effects.sem: Total, Direct, and Indirect Effects for Structural Equation Models

Description

The sem method for the standard generic function effects computes total, direct, and indirect effects for a fitted structural equation model according to the method described in Fox (1980).

Usage

# S3 method for sem
effects(object, ...)
# S3 method for msem
effects(object, ...)
# S3 method for semeffects
print(x, digits = getOption("digits"), ...)
# S3 method for semeffectsList
print(x, digits = getOption("digits"), ...)

Arguments

object

a fitted structural-equation model object produced by the sem function.

an object of class semeffects or semeffectsList, produced by effects.

digits

digits to print.

…

not used.

Value

effect.sem returns an object of class semeffects with Total, Direct, and Indirect elements.

References

Fox, J. (1980) Effect analysis in structural equation models: Extensions and simplified methods of computation. Sociological Methods and Research 9, 3--28.

Examples

Run this code

# NOT RUN {
    
# }
# NOT RUN {
# These examples are from Fox (1980)

# In the first pair of examples, readMoments() and specifyModel() read from the
# input stream. These examples cannot be executed via example() but can be entered
# at the command prompt. The Blau and Duncan example is repeated using file input;
# this example can be executed via example(). 

# The recursive Blau and Duncan basic stratification model:
#  x1 is father's education, x2 father's SES, y3 respondent's education,
#  y4 SES of respondent's first job, y5 respondent's SES in 1962

R.bd <- readMoments(names=c("x1", "x2", "y3", "y4", "y5"))
1
.516 1
.453 .438 1
.332 .417 .538 1
.322 .405 .596 .541 1

mod.bd <- specifyModel()
y3 <- x1, gam31
y3 <- x2, gam32
y4 <- x2, gam42
y4 <- y3, beta43
y5 <- x2, gam52
y5 <- y3, beta53
y5 <- y4, beta54

sem.bd <- sem(mod.bd, R.bd, N=20700, fixed.x=c("x1", "x2"))
summary(sem.bd)
effects(sem.bd)


# The nonrecursive Duncan, Haller, and Portes peer-influences model for observed variables:

R.DHP <- readMoments(diag=FALSE, names=c("ROccAsp", "REdAsp", "FOccAsp", 
"FEdAsp", "RParAsp", "RIQ", "RSES", "FSES", "FIQ", "FParAsp"))
.6247     
.3269  .3669       
.4216  .3275  .6404
.2137  .2742  .1124  .0839
.4105  .4043  .2903  .2598  .1839
.3240  .4047  .3054  .2786  .0489  .2220
.2930  .2407  .4105  .3607  .0186  .1861  .2707
.2995  .2863  .5191  .5007  .0782  .3355  .2302  .2950
.0760  .0702  .2784  .1988  .1147  .1021  .0931 -.0438  .2087

model.dhp <- specifyModel()
RIQ      -> ROccAsp, gam51,  NA
RSES     -> ROccAsp, gam52,  NA
FSES     -> FOccAsp, gam63,  NA
FIQ      -> FOccAsp, gam64,  NA
FOccAsp  -> ROccAsp, beta56, NA
ROccAsp  -> FOccAsp, beta65, NA
ROccAsp <-> ROccAsp, ps55,   NA
FOccAsp <-> FOccAsp, ps66,   NA
ROccAsp <-> FOccAsp, ps56,   NA


# Note: The following generates a warning because not all of the variables
#       in the correlation matrix are used
sem.dhp <- sem(model.dhp, R.DHP, 329,
                fixed.x=c('RIQ', 'RSES', 'FSES', 'FIQ'))
summary(sem.dhp)
effects(sem.dhp)
    
# }
# NOT RUN {
    
## the following example may be executed via example()

etc <- system.file(package="sem", "etc") # path to data and model files

# The recursive Blau and Duncan basic stratification model:
#  x1 is father's education, x2 father's SES, y3 respondent's education,
#  y4 SES of respondent's first job, y5 respondent's SES in 1962

(R.bd <- readMoments(file=file.path(etc, "R-Blau-Duncan.txt"),
					names=c("x1", "x2", "y3", "y4", "y5")))
(mod.bd <- specifyModel(file=file.path(etc, "model-Blau-Duncan.txt")))
sem.bd <- sem(mod.bd, R.bd, N=20700, fixed.x=c("x1", "x2"))
summary(sem.bd)
effects(sem.bd)
# }