Deck 12: Factor Analysis, Path Analysis, and Structural Equation Modeling

Question 1

Describe the purposes of factor analysis, distinguishing between exploratory and confirmatory factor analysis.

Answer 1

No Answer

Question 2

For what two purposes can exploratory factor analysis be used?

Answer 2

No Answer

Question 3

Describe a study that would be analyzed using exploratory factor analysis. What would the goals of that analysis be?

Answer 3

No Answer

Question 4

List two threats to the validity of correlational research that also can threaten the validity of the results of an exploratory factor analysis.

Question 5

Describe the preliminary analyses that should be completed before an exploratory factor analysis is conducted. What would the researcher look for in those analyses?

Question 6

What is factor extraction? What determines the order in which factors are extracted in an exploratory factor analysis?

Question 7

What is factor rotation? Name and define the two general categories of factor rotation. Under what conditions will those two types of rotation produce similar results?

Question 8

Describe the process a researcher would use to determine how many factors underlie a set of correlations.

Question 9

What are factor eigenvalues and how are they interpreted?

Question 10

What is the purpose of creating a scree plot in exploratory factor analysis?

Question 11

Define parallel analysis and explain why a researcher would conduct this type of analysis.

Question 12

Describe how a researcher would use the matrix of factor loadings produced by an exploratory factor analysis.

Question 13

Describe the two ways factor scores can be created from the results of an exploratory factor analysis.

Question 14

Describe a study that would be analyzed using confirmatory factor analysis. What would the goals of that analysis be?

Question 15

Explain how a confirmatory factor analysis can be used to test theoretical hypotheses.

Question 16

Explain how a confirmatory factor analysis can be used to assess the structural validity of a measure.

Question 17

Explain how a confirmatory factor analysis can be used to test the generalizability of a measure.

Question 18

What is goodness of fit and how is it assessed in a confirmatory factor analysis?

Question 19

Name and define the three approaches to testing mediational hypotheses.

Question 20

Describe how the causal steps strategy can be used to test mediational hypotheses.

Question 21

Describe how a path analysis can be used to test mediational hypotheses.

Question 22

What is structural equation modeling? Why would a researcher use this type of statistical analysis?

Question 23

Describe a research study that could be analyzed using structural equation modeling.

Question 24

What is prospective research? Why would a researcher use this type of statistical analysis?

Question 25

Describe two limitations of path analysis and structural equation modeling.

Question 26

Explain why it would be important to examine whether there are any curvilinear or nonadditive relationships between the variables you plan to analyze using path analysis or structural equation modeling.

Question 27

Explain why omitting relevant variables from a path analysis or structural equation model can be problematic.

Question 28

Explain why it is important to test both preferred and plausible alternative models when analyzing data using path analysis and structural equation modeling.

Question 29

List two ways alternative models can be ruled out in path analysis and structural equation modeling other than directly testing those models.

Question 30

Factor analysis is a

A) form of experimental design in which the experimenter manipulates two or more independent variables but does not propose any specific hypotheses.
B) statistical technique that can be applied to a set of variables to identify clusters of variables that correlate with each other but that are relatively uncorrelated with the variables in the other clusters.
C) statistical analysis used to test hypotheses about which dimensions the items in a set of items will load on.
D) study in which the independent variable is manipulated but research participants are not randomly assigned to conditions of the independent variable.

Question 31

In factor analysis, a factor is

A) another term for independent variable.
B) equivalent to a predictor variable in multiple regression analysis.
C) a dimension that underlies a set of inter-correlated variables.
D) a manifest variable used to organize scores on a set of inter-correlated variables.

Question 32

Exploratory factor analysis can be used to

A) summarize the pattern of correlations among a set of variables.
B) condense a large number of variables into a few variables to simplify data analysis.
C) examine the structural validity of a new measure.
D) do all of the above.

Question 33

Exploratory factor analysis can be used to

A) test hypotheses about the relationships among a set of variables.
B) condense a large number of variables into a few variables to simplify data analysis
C) examine the generalizability of a measure.
D) do any of the above.

Question 34

If your goal is to condense scores on a large number of dependent measures into a smaller set of variables in order to simply data analysis, the best approach to take is

A) principle components analysis.
B) common factors analysis.
C) confirmatory factor analysis.
D) parallel analysis.

