Path Analysis

Overview
Path analysis is an extension of the regression model, used to test the fit of the
correlation matrix against two or more causal models which are being compared
by the researcher. The model is usually depicted in a circle-and-arrow figure in
which single-headed arrows indicate causation. A regression is done for each
variable in the model as a dependent on others which the model indicates are
causes. The regression weights predicted by the model are compared with the
observed correlation matrix for the variables, and a goodness-of-fit statistic is
calculated. The best-fitting of two or more models is selected by the researcher as
the best model for advancement of theory.
Path analysis requires the usual assumptions of regression. It is particularly
sensitive to model specification because failure to include relevant causal
variables or inclusion of extraneous variables often substantially affects the path
coefficients, which are used to assess the relative importance of various direct and
indirect causal paths to the dependent variable. Such interpretations should be
undertaken in the context of comparing alternative models, after assessing their
goodness of fit discussed in the section on structural equation modeling (SEM
packages are commonly used today for path analysis in lieu of stand-alone path
analysis programs). When the variables in the model are latent variables measured
by multiple observed indicators, path analysis is termed structural equation
modeling, treated separately. We follow the conventional terminology by which
path analysis refers to modeling single-indicator variables.

Key Concepts and Terms

o

Estimation. Note that path estimates may be calculated by OLS regression or by MLE maximum likelihood
estimation, depending on the computer package. Two-Stage Least Squares (2SLS), discussed separately, is
another path estimation procedure designed to extend the OLS regression model to situations where nonrecursivity is introduced because the researcher must assume the covariances of some disturbance terms are
not 0 (this assumption is discussed below). Click here for a separate discussion of 2SLS.
Path model. A path model is a diagram relating independent, intermediary, and dependent variables. Single
arrows indicate causation between exogenous or intermediary variables and the dependent(s). Arrows also
connect the error terms with their respective endogenous variables. Double arrows indicate correlation
between pairs of exogenous variables. Sometimes the width of the arrows in the path model are drawn in a
width which is proportional to the absolute magnitude of the corresponding path coefficients (see below).
Causal paths to a given variable include (1) the direct paths from arrows leading to it, and (2) correlated
paths from endogenous variables correlated with others which have arrows leading to the given variable.
Consider this model:

This model has correlated exogenous variables A, B, and C, and endogenous variables D and E. Error terms
are not shown. The causal paths relevant to variable D are the paths from A to D, from B to D, and the paths
reflecting common anteceding causes -- the paths from B to A to D, from C to A to D, and from C to B to D.
Paths involving two correlations (C to B to A to D) are not relevant. Likewise, paths that go backward (E to
B to D, or E to B to A to D) reflect common effects and are not relevant.

Exogenous and endogenous variables. Exogenous variables in a path model are those with no explicit
causes (no arrows going to them, other than the measurement error term). If exogenous variables are
correlated, this is indicated by a double-headed arrow connecting them. Endogenous variables, then, are those
which do have incoming arrows. Endogenous variables include intervening causal variables and dependents.
Intervening endogenous variables have both incoming and outgoing causal arrows in the path diagram. The
dependent variable(s) have only incoming arrows.
Path coefficient/path weight. A path coefficient is a standardized regression coefficient (beta) showing the
direct effect of an independent variable on a dependent variable in the path model. Thus when the model has
two or more causal variables, path coefficients are partial regression coefficients which measure the extent of
effect of one variable on another in the path model controlling for other prior variables, using standardized
data or a correlation matrix as input. Recall that for bivariate regression, the beta weight (the b coefficient for
standardized data) is the same as the correlation coefficient, so for the case of a path model with a variable as
a dependent of a single exogenous variable (and an error residual term), the path coefficient in this special
case is a zero-order correlation coefficient.
Consider this model, based on Bryman, A. and D. Cramer (1990). Quantitative data analysis for social
scientists, pp. 246-251.

