Using the SmartPLS Software

“Structural Model Assessment”

Joe F. Hair, Jr.

Founder & Senior Scholar

All rights reserved ©. Cannot be reproduced or distributed without express written permission 1
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Structural Model Assessment
Once the construct measures have been confirmed as reliable and
valid, the next step is to assess the structural model results. This
involves examining the model’s predictive capabilities and the
relationships between the constructs. Exhibit 6.1 (next slide) shows a
systematic approach to the assessment of structural model results.
Before assessing the structural model, you must examine the
structural model for collinearity (Step 1). The reason is that the
estimation of path coefficients in the structural model is based on OLS
regressions of each endogenous latent variable on its corresponding
predecessor constructs. Just as in a regular multiple regression, the
path coefficients may be biased if the estimation involves significant
levels of collinearity among the predictor constructs.
The key criteria for assessing the structural model in PLS-SEM
are: Step 2 – the significance of the path coefficients, Step 3 – the level
of the R² values, Step 4 – the f² effect size, and Step 5 – the predictive
relevance (Q² & the q² effect size).

All rights reserved ©. Cannot be reproduced or distributed without express written permission 2
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Step 1: Collinearity Assessment
To assess collinearity, we apply the same measures as in the
evaluation of formative measurement model indicators (i.e., tolerance
and VIF values). To do so, we need to examine each set of predictor
constructs separately for each subpart of the structural model. For
instance, in the model shown in Exhibit 6.2 (next slide), Y1 and Y2
jointly explain Y3. Likewise, Y2 and Y3 act as predictors of Y4.
Therefore, you need to check whether there are significant levels of
collinearity between each set of predictor variables (constructs). In other
words, you need to check the collinearity between Y1 and Y2 as well as
between Y2 and Y3.
Similar to the assessment of formative measurement model
indicators, we consider tolerance levels below 0.20 (VIF above 5.00) in
the predictor constructs as indicative of collinearity that is too high. If
collinearity is exceeds these thresholds, you should consider eliminating
constructs, merging predictors into a single construct, or creating
higher-order constructs to deal with collinearity problems.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 4
from Sage, Prentice-Hall, SmartPLS, and session presenters.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 5
from Sage, Prentice-Hall, SmartPLS, and session presenters.
Evaluating Construct Collinearity – Extended Reputation Model
The following sets of
(predictor) constructs are
run to assess collinearity:
(1) ATTR, CSOR, PERF,
and QUAL as predictors of
COMP; (2) COMP and LIKE
as predictors of CUSA; and
(3) COMP, LIKE, and CUSA
as predictors of CUSL.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 6
from Sage, Prentice-Hall, SmartPLS, and session presenters.
Evaluating Construct Collinearity – Extended Reputation Model

To assess construct collinearity, run the extended reputation model and open
the default report by going to Menu → Report → Default Report. Next, you
need to extract the latent variable scores from the default report, which you can
find under PLS → Calculation Results → Latent Variable Scores. The scores
are shown on this slide. Copy these scores to an SPSS file to run this analysis.
Using the SPSS linear regression option, the following sets of (predictor)
constructs are run to assess collinearity: (1) ATTR, CSOR, PERF, and QUAL as
predictors of COMP; (2) COMP and LIKE as predictors of CUSA; and (3) COMP,
LIKE, and CUSA as predictors of CUSL.
The SPSS steps for testing collinearity are shown on the next slide for the first
All rights reserved ©. Cannot be reproduced or distributed without express written permission 7
regression
from Sage,run.Prentice-Hall, SmartPLS, and session presenters.
 Using SPSS to Assess Collinearity