Question 35

Which of the following statements about exploratory factor analysis is TRUE?

A) Multicollinearity is a problem for exploratory factor analysis just as it is for multiple regression analysis.
B) Participant sample size is not an issue in exploratory factor analysis.
C) Unlike multiple regression analysis, it is desirable to have several outliers in the data for an exploratory factor analysis.
D) There is only one generally accepted way to conduct an exploratory factor analysis.

Question 36

In exploratory factor analysis, factor extraction

A) determines the number of factors to be interpreted.
B) creates an initial set of factors from the raw data.
C) clarifies the content of factors following factor extraction.
D) identifies and eliminates outliers from the data set.

Question 37

In exploratory factor analysis, factor rotation

A) determines the number of factors to be interpreted.
B) creates an initial set of factors from the raw data.
C) clarifies the content of factors following factor extraction.
D) identifies and eliminates outliers from the data set.

Question 38

Which of the following statements about orthogonal rotation in exploratory factor analysis is TRUE?

A) Although orthogonal rotation is a classic technique, it is rarely used any more.
B) Orthogonal rotation creates a set of factors that can be correlated with one another if that is consistent with the data.
C) Orthogonal rotation is best used when there are very few strong correlations in the data set.
D) Orthogonal rotation creates a set of factors that are uncorrelated with one another.

Question 39

Which of the following statements about oblique rotation in exploratory factor analysis is TRUE?

A) Although oblique rotation is a classic technique, it is rarely used any more.
B) Oblique rotation creates a set of factors that can be correlated with one another if that is consistent with the data.
C) Oblique rotation is best used when multicollinearity might be an issue in the data set.
D) Oblique rotation creates a set of factors that are uncorrelated with one another.

Question 40

In factor analysis, an eigenvalue represents the

A) number of factors that should be interpreted.
B) extent to which outliers might be influencing the results.
C) amount of variance in the data accounted for by a given factor.
D) statistical significance of a given factor.

Question 41

A scree plot is used to

A) determine the order in which factors should be extracted.
B) determine the appropriate number of factors to interpret.
C) visually portray the correlations between extracted factors.
D) visually portray the results of a factor analysis.

Question 42

Parallel analysis is used to

A) determine the appropriate number of factors to interpret.
B) determine the order in which factors should be extracted.
C) calculate the correlations between extracted factors.
D) calculate factor scores.

Question 43

________ is generally considered to be the soundest way to decide how many factors to interpret once an exploratory factor analysis has been conducted.

A) Parallel analysis
B) The number of factors with eigenvalues greater than one
C) Interpretation of the scree plot of eigenvalues
D) Conducting multiple analyses using different methods and averaging the results

Question 44

Which of the following statements about factor loadings in exploratory factor analysis is TRUE?

A) Lower factor load?ings indicate items with stronger associations with the factor.
B) Parallel analysis is a statistical technique for determining which factor loadings are meaningful.
C) Factor loadings represent the correlation of each item with the underlying factor.
D) Negative factor loadings indicate a problem with the factorability of the data set.

Question 45

Which of the following statements about factor loadings in exploratory factor analysis is/are TRUE?

A) Higher factor load?ings indicate items with stronger associations with the factor.
B) Items with nega?tive factor loadings are usually ones that are worded in the opposite direction of items with positive loadings.
C) It is possible for an item to load on more than one factor.
D) All of the above are true.

Question 46

The combined score for the items loading on a factor is called the research participant's

A) factor loading.
B) factor score.
C) eigenvalue.
D) extracted score.

Question 47

In confirmatory factor analysis, the researcher

A) hypothesizes which dimensions the items in a set will load on and uses statistical analyses to test that hypothesis.
B) examines a set of item scores to see what dimensions underlie the items and which items load on which dimensions.
C) conducts a study with two independent variables to test the hypothesis that one independent variable moderates the effects of the other.
D) examines a set of independent variables to determine which of those variables best predict one or more dependent variables.

Question 48

Confirmatory factor analysis can be used to test

A) hypotheses about the relationships among a set of variables.
B) the structural validity of a measure.
C) the generalizability of a measure.
D) any of the above.

Question 49

Confirmatory factor analysis is most likely be used to

A) summarize the pattern of correlations among a set of variables.
B) test hypotheses about the relationship among a set of variables.
C) examine the reliability of a new measure.
D) condense a large number of variables into a few variables to simplify data analysis.