This model is specified by the following path equations:

Equation
1.
satisfaction
=
Equation
2.
income
Equation 3. autonomy = b31age + e3

b11age
=

+
b12autonomy
b21age
+

+
b13
income
+
b22autonomy
+

e1
e2

where the b's are the regression coefficients and their subscripts are the equation number and variable number
(thus b21 is the coefficient in Equation 2 for variable 1, which is age.
Note: In each equation, only (and all of) the direct priors of the endogenous variable being used as the
dependent are considered. The path coefficients, which are the betas in these equations, are thus the
standardized partial regression coefficients of each endogenous variable on its priors. That is, the beta for any
path (that is, the path coefficient) is a partial weight controlling for other priors for the given dependent
variable.
Formerly called p coefficients, now path coefficients are called simply beta weights, based on usage in
multiple regression models. Bryman and Cramer computed the path coefficients = standardized regression
coefficients = beta weights, to be:

Correlated Exogenous Variables. If exogenous variables are correlated, it is common to label the
corresponding double-headed arrow between them with its correlation coefficient.
Disturbance terms.The residual error terms, also called disturbance terms, reflect unexplained
variance (the effect of unmeasured variables) plus measurement error. Note that the dependent in
each equation is an endogenous variable (in this case, all variables except age, which is exogenous).
Note also that the independents in each equation are all the variables with arrows to the dependent.
The effect size of the disturbance term for a given endogenous variable, which reflects unmeasured
variables, is (1 - R2), and its variance is (1 - R2) times the variance of that endogenous variable,
where R2 is based on the regression in which it is the dependent and those variables with arrows to
it are independents. The path coefficient is SQRT(1 - R2).
The correlation between two disturbance terms is the partial correlation of the two endogenous
variables, using as controls all their common causes (all variables with arrows to both). The
covariance estimate is the partial covariance: the partial correlation times the product of the
standard deviations of the two endogenous variables.

Path multiplication rule: The value of any compound path is the product of its path coefficients. Imagine a
simple three-variable compound path where education causes income causes conservatism. Let the regression
coefficient of income on education be 1000: for each year of education, income goes up $1,000. Let the
regression coefficient of conservatism on income be .0002: for every dollar income goes up, conservativism
goes up .0002 points on a 5-point scale. Thus if education goes up 1 year, income goes up $1,000, which
means conservatism goes up .2 points. This is the same as multiplying the coefficients: 1000*.0002 = .2. The
same principle would apply if there were more links in the path. If standardized path coefficients (beta
weights) were used, the path multiplication rule would still apply, but the the interpretation is in standardized
terms. Either way, the product of the coefficients along the path reflects the weight of that path.
Effect decomposition. Path coefficients may be used to decompose correlations in the model into direct and
indirect effects, corresponding, of course, to direct and indirect paths reflected in the arrows in the model.
This is based on the rule that in a linear system, the total causal effect of variable i on variable j is the sum of
the values of all the paths from i to j. Considering "satisfaction" as the dependent in the model above, and
considering "age" as the independent, the indirect effects are calculated by multiplying the path coefficients
for each path from age to satisfaction:
age
->
income
age
->
autonomy
age
->
autonomy
->
total indirect effect = .45

->
->
income

satisfaction
satisfaction
->
satisfaction

is
is
is

.57*.47
.28*.58
.28*.22
x

=
=
.47

.26
.16
.03

That is, the total indirect effect of age on satisfaction is plus .45. In comparison, the direct effect is only
minus .08. The total causal effect of age on satisfaction is (-.08 + .45) = .37.

Effect decomposition is equivalent to effects analysis in regression with one dependent variable. Path
analysis, however, can also handle effect decomposition for the case of two or more dependent variables.
In general, any bivariate correlation may be decomposed into spurious and total causal effects, and the total
causal effect can be decomposed into a direct and an indirect effect. The total causal effect is the coefficient
in a regression with all of the model's prior but not intervening variables for x and y controlled (the beta
coefficient for the usual standardized solution, the partial b coefficient for the unstandardized or raw
solution). The spurious effect is the total effect minus the total causal effect. The direct effect is the partial
coefficient (beta for standardized, b for unstandardized) for y on x controlling for all prior variables and all
intervening variables in the model. The indirect effect is the total causal effect minus the direct effect, and
measures the effect of the intervening variables. Where effects analysis in regression may use a variety of
coefficients (partial correlation or regression, for instance), effect decomposition in path analysis is restricted
to use of regression.
For instance, imagine a five-variable model in which the exogenous variable Education is correlated with the
exogenous variable Skill Level, and both Education and Skill Level are correlated with the exogenous
variable Job Status. Further imagine that Education and each of the other two exogenous variables are
modeled to be direct causes of Income and also of Median House Value, which are the two dependent
variables. We might then decompose the correlation of Education and Income:
1.
2.

