Question 1

Which of the following statements is false:
A) Correspondence analysis (CA) can be used for research on the Internet.
B) CA uses the same distance formula as MDS or cluster analysis.
C) CA provides a perceptual map like MDS but the data can come from checklists.
D) CA provides a perceptual map like MDS but the points represent market shares.
E) both c and d are false

Accepted Answer

Correspondence analysis (CA) provides a perceptual map where the points represent the relationships between categories of variables, not market shares.

Question 2

Which of the following statements is true:
A) Correspondence analysis (CA) tracks the flow of correspondence, and these days is mostly used to analyze patterns of email.
B) CA is used to track the correspondence between consumers' attitudes and their behaviors.
C) CA is a mapping technique.
D) CA maps consumer preferences to market segments.
E) None of these statements are true.

Accepted Answer

Correspondence Analysis (CA) is indeed a mapping technique used in statistics to graphically display the relationships among categorical variables in a dataset. It does not specifically track email patterns, consumer attitudes directly to behaviors, or map consumer preferences to market segments in the ways described in the other options.

Question 3

Structural equations models:
A) are like regression in that some variables predict others.
B) are like factor analysis in that multiple indicator variables may be included.
C) are like discriminant analysis in that group differences are modeled.
D) a and b
E) a, b and c

Accepted Answer

Structural equation models (SEMs) are similar to regression models because they involve predicting one variable from others, and they are also similar to factor analysis because they often include multiple indicator variables to represent latent constructs. However, SEMs are not specifically designed to model group differences like discriminant analysis, making options A and B correct.

Question 4

Structural equations models:
A) are the best tool to study customer loyalty.
B) can be modeled in Excel, if you specify the regression statements properly.
C) are like regressions but the regression equation is more "structured."
D) is a mapping technique.
E) can fit models where z=fn(y) and y=fn(x).

Accepted Answer

Structural equation models (SEMs) are capable of fitting complex models where variables can be both independent and dependent in different parts of the model, such as z being a function of y and y being a function of x. This allows for the examination of indirect effects and the modeling of more complex relationships than simple regression models.

Question 5

Neural networks:
A) are believed to describe fairly accurately neural transmissions in the brain.
B) require assumptions of multivariate normality.
C) require large data bases.
D) require ratio level data.
E) require the user to specify the beta-weights.

Accepted Answer

Neural networks are known for their ability to handle and learn from large datasets, which is crucial for their performance and accuracy. They do not accurately describe neural transmissions in the brain, do not require assumptions of multivariate normality, do not specifically require ratio level data, and do not require the user to specify the beta-weights, as these are learned during the training process.

Question 6

Which of the following is not true:
A) neural networks can learn relationships among variables
B) neural networks can be tested in cross-validation
C) neural nets have few methodological assumptions
D) neural networks are like social networks but they map cognitive connections, not social ones
E) neural networks come out of expert systems

Accepted Answer

Neural networks are computational models inspired by the human brain's structure and function, not models of cognitive connections in the context of social networks. They are used for pattern recognition and predictive modeling among other applications, not for mapping social connections.

Question 7

Which of the following statements is true?
A) Social networks model connections between variables.
B) Social networks model connections between employees.
C) Social networks could be used to model the word-of-mouth communications among consumers.
D) a and b
E) b and c

Accepted Answer

Social networks can indeed model connections between employees as well as be used to model word-of-mouth communications among consumers. They are not limited to modeling connections between variables; that description fits more with statistical or mathematical models.

Question 8

Which statement is false:
A) Social networks finds actors who are central to the network.
B) Social networks find cliques and coalitions in networks.
C) Social networks describe connections between people, but unfortunately do not describe the qualities of the people very well.
D) Social networks can be used to describe the relationships between people, or relationships between organizations.
E) Social networks could be used to describe interactions between CEOs and their VPs.

Accepted Answer

Social networks not only describe connections between individuals or entities but also can provide insights into the qualities of the people involved through analysis of their interactions, network positions, and other attributes.

Quiz 18: Appendix: More Multivariate Statistical Techniques

Which of the following statements is false:

Which of the following statements is true:

Structural equations models:

Structural equations models:

Neural networks:

Which of the following is not true:

Which of the following statements is true?

Which statement is false: