Advanced and Multivariate Statistical Methods: Practical Application and Interpretation

Advanced and Multivariate
Statistical Methods
Practical Application and Interpretation
Sixth Edition
Craig A. Mertler and Rachel Vannatta Reinhart
© 2017 Taylor & Francis

Chapter 2
A Guide to Multivariate Techniques

Multivariate (and Other) Techniques
• Categories of statistical tests

o Degree of relationship among variables
o Significance of group differences
o Prediction of group membership
o Structure

Degree of Relationship Among Variables
• Bivariate correlation and regression

o Evaluate the degree of relationship between two quantitative variables
o Pearson r is the most commonly used measure of bivariate relationship (without
distinguishing between the IV and DV)
o Bivariate regression uses the relationship between the IV and DV to predict
scores on the DV from knowledge of scores on the IV

• Multiple regression
o Identifies the best combination of predictor variables (IVs) of a single DV
o Appropriate research situation involves several quantitative IVs and one
quantitative DV

• Path analysis
o Utilizes multiple applications of multiple regression in order to estimate causal
relationships (direct and indirect) among several variables and to test the
acceptability of a causal model hypothesized by the researcher
o A complex analytical technique, due to the fact that there are numerous and
subsequent multiple regressions being conducted, all of which contribute to the
development and testing of the model

Significance of Group Differences
• t Test
o Most basic group comparison test; used to analyze significant differences
between two group means
o Appropriate when the IV is defined by having two categories and the DV is
quantitative

• One-way analysis of variance (ANOVA)

o Tests the significance of group differences between two or more means, while
also analyzing the variation in between and within each group
o Appropriate when the IV has two or more categories and the DV is quantitative
• One-way analysis of covariance (ANCOVA)

o Similar to ANOVA in that two or more groups are being compared on the mean of
a quantitative DV, but adds a control variable, known as a covariate
o The covariate is controlled because it may potentially influence the DV

• Factorial analysis of variance

o Extension of ANOVA to research scenarios with two or more IVs that are
categorical and one quantitative DV
• Factorial analysis of covariance

o An extension of factorial ANOVA where group differences are examined on a
single quantitative DV based upon two or more categorical IVs, while controlling
for a covariate that may influence the DV

• One-way multivariate analysis of variance (MANOVA)

o Similar to ANOVA, but utilized to simultaneously study two or more related DVs
while controlling for correlations among those DVDs
o Appropriate in situations involving only one IV with two or more categories
• One-way multivariate analysis of covariance (MANCOVA)

o An extension of ANCOVA where group differences are being investigated among
the several DVs while also controlling for covariates that may influence the DVs
o Appropriate in research scenarios involving multiple DVs and one IV with two or
more categories
• Factorial multivariate analysis of variance

o Extension of MANOVA to scenarios with two or more IVs that are categorical,
along with two or more quantitative DVs
• Factorial multivariate analysis of covariance
o Extension of factorial MANOVA to scenarios that require control of one or more
covariates on the DV

Prediction of Group Membership
• Discriminant analysis
o Sometimes seen as the reverse of MANOVA
o Attempts to identify a combination of quantitative IVs that best predicts group
membership as defined by a single DV that has two or more categories

Prediction of Group Membership
• Logistic regression
o Similar to discriminate analysis in that both identify a set of IVs that best predicts
group membership
o However, logistic regression is appropriate in situations where the DV is a
dichotomous variable

Underlying Structure
• Factor analysis and principal components analysis

o Techniques used to explain or describe in the underlying structure of an
instrument or to possibly reduce the number of IVs resulting from the
administration of an instrument
o Factor analysis allows a researcher to explore the underlying structure of an
instrument or data set; the underlying structures are typically referred to as latent
factors (unobservable variables, traits, or characteristics)
o Principal components analysis is typically used to reduce the number of IVs,
which is often advantageous when conducting multivariate techniques where
numerous IVs may be highly correlated
Statistical Decision-Making Tools
• Table of Statistical Tests (Fig. 2.1)

o Begin by identifying the variables in the research question
o Indicate which variables are the IVs, DVs, and covariates
o Determine whether each variable is categorical or quantitative; if categorical,
determine the number of categories
o Then, use the table to identify:
§ The appropriate role for the IVs,
§ The appropriate column for the DVs, and
§ The row and column intersection that indicates the statistical test to be used
Statistical Decision-Making Tools
• Decision-Making Tree for Statistical Tests (Fig. 2.2)

o Begin by identifying the variables in the research question
o Indicate which variables are the IVs, DVs, and covariates
o Determine whether each variable is categorical or quantitative; if categorical,
determine the number of categories
o Determine the purpose of the research question
o Apply the information from the preceding steps to the Tree following the process
of decisions in order to identify the appropriate statistical test

Advanced and Multivariate Statistical Methods: Practical Application and Interpretation

Uploaded by

Copyright:

Available Formats

Advanced and Multivariate Statistical Methods: Practical Application and Interpretation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advanced and Multivariate Statistical Methods: Practical Application and Interpretation

Uploaded by

Copyright:

Available Formats

Advanced and Multivariate

Craig A. Mertler and Rachel Vannatta Reinhart

© 2017 Taylor & Francis

© 2017 Taylor & Francis

• Categories of statistical tests

© 2017 Taylor & Francis

• Bivariate correlation and regression

© 2017 Taylor & Francis

© 2017 Taylor & Francis

© 2017 Taylor & Francis

© 2017 Taylor & Francis

• One-way analysis of variance (ANOVA)

• One-way analysis of covariance (ANCOVA)

© 2017 Taylor & Francis

• Factorial analysis of variance

• Factorial analysis of covariance

© 2017 Taylor & Francis

• One-way multivariate analysis of variance (MANOVA)

• One-way multivariate analysis of covariance (MANCOVA)

• Factorial multivariate analysis of variance

© 2017 Taylor & Francis

© 2017 Taylor & Francis

© 2017 Taylor & Francis

• Factor analysis and principal components analysis

• Table of Statistical Tests (Fig. 2.1)

• Decision-Making Tree for Statistical Tests (Fig. 2.2)

You might also like