3.

Direct effect of Education on Income, indicated by the path coefficient of the single-headed arrow
from Education to Income.
Indirect effect due to Education's correlation with Skill Level, and Skill Level's direct effect on
Income, indicated by multiplying the correlation of Education and Skill Level by the path
coefficient from Skill Level to Income.
Indirect effect due to Education's correlation with Job Status, and Job Status's direct effect on
Income, indicated by multiplying the correlation of Education and Job Status by the path
coefficient from Job Status to Income.

As a second example decomposition for the same five-variable model is a bit more complex if we wish to
break down the correlation of the two dependent variables, Income and Median House Value. Since here
somewhat implausibly the two dependents are modeled not to have a direct effect from Income to House
Value, the true correlation is hypothesized to be zero and all correlations are spurious.
4.

5.

6.

7.

8.

9.

The spurious direct effect of Education as a common anteceding variable directly causing both
dependents, indicated by multiplying the path coefficient from Education to Income by the path
coefficient of Education to House Value.
The spurious direct effect of Skill Level as a common anteceding variable directly causing both
dependents, indicated by multiplying the path coefficient from Skill Level to Income by the path
coefficient of Skill Level to House Value.
The spurious direct effect of Job Status as a common anteceding variable directly causing both
dependents, indicated by multiplying the path coefficient from Job Status to Income by the path
coefficient of Job Status to House Value.
The spurious indirect effect of Education and Skill Level as a common antecedings variable
directly causing both dependents, indicated by multiplying the path coefficient from Education to
Income by the correlation of Education and Skill Level by the path from Skill Level to House
Value and adding the product of the path from Skill Level to Income by the correlation of
Education and Skill Level by the path from Education to Median House Value.
The spurious indirect effect of Education and Job Status as a common anteceding variables directly
causing both dependents, indicated by multiplying the path coefficient from Education to Income
by the correlation of Education and Job Status by the path from Job Status to House Value and
adding the product of the path from Job Status to Income by the correlation of Education and Job
Status by the path from Education to Median House Value..
The spurious indirect effect of Skill Level and Job Status as a common anteceding variables
directly causing both dependents, indicated by multiplying the path coefficient from Skill Level to
Income by the correlation of Skill Level and Job Status by the path from Job Status to House Value
and adding the product of the path from Job Status to Income by the correlation of Skill Level and
Job Status by the path from Skill Level to Median House Value..

10. The residual effect is the difference between the correlation of Income and Median House Value
and the sum of the spurious direct and indirect effects.
Correlated exogenous variables. The path weights connecting correlated exogenous variables are equal to the
Pearson correlations. When calculating indirect paths, not only direct arrows but also the double-headed
arrows connecting correlated exogenous variables, are used in tracing possible indirect paths, except:
Tracing rule: An indirect path cannot enter and exit on an arrowhead. This means that you cannot have a
direct path composed of the paths of two correlated exogenous variables.

Significance and Goodness of Fit in SEM Path Models

OLS vs. SEM While a series of OLS regressions may be used to implement path analysis, testing
individual path coefficients using the standard t or F test from regression output, today it is far more
common to use structural equation modeling (SEM) software. This section uses AMOS with a
model based on Ingram et al. (2000), used with the kind permission of Karl Wuensch. Use of
AMOS is described more fully in the section on structural equation modeling. A structural equation
model with simple rather than latent variables is a path model.
The SEM path model. The path model is drawn as usual in SEM. Illustrated below is the model for
the Ingram data, which deals with application to graduate schools. In this model, Attitude,
SubNorm, and PBC all predict Intent, while the ultimate dependent variable, Behavior, is predicted
by Intent and also directly by PBC. As customary, the straight arrows represent regression paths for
presumed causal relationships, which the curved double-headed arrows represent assumed
covariances among the exogenous variables. The endogenous variables are depicted with associated
error terms.

In this model, Attitude is the individual's attitude toward graduate school; SubNorm is subjective
norms, reflected by attitudes toward graduate school of others around the individual; PBC is
planned behavioral control, which is the individual's level of control over behaviors related to
graduate school. Intent is the individual's intent to go to graduate school. Behavior is applying to
graduate school.

