?
It explains the relation between two variables and why the relation exists.
12. How is a mediator different from a moderator? Be sure you can recognize the difference in an example.
A moderator’s third variables are external to the bivariate correlation while a mediator’s are internal to
the causal variable. The moderators ask “is there a relation?” and a mediator asks “why is there a
relation”. The moderator can change the relationship while the mediator explains the relationship. An
example is, when children view violent content of TV they express more aggressive behavior when there
is no parental discussion about it while with parental discussion there is less aggressive behavior being
expressed (moderator); children express more aggressive behavior when they watch violent content on
TV because they become desensitized to violence (mediator).
13. Discuss multivariate designs and the four big validities:
a. Construct: How well is each variable measured?
b. Internal: What are the third variables? Is there temporal precedence?
c. External: How were the participants sampled (randomly?) and from what population? Is
the sample generalizable?
2
