Deck 15: Multivariate Statistical Analysis

Question 1

In factor analysis, communality is a measure of the percentage of variance in each variable that can be explained by all of the factors.

Answer 1

True

Question 2

In factor analysis, the factor scores are a measure of the importance of each variable used to measure each factor.

Answer 2

False

Question 3

One of the purposes of dependence methods of analysis is to group things together.

Answer 3

False

Question 4

In factor analysis, the factor loadings are correlations of each original variable with the factor.

Question 5

Multivariate statistical methods allow us to consider the effects of more than one variable at the same time.

Question 6

Cluster analysis is designed to discover natural groups of cases.

Question 7

The n-way univariate analysis of variance (ANOVA) is appropriate when the dependent variable is metric and the independent variables are nonmetric.

Question 8

Partial correlation analysis allows the measurement of the association between two curvilinear variables after controlling for the effects of other variables.

Question 9

N-way univariate analysis of variance (ANOVA) independently tests for the differences in the mean of a metric dependent variable among two or more nonmetric independent variables.

Question 10

Multivariate dependence methods include logistic regression.

Question 11

Marketing problems are inherently uni-dimensional.

Question 12

Binary logistic regression is appropriate when the dependent variable and the independent variables are nonmetric.

Question 13

In discriminant analysis, both the dependent and the independent variables must be nonmetric.

Question 14

The most common interdependence methods include exploratory factor analysis.

Question 15

Partial correlation can only be applied to one control variable at a time.

Question 16

An underlying assumption of multiple regression is that all the independent variables are dependent of each other.

Question 17

In multiple regression, comparing the coefficients will identify which independent variable has the greatest influence on the dependent variable.

Question 18

In multiple regression, parameter estimates become highly reliable when independent variables are highly correlated with each other.

Question 19

When a researcher attempts to predict a dependent variable from two or more independent variables, the researcher is studying interdependence.

Question 20

In cluster analysis, each cluster should have low internal homogeneity and high external heterogeneity.

Question 21

In cluster analysis, relationships are predicted by defining the composition of the groups.

Question 22

Multicollinearity exists when two or more independent variables are highly correlated with each other, causing the regression estimates to be unreliable.

Question 23

Logistic regression:

A) simultaneously investigates the effects of several independent variables on a dependent variable.
B) is used to predict the probability that an object or individual will belong in one of two mutually exclusive categories.
C) is the measure of association between two linear variables after controlling for the effects of other variables.
D) simultaneously tests for the mean difference of the dependent variable with two or more nonmetric independent variables.

Question 24

Multidimensional scaling is a statistical technique that locates objects in multidimensional space on the basis of measures of the similarity of objects.

Question 25

Which of the following multivariate methods' goals is to achieve parsimony?

A) Multidimensional scaling
B) Factor analysis
C) Cluster analysis
D) Discriminant analysis

Question 26

The purpose of factor analysis is to categorise respondents into mutually exclusive groups.

Question 27

n-way analysis of variance:

A) uncover the effects of several categorical variables on an interval-scaled dependent variable.
B) remove the effects of a third variable on the relationship between two ratio-scaled variables.
C) remove the effects of a third variable on the relationship between two nominal-scaled variables.
D) Investigate simultaneously the effects of several independent variables on a dependent variable.

Question 28

In multidimensional scaling, the labelling of the dimension axis is statistically determined.

Question 29

All of the following are examples of interdependence methods of analysis, except:

A) factor analysis.
B) cluster analysis.
C) multidimensional scaling.
D) discriminant analysis.

Question 30

A problem in multiple regression when the independent variables are correlated with each other, causing the parameters estimates to be unreliable is called:

A) multi-collinearity
B) interdependence.
C) causality.
D) prediction.

Question 31

To remove the effects of a third (or more) variables on the relationship between two interval or ratio-scaled variables is a purpose of:

A) partial correlation technique.
B) n-way cross tabulation technique
C) multiple regression technique.
D) logistic regression technique.

Question 32

All of the following are examples of dependence methods of analysis, except:

A) multiple regression analysis.
B) multiple discriminant analysis.
C) exploratory factor analysis.
D) multivariate analysis of variance.

Question 33

An n-way cross tabulation investigates the effect of several categorical variables on an interval-scaled dependent variable.

Question 34

In hierarchical cluster analysis, objects can leave one cluster and join another, as clusters are being formed until they are finalised.

Question 35

When a multivariate statistical technique is used to predict a dependent variable from two or more independent variables, the researcher is studying:

A) dependence.
B) independence.
C) interdependence.
D) interrelatedness.

Question 36

Binary logistic regression can be used to predict the probability that an individual belongs to one of two mutually exclusive groups.

Question 37

If the analysis contains one dependent variable that is metric and several independent variables that are metric, then the appropriate statistical analysis is:

A) multiple discriminant analysis.
B) conjoint analysis.
C) binary logistic regression.
D) multiple regression analysis.

Question 38

If the analysis contains several independent variables that are nonmetric and a dependent variable that is metric, then the appropriate statistical technique is:

A) multiple regression analysis.
B) multiple discriminant analysis.
C) conjoint analysis.
D) n-way univariate ANOVA.

Question 39

Hierarchical cluster analysis works very well when dealing with a relatively large number of cases.

Question 40

When using interdependence methods of analysis, no one variable is to be predicted from or explained by the other variables.

Question 41

If the regression equation is: Y = 12.43 + 0.64X1 + 0.25X2, then the predicted value for Y when X1 = 2 and X2 = 4 is approximately:

A) 10.15.
B) 12.71.
C) 15.49.
D) 14.71.