Select outputs. Statistical tests and other outputs are selected under View, Analysis Properties, in
the AMOS menu system, yielding the dialog shown below:

Path coefficients in standardized and unstandardized form are generated by AMOS by selecting
Analyze, Compute Estimates. Un like the OLS regression method, all parameters are calculated
simultaneously. These coefficients may be caused to be displayed on the path diagram, and also
appear in the output, which is obtained in the menu system by selecting View, Text Output. For this
example, note that the paths from SubNorm and PBC to Intent are not significant. That is, Intent is
primarily a function of Attitude.

Overall test of the model. The likehood ratio chi-square test, also called the model chi-square test or
deviance test, assesses the overall fit of the model. A finding of nonsignificance corresponds to an
adequate model - one whose model-implied covariance matrix does not differ from the observed
covariance matrix. For this example, there is adequate fit:

Result (Default model)

Minimum was achieved
Chi-square = .847
Degrees of freedom = 2
Probability level = .655
However, the likelihood ratio chi-square test cannot be relied upon alone, particularly for large
samples, because a finding of significance (rejecting the model) can occur even with very small
differences of the model-implied and observed covariance matrices (note below that AMOS labels
the likelihood ratio chi-square CMIN). Therefore a large variety of other goodness of fit measures
have been devised. Their use and relative merits are described in the section on structural equation

modeling. In the output below, the "default model" is the researcher's model. The "saturated model"
is the perfectly explanatory but trivial model with all possible arrows. he "independence model" is
the model with no regression arrows (straight arrows). Suffice it to say, these goodness of fit
measures support the adequacy of the model in the example (for example, RMSEA should be <.05
for a good model and here is .000). Note, however, that in spite of statistically adequate goodness
of fit, normally the researcher would drop non-significant structural arrows indicated above in the
path coefficients section.

Correlations. Also included in AMOS output are the correlations among the exogenous variables
(correlations in the upper output) and between the exogenous variables and the endogenous
variables (squared multiple correlation in lower output) as illustrated below. The model explains
34.3% of the variance in the dependent variable, Behavior.

Correlations: (Group number 1 - Default model)

Estimate
SubNorm <--> PBC .505
Attitude <--> SubNorm .472
Attitude <--> PBC .665
Squared Multiple
Default model)
Estimate
Intent
.600
Behavior
.343

Correlations:

(Group

number

10

Direct and indirect effects. AMOS will use the muliplication rule automatically to partition overall effects
into direct and indirect effects for the endogenous variables (for Intent and Behavior in this example).

Modification indexes. Modificiation indexes (MI) may be used to add arrows to the model. The
larger the MI, the more adding the model will improve model fit. As discussed in the section on
structural equation modeling, MIs should be interpreted in relation to critical ratios (CR), which are
a measure of the change in likelihood ratio chi square. And as noted above, nonsignificance of path

11

coefficients is used to drop arrows in a model-building and model-trimming process discussed in

the section on structural equation modeling. The conservative approach calls for adding or dropping
one arrow at a time as each change will affect the coefficients. For this example, the MIs are so
small that no addition of arrows is warranted. In fact, all MIs are well below the usual lower
threshold of 4.0.

Assumptions
o
o
o

o
o

Linearity: relationships among variables are linear (though, of course, variables may be nonlinear
transforms).
Additivity: there are no interaction effects (though, of course, variables may be interaction crossproduct
terms)
Interval level data for all variables, if regression is being used to estimate path parameters. As in other forms
of regression modeling, it is common to use dichotomies and ordinal data in practice. If dummy variables are
used to code a categorical variable, one must be careful that they are represented as a block in the path
diagram (ex., if an arrow is drawn to one dummy it must be drawn to all others in the set). If an arrow were to
be drawn from one dummy variable to another dummy variable in the same set, this would violate the
recursivity assumption discussed below.
Residual (unmeasured) variables are uncorrelated with any of the variables in the model other than the
one they cause.
Disturbance terms are uncorrelated with endogenous variables. As a corollary of the previous
assumption, path analysis assumes that for any endogenous variable, its distubance term is uncorrelated with
any other endogenous variable in the model. This is a critical assumption, violation of which may make
regression inappropriate as a method of estimating path parameters. This assumption may be violated due to
measurement error in measuring an endogenous variable; when an endogenous variable is actually a direct or
indirect cause of a variable which the model states is the cause of that endogenous variable (reverse