These are the

four exogenous
constructs that are
being tested for
multicollinearity.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 8
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Evaluating Collinearity Among Exogenous Constructs
Below are the SPSS collinearity results from using
ATTR, CSOR, PERF, and QUAL as predictors of COMP.
All VIF values
are clearly below
the threshold of 5.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 9
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Evaluating Collinearity Among Exogenous Constructs
Below are the collinearity results from using COMP and LIKE as
predictors of CUSA (top table), and COMP, LIKE and CUSA as predictors
of CUSL (bottom table). All VIF values are well below the threshold of 5.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 10
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Step 2: Assess Significance and Relevance
of the Structural Model Relationships
After applying the PLS-SEM algorithm, estimates are
obtained for the structural model relationships (the path
coefficients), which represent the hypothesized relationships
between the constructs.
The path coefficients for the structural model are shown
in the next several slides. These results were obtained from
the SmartPLS Default Report with the following sequence:
PLS Algorithm → Calculation Results → Path Coefficients

Before examining the sizes of the path coefficients we

will first examine their significance. To do so, we must first
run the Bootstrapping option.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 11
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Reputation Model Results
– Path Coefficients –

All rights reserved ©. Cannot be reproduced or distributed without express written permission 12
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Bootstrapping
Which path coefficients
are significant?

Click here to run

Bootstrapping

When you run

bootstrapping select
mean replacement for
missing data, no sign
changes, your actual
sample size = cases, and
5,000 samples.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 13
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Bootstrapping Default Report – Significance of Path Coefficients

T Statistics are found

beside the tab below.

The above results show the significance of the path coefficients. Recall that
a T Statistic > 1.96 is significant with a two-tailed test, and >.98 is significant for a
one-tailed test. Note: path coefficients are shown in Original Sample column.
The results indicate that all paths are statistically significant using a one-tailed
test except COMP – CUSL. But eight of the thirteen structural paths are
significant based on a two-tailed test.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 14
from Sage, Prentice-Hall, SmartPLS, and session presenters.
Obtaining the Default Report to
Evaluate Path Coefficients
Click here to obtain
Default Report

All rights reserved ©. Cannot be reproduced or distributed without express written permission 15
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 SmartPLS Default Report – Path Coefficients

After examining the significance of relationships, it is important to assess the

relevance of significant relationships. Path coefficients in the structural model
may be significant, but their size may be so small that they do not warrant
managerial attention.
Structural model path coefficients can be interpreted relative to one another.
If one path coefficient is larger than another, its effect on the endogenous latent
variable is greater. More specifically, the individual path coefficients of the path
model can be interpreted just as the standardized beta coefficients in an OLS
regression. These coefficients represent the estimated change in the
endogenous construct for a unit change in a predictor construct.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 16
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 PLS Algorithm Default Report – Path Coefficients

Looking at the relative importance of the exogenous driver constructs in

predicting the dependent construct perceived competence (COMP), we see that
customer perceptions of the company’s quality of products and services (QUAL =
0.4297) is most important, followed by performance (PERF = 0.2955). In
contrast, the perceived attractiveness (ATTR = 0.0861) and degree to which the
company acts in socially conscious ways (CSOR = 0.0589) have very little
influence on COMP. These last two drivers are, however, more important in
predicting a company’s likeability (LIKE) – ATTR = 0.1671 and CSOR = 0.1784,
but the most important predictor of LIKE is QUAL (0.3800).
When you examine the endogenous likability construct (LIKE) you see that it
is the primary driver (predictor) of customer satisfaction (CUSA = 0.4357), and a
meaningful predictor
All rights reserved of loyalty
©. Cannot (CUSL
be reproduced = 0.3440),
or distributed and COMP
without express has little impact on17
written permission
CUSL from(0.0057).
Sage, Prentice-Hall, SmartPLS, and session presenters.
 Understanding Direct and Indirect Effects
Researchers are often interested in evaluating not only one
construct’s direct effect on another but also its indirect effects via
one or more mediating constructs. The sum of direct and indirect
effects is referred to as the total effect.
In Exhibit 6.4 on the next slide, constructs Y1 and Y3 are
linked by a direct effect (p13 = 0.20). In addition, there is an
indirect effect between the two constructs via the mediating
construct Y2. This indirect effect can be calculated as the product
of the two effects p12 and p23 (p12 x p23 = 0.80 x 0.50 = 0.40). The
total effect is 0.60, which is calculated as p13 + (p12 x p23) = 0.20 +
(0.80 x 0.50) = 0.20 + 0.40 = 0.60.
Although the direct effect of Y1 to Y3 is not very strong (i.e.,
0.20), the total effect (both direct and indirect combined) is quite
pronounced (i.e., 0.60), indicating the relevance of Y1 in explaining
Y3. This type of result suggests that the direct relationship from Y1
to Y3 is mediated by Y2.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 18
from Sage, Prentice-Hall, SmartPLS, and session presenters.
.
 PLS Algorithm Default Report – Total Effects = Sizes