Question 50

An observed variable is

A) directly measured.
B) represented only by a factor score.
C) estimated statistically, but not directly measured.
D) used only in exploratory factor analysis.

Question 51

A latent variable is

A) another term for a measured or observed variable.
B) the square of a variable's loading on a factor in exploratory factor analysis.
C) a variable that is not directly measured but which is estimated statistically from scores on measured variables.
D) an index of goodness of fit used in confirmatory factory analysis.

Question 52

Latent variables are

A) measured directly, such as by having research participants fill out a set of questionnaires.
B) derived from a set of observed scores through the use of confirmatory factor analysis.
C) components of measured variables that generalize across participant populations.
D) the sums of participants' scores on the set of variables loading on a given factor in a factor analysis.

Question 53

In confirmatory factor analysis, the concept of goodness of fit represents

A) how well a hypothesized factor structure matches the actual factor structure of the data set.
B) the number of factors with eigenvalues greater than one.
C) how well any one measured variable represents a given latent variable.
D) the reliability of a latent variable.

Question 54

Which of the following statements about goodness of fit is TRUE?

A) The root mean square error of approximation (RMSEA) is the one best statistic for assessing goodness of fit.
B) There are a number of ways of assessing goodness of fit, each of which has its strengths and limitations.
C) Although goodness of fit is an important theoretical concept, there is no way to assess it in practice.
D) In evaluating the validity of measures, goodness of fit is relevant to questions of structural validity but not to questions of generalizability.

Question 55

The hypothesis that variable C mediates the relationship between variables A and B can be tested using

A) the causal steps strategy.
B) path analysis.
C) structural equation modeling,
D) any of the above depending on the nature of the data.

Question 56

Which of the following statements about testing mediational models is FALSE?

A) The techniques available for testing mediational models cannot be used when one or more of the variables involved is manipulated experimentally.
B) The causal steps approach is best suited for simple, three-variable mediational models.
C) Path analysis can be used to test mediational models involving more than three variables.
D) Structural equation modeling is a technique for testing mediational models using latent variables.

Question 57

Path analysis is a method for

A) determining causal relationships using correlational data.
B) conducting complex exploratory factor analyses.
C) testing mediational hypotheses.
D) evaluating goodness of fit for confirmatory factor analyses.

Question 58

Structural equation modeling is a

A) way of testing mediational hypotheses using observed variables.
B) way of testing mediational hypotheses using latent variables.
C) form of exploratory factor analysis using latent variables.
D) technique for determining the causal relationships among a set of latent variables.

Question 59

Which of the following statements about prospective research is/are TRUE?

A) It allows researchers to draw firm causal conclusions from correlational data by testing the time precedence of a proposed cause relative to its proposed effect.
B) It allows researchers to draw firm causal conclusions from correlational data by controlling the third variable problem.
C) It examines the correlation of a hypothesized cause at Time 1 with its hypothesized effect at Time 2 to investigate the time precedence of a possible causal variable.
D) All of the above are true.

Question 60

__________ examines the correlation between a hypothesized cause at Time 1 with its hypothesized effect at time 2.

A) Latent variable analysis
B) Confirmatory factor analysis
C) Prospective research
D) Alternative analysis modeling

Question 61

Which of the following statements about path analysis and structural equation modeling is/are TRUE?

A) If all relevant variables are not included in the model, then the results of the analysis might be misleading.
B) In general, path models and structural equation models do not take interactions between variables into account.
C) There can be more than one model that will fit a given data set equally well.
D) All of the above are true.

Question 62

You hypothesize that gender of research participant MODERATES the relationship between peer pressure to conform and conformity to group norms. Of the following designs, which would be the most appropriate to use to test your hypothesis?

A) one-way analysis of variance (ANOVA)
B) path analysis
C) quasi-experiment
D) factorial ANOVA

Question 63

People who score higher on empathy tend to score low on prejudice; women also tend to have lower prejudice scores than men. You hypothesize that empathy MEDIATES the relationship between gender and prejudice. Of the following designs, which would be the most appropriate to use to test your hypothesis?

A) one-way analysis of variance (ANOVA)
B) path analysis
C) quasi-experiment
D) factorial ANOVA