12

o
o

o
o
o

causation); or when a variable not in the model is a cause of an endogenous variable and a variable the model
specifies as a cause of that endogenous variable (spurious causation).
Low multicollinearity (otherwise one will have large standard errors of the b coefficients used in removing
the common variance in partial correlation analysis).
No underidentification or underdetermination of the model is required. For underidentified models there
are too few structural equations to solve for the unknowns. Overidentification usually provides better
estimates of the underlying true values than does just identification.
Recursivity: all arrows flow one way, with no feedback looping. Also, it is assumed that
disturbance (residual error) terms for the endogenous variables are uncorrelated. Recursive models
are never underidentified.
Proper specification of the model is required for interpretation of path coefficients. Specification error
occurs when a significant causal variable is left out of the model. The path coefficients will reflect the shared
covariance with such unmeasured variables and will not be accurately interpretable in terms of direct and
indirect effects. In particular, if a variable specified as prior to a given variable is really consequent to it, "we
can do ourselves considerable damage" (Davis, 1985: 64) because if a variable is consequent it would be
estimated to have no path effect, whereas when it is included as a prior variable in the model, this erroneously
changes the coefficients for other variables in the model. Note, however, that while interpretation of path
coefficients is inaccurate under specification error, it is still possible to compare the relative fit of two
models, perhaps both with specification error.
Appropriate correlation input. When using a correlation matrix as input, it is appropriate to use Pearsonian
correlation for two interval variables, polychoric correlation for two ordinals, tetrachoric for two dichotomies,
polyserial for an interval and an ordinal, and biserial for an interval and a dichotomy.
Adequate sample size is needed to assess significance. Kline (1998) recommends 10 times as many cases as
parameters (or ideally 20 times). He states that 5 times or less is insufficient for significance testing of model
effects.
The same sample is required for all regressions used to calculate the path model. This may require reducing
the data set down so that there are no missing values for any of the variables included in the model. This
might be achieved by listwise dropping of cases or by data imputation.

Frequently Asked Questions

o

Does path analysis confirm causation in a model?

No, although this is sometimes said. Everitt and Dunn (1991) note, "However convincing,
respectable and reasonable a path diagram... may appear, any causal inferences extracted are rarely
more than a form of statistical fantasy". The authors are referring to the fact that ultimately path
analysis deals with correlation, not causation of variables. The arrows in path models do indeed
reflect hypotheses about causation. However, many models may be consistent with a given dataset.
Path analysis merely illuminates which of two or more competing models, derived from theory, is
most consistent with the pattern of correlations found in the data. The competing theories may be
represented in separate path models with separate path analyses, or may be combined in a single
path diagram, in which case the researcher is concerned with comparing the relative importance of
different paths within the diagram.

Can path analysis be used for exploratory rather than confirmatory purposes?
Methodologists favor a priori formulation of hypotheses about the results to be obtained from path
analysis, not post factum conclusions based on the results. That is, the researcher should be seeking
to confirm hypotheses made beforehand. At a minimum, the researcher should posit the sign of
each relationship (arrow) in the model, and ideally should go further to posit the arrangement by
magnitude of the importance of the independents, or even better yet, the intervals within which the
path coefficients will be expected to lie. Since data can support multiple models, path analysis
should focus on determining which of two or more theoretically-derived models most conform to
the underlying data.

13

If used in an exploratory way, note that the more models you test, the more likely you will confirm
one just by chance. Thus the actual confidence level is equal to the individual test confidence level
to the power of the number of models tested, here .95 cubed = .86. Thus testing 3 models at .95
would mean we were actually operating at about .86 confidence. At the .99 level, however, actual
confidence would still be .95 with 5 models.

How do I know if my model is "underidentified" and what difference does it make? How does this
relate to "recursivity?
A unique path solution cannot be calculated if a model is underidentified. Identification is defined
and steps the researcher can take to deal with underidentification are dealt with in a section on
structural equation modeling. How to determine beforehand if a model is underidentified, other
than by running a path analysis program for sample or fictional data and seeing if there are error
flags, is discussed more fully under a second section on structural equation modeling.

How does the significance of a path coefficient compare with the significance of the corresponding
regression coefficient?
They are identical. The path coefficient is the beta for the regression of the dependent or other
endogenous variable on the other variables with arrows to it. The significance of the beta and b
coefficients will be the same, and is displayed on the same line in SPSS regression output.
Naturally, all paths in the model should be significant!

