Exam 22: Primer on Structural Equations Modeling

If there are five measured variables,how many unique moments exist?

What technique combines a PCA-based measurement approach with a least-squares based regression approach?

Which technique provides a means for measuring objects in a multidimensional space based on respondents' judgments of how similar different objects are?

Discuss partial least squares estimation and explain when it is particularly useful.

A type of factor analysis in which the researcher specifies the measured variables that indicate specific constructs based on theory is called ____.

Because the χ² test used in SEM is a function of ____,a good model could get rejected just because the amount of power.

Explain the comparative fit index (CFI)and how it is used to determine fit.

A(n)____________________ is a unique piece of information that serves as input to a SEM model.

The latent variable equivalent of dependent variables is called a ____.

The comparative fix index (CFI)provides an index showing the fit of a model relative to a null model.

As the sample size increases in a χ² test,the likelihood of finding a significant result becomes greater.

SEM conventionally uses maximum likelihood estimation.

Structural equation modeling (SEM)is unique in that it is purely an interdependence technique.

Partial least squares (PLS)is particularly useful when the data are messy,difficult to obtain,and/or the measurement quality is not strong enough for use in SEM.

Endogenous constructs are the latent variable equivalent of ____________________ variables.

Which CFI value indicates that a model is a relatively good fit?

Analysts apply SEM when prediction is more important than explanation.

Which of the following is a statistic providing an indication of the average squared residual resulting from fitting an SEM model?

The ____________________ is a statistic providing an indication of the average squared residual resulting from fitting an SEM model.

Which method is particularly likely to be plagued by data anomalies such as implausible results or unstable solutions?