Path Analysis (Statistical Associates Blue Book Series 22)

An illustrated tutorial and introduction to path analysis using SPSS, AMOS, SAS, or Stata. Suitable for introductory graduate-level study.

The 2014 edition is a major update to the 2012 edition. Among the new features are these:

* Over double the page length (74 pp. rather than 35)

* Over four times as many illustrative figures (45 rather than 11)

* Now covers SAS and Stata as well as SPSS and AMOS

* Covers both regression-based and SEM-based methods

* Totally rewritten and reformatted, with new FAQs added

* Links to download datasets used in the text.

Below is the unformatted table of contents.

PATH ANALYSIS
Overview 6
Key Concepts and Terms 7
Estimation 7
Path model 7
Exogenous and endogenous variables 8
Disturbance terms 8
Causal paths 9
Path Coefficients 9
Path coefficents/weights 9
Path multiplication rule 11
Effect decomposition 11
Path analysis example in OLS regression 13
Overview 13
SPSS regression method 13
SAS regression method 16
Stata regression method 17
Path analysis example in SEM 17
SEM example data 17
SEM in SPSS Amos, SAS, and Stata 18
Path analysis in SPSS AMOS 19
AMOS Inputs 19
AMOS outputs: Path estimates 25
AMOS outputs: Partition of total effects 27
AMOS outputs: Goodness of Fit measures 29
AMOS outputs: Correlations 34
AMOS outputs: Modification indexes 35
Path Analysis in SAS PROC CALIS 36
Overview 36
SAS inputs 37
SAS outputs: Path estimates 39
SAS outputs: Partition of total effects 40
SAS outputs: Goodness of fit measures 41
SAS outputs: Correlations 45
SAS outputs: Modification indices 46
Path Analysis in Stata 47
Overview 47
Stata inputs 48
Stata outputs: Path estimates 49
Stata outputs: Partition of total effects 51
Stata outputs: Goodness of fit measures 53
Stata outputs: Correlations 55
Stata outputs: Modification indices 55
Assumptions 57
Linearity 57
Interval level variables 57
Uncorrelated error 58
Proper specification 59
Normally distributed error 59
Low multicollinearity 60
Identification of the model 60
Recursivity 60
Adequate sample size 60
Frequently Asked Questions 61
Does path analysis confirm causation in a model? 61
Can path analysis be used for exploratory rather than confirmatory purposes? 61
How does the significance of a path coefficient compare with the significance of the corresponding regression coefficient? 62
Why might the direct effect be zero? 62
What is Simpson's paradox? 62
What is Berkson's paradox (case-control bias)? 63
How are path coefficients related to a covariance/correlation matrix for purposes of testing a model? 64
How, exactly, can I compute path coefficients in the regression method? 65
How do I compute the value of the path from an error term to an endogenous variable? 66
How can multiple group path analysis determine if the path model differs across groups in my sample? 66
Could I substitute logistic regression when doing effect decomposition? 67
Can path analysis handle hierarchical/multilevel data? 67
What is partial least squares path analysis? 67
What is the historical origin of path analysis? 67
Bibliography 69
Pagecount: 74