How do you assess the significance of the total (direct and indirect) effect of exogenous variable x on
endogenous variable y?
Run a regression with y as dependent and all others as independents, leaving out any variable
which mediates between x and y. The significance of the b or beta for x in this equation is a test of
the significance of the total effect.

Why might the direct effect be zero?

There is a fully controlling mediating effect or fully controlling anteceding effect. See further
discussion in the section on partial correlation.

How are path coefficients related to the correlation matrix for purposes of testing a model?
First, recognize that computation of the model-estimated correlations and their comparison with
observed correlations is best done by relying on a model-estimating program such as LISREL or
AMOS. The model path coefficients can be compared to the predicted path coefficients as
computed from the correlation matrix, following which the model coefficients can be tested for
goodness-of-fit with the predicted coefficients.
The tracing rule is a rule for identifying all the paths, the sum of effects of which is the estimated
correlation between two variables in the model. This model-estimated correlation can be compared
to the observed correlation to assess the fit of the model to the data. The tracing rule is simply that
the model-implied correlation between two variables in a model is the sum of all valid paths
(tracings) between the two variables. These include the total effect (which is the sum of direct and
indirect effects) plus any associational effects due to correlated exogenous variables. These
associational effects are calculated by multiplying the correlation between the exogenous variable
under consideration with a second exogenous variable, by this second exogenous variable"s total
effect on the target variable under consideration.
For simplicity, consider this simple model:

14

The actual (observed) correlation matrix might look like this:

A
A
B
C

B
.379
1

D
-.652
-.451
1

When running AMOS or another path analysis program, the path coefficients (standardized
regression coefficients) would be:

A
A
B
C

B
.379
1

D
-.562
-.238
1

If one had only the path output and wanted to estimate back to the correlation matrix, one would
use these equations, one for each path:

r(AB) = n = .379 (the correlation is the standardized path

coefficient, in the bivariate case
where the independent is exogenous
and the dependent has no other inputs)
r(BD) = q + np = -.238 + .379(-.562) = -.451
r(AD) = p + nq = -.562 + .379(-.238) = -.652
In testing a model, somewhat similar reasoning is followed to compare model-implied covariance
matrices with observed covariance matrices, with smaller differences indicating better goodness of
fit.

How, exactly, can I compute path coefficients in SPSS?

The recommended method is to enter a path model into AMOS, which is the SPSS program for
structural equation modeling, discussed in that section. However, in the SPSS regression module,
for a recursive model, let VARA cause VARB and VARC, and let VARB cause VARC. A series of
regressions is conducted with each non-exogenous variable considered as a dependent in turn. For
the foregoing model, the path coefficient from VARA to VARB is given by this SPSS code:

REGRESSION
/MISSING LISTWISE
/STATISTCS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT VARB
/METHOD=ENTER VARA.

15

The path coefficients from VARA to VARC and from VARB to VARC are given by this second
regression command:

REGRESSION
/MISSING LISTWISE
/STATISTCS COEFF OUTS R ANOVA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT VARC
/METHOD=ENTER VARA VARB.

How do I compute the value of the path from an error term to an endogenous variable?
The path is the square root of (1 - R-squared), where R-squared is from the regression equation for
the corresponding dependent variable. Do not use adjusted R-squared.

How can multiple group path analysis determine if the path model differs across groups in my sample?
Multiple group path analysis may be accomplished simply by running separate path analysis for
each group in the sample, then comparing the path estimates. A more sophisticated approach
supported by some path analysis and SEM packages involves a second step: to impose a crossgroup equality constraint on the path estimates, then run the analysis separately for each group,
then see if the goodness-of-fit for the constrained models is as good as for the unconstrained
models. If the fit of the constrained model is worse than that for the corresponding unconstrained
model, then the researcher concludes that model direct effects differ by group.

Could I substitute logistic regression in doing effect decomposition?

No. Forward paths cannot be decomposed accurately with log linear techniques. See Davis (1985):
48, 59.

Can path analysis handle hierarchical/multilevel data?

Multilevel path analysis is available in Mplus but not yet in SPSS or SAS as of 2011.

Are regression and SEM the only approaches to path analysis?

Partial least squares path analysis is also available, through custom software (not SPSS or SAS,
which only support PLS regression). PLS can support small sample models, even where there are
more variables than observations, but it is lower in power than SEM approaches.

16