Question 42

In multiple regression, the coefficients β1, β2, and so on, are known as:

A) correlation coefficients.
B) standardised coefficients.
C) coefficients of multiple determination.
D) coefficients of partial regression.

Question 43

In order to remove the effects of a third (or more) variable on the relationship between two interval- or ratio-scaled variables, which of the following techniques is appropriate?

A) n-way analysis of variance
B) Partial correlation
C) Logistic regression
D) n-way cross tabulation

Question 44

In cluster analysis, the researcher wants clusters to have high ___________ within clusters and high between-cluster ____________.

A) homogeneity; homogeneity
B) heterogeneity; heterogeneity
C) heterogeneity; homogeneity
D) homogeneity; heterogeneity

Question 45

If the analysis contains several independent variables that are nonmetric and a dependent variable that is nonmetric then the appropriate statistical technique is:

A) multiple regression analysis.
B) multiple discriminant analysis.
C) n-way univariate ANOVA.
D) n-way cross tabulation.

Question 46

The goal of factor analysis is to reduce a large number of variables to as few dimensions or constructs as possible. This is known as:

A) prudence.
B) munificence.
C) parsimony.
D) summation.

Question 47

Which of the following is not a dependence method of analysis?

A) Multidimensional scaling
B) Multiple regression analysis
C) Discriminant analysis
D) N-way cross tabulation

Question 48

In factor analysis, which of the following provides a means for interpreting and labelling the factors?

A) Factor scores
B) Eigenvalues
C) Factor loadings
D) Communalities

Question 49

When several independent variables are highly correlated with each other, the effect is known as:

A) parsimony.
B) multicollinearity.
C) communality.
D) partiality.

Question 50

Which of the following statements regarding cluster analysis is not true?

A) Cluster analysis is used to seek constructs that underlie the variables.
B) The groups are not predefined in cluster analysis.
C) Cluster analysis does not predict relationships.
D) Cluster analysis requires the data to be interval-scaled.

Question 51

In multiple regression, to discover which independent variable has the greatest influence:

A) is to consult the coefficients to find the largest number.
B) is to consult the standardised coefficients to determine to largest number.
C) is to divide each corresponding coefficient by its standard error to determine the largest number.
D) is to consult the variance inflation factors (VIF) to find the largest number.

Question 52

If the coefficient of multiple determination is 0.532, then the variation in the independent variables accounts for approximately ______ per cent of the variance in the dependent variable.

A) 73
B) 53
C) 28
D) 27

Question 53

In factor analysis, communalities are:

A) correlations of the original variables with the factor.
B) measures of the percentage of variance in each variable explained by all of the factors.
C) sums of the squared factor loadings for each factor.
D) scores representing each observation's calculated value on each factor.

Question 54

If a researcher wishes to discover natural groups of variables, he or she should use:

A) factor analysis.
B) cluster analysis.
C) discriminant analysis.
D) logistic regression analysis.

Question 55

A variable that is coded as either zero or one and has two distinct levels is called a(n):

A) regression variable.
B) dummy variable.
C) continuous variable.
D) eigenvalue.

Question 56

Which of the following is an appropriate statistical technique to use in order to uncover the effect of several categorical variables on an interval-scaled dependent variable?

A) Logistic regression
B) Partial correlation
C) n-way crosstabulation
D) n-way analysis of variance

Question 57

Which of the following is an appropriate technique to use in order to predict the probability that an object or individual will belong in one of two mutually exclusive categories, based on several independent variables?

A) Multiple regression analysis
B) Logistic regression analysis
C) Multiple discriminant analysis
D) Multidimensional scaling

Question 58

Which of the following technique refers to a summarisation of information contained in a large number of variables into a smaller number of factors?

A) Cluster analysis
B) Exploratory factor analysis
C) Metric multidimensional scaling
D) Non-metric multidimensional scaling

Question 59

When a researcher is attempting to predict sales volume by using building permits, amount of advertising and the income levels of residents, he or she is using:

A) factor analysis.
B) logistic regression analysis.
C) multiple regression analysis.
D) multiple discriminant analysis.

Question 60

In an n-way cross tabulation, the independent variable(s) are ______ and the dependent variable(s) are ______.

A) metric; nonmetric
B) metric; metric
C) nonmetric; metric
D) nonmetric; nonmetric

Question 61

To facilitate market segmentation by identifying subjects or individuals who have similar needs, lifestyles or responses to marketing strategies, which of the following statistical procedures would be appropriate?

A) Multidimensional scaling
B) Discriminant analysis
C) Cluster analysis
D) Factor analysis

Question 62

A type of graphical representation that helps the researcher to decide how many groups are appropriate is called a:

A) perceptual map.
B) dendogram.
C) crosstabulation.
D) correlation matrix.

Question 63

In a prestige car study, respondents were asked to rate the similarity of a BMW to a Mercedes, a Jaguar and other vehicles. Which statistical procedure would you recommend to analyse the data collected?

A) Multidimensional scaling
B) Discriminant analysis
C) Cluster analysis
D) Conjoint analysis

Question 64

Which of the following is not an interdependence method of analysis?

A) Exploratory factor analysis
B) Discriminant analysis
C) Cluster analysis
D) Multidimensional scaling

Question 65

Which of the following statements about multidimensional scaling is not true?

A) It shows how objects are perceived by consumers through a visual display.
B) The perceptual difference among objects is reflected in the relative distance among objects in the multidimensional space.
C) It provides a means of measuring objects in multidimensional space on the basis of measures of the similarity of objects.
D) The labelling of the dimension axis is statistically determined.

Question 66

If a researcher wishes to discover natural groups of cases, then he or she should use:

A) factor analysis.
B) cluster analysis.
C) conjoint analysis.
D) discriminant analysis.