The four driver constructs for loyalty (CUSL) are the exogenous
constructs on the left side of the SEM model (these constructs are
actionable because they are formative and thus of primary concern
for the total effects analysis). The findings shown above indicate that
quality (QUAL = 0.2483) has the strongest total effect on loyalty,
followed by corporate social responsibility (CSOR = 0.1053),
attractiveness (ATTR = 0.1010), and performance (PERF = 0.0894).
Note that there are three mediating constructs (COMP, CUSA &
LIKE) between the exogenous driver constructs and the endogenous
construct CUSL whose role must be considered, but they are
relatively less actionable because they are reflective – not formative.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 20
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Bootstrapping Default Report – Total Effects = Significance

You can also examine the significance of the total effects. Above are
the T statistics for the total effects from the bootstrapping default report.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 21
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 .

Note: In the above table we only show the total effects for the four exogenous
driver constructs to the satisfaction and loyalty endogenous constructs,
All rights reserved ©. Cannot be reproduced or distributed without express written permission 22
because these
from Sage, Prentice-Hall, areand
SmartPLS, thesession
outcomes the company is most interested in managing.
presenters.
 To get outer weights:
PLS → Calculation Results → Outer Weights

Identifying actionable
strategies based on
sizes of exogenous
construct weights.

By examining the outer weights of the construct indicators, we can identify which
specific element of quality (QUAL) needs to be addressed. Note that qual_6 has the
highest outer weight (0.3980). This survey question was “[the company] seems to be
a reliable partner for customers” so perceptions of reliability should be enhanced and
23
communicated to customers.
 Comment on mediating effect in
reputation model ??

All rights reserved ©. Cannot be reproduced or distributed without express written permission 24
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Step 3: Assess the Level of R2
The R² values of the endogenous latent variables are available in the
PLS Algorithm default report (PLS → Quality Criteria → Overview), as
shown below. The R2 values of COMP (0.6309), CUSL (0.5620), and
LIKE (0.5576) can be considered moderate. In contrast, the R² value of
CUSA (0.2919) is rather weak.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 25
from Sage, Prentice-Hall, SmartPLS, and session presenters. 25
Note: calculation of Q2 value is shown in a later slide.
26
 Step 4: Assessing Effect Size – ƒ 2
The ƒ² effect size is a measure of the impact of a specific
predictor construct on an endogenous construct. In addition to
evaluating the size of the R² values of all endogenous constructs, the
ƒ² effect size can be calculated (it is not available from the SmartPLS
software output). The ƒ² effect size measures the change in the R²
value when a specified exogenous construct is omitted from the model.
It is used to evaluate whether the omitted predictor construct has a
substantive impact on the R² values of the endogenous construct(s).
The formula for calculating the ƒ² effect size is shown on the next
slide. Guidelines for assessing ƒ2 values for the exogenous latent
constructs in predicting the endogenous constructs are:
Value Effect Size
0.02 = small
0.15 = medium
0.35 = large
(Cohen, 1988)

All rights reserved ©. Cannot be reproduced or distributed without express written permission 27
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Calculating the Effect Size – ƒ 2
The f 2 effect size can be calculated as shown below. In the
formula R2included and R2excluded are the R² values of the
endogenous latent variable when a selected exogenous latent variable
is included or excluded from the model. The change in the R² values is
calculated by estimating the PLS path model twice. It is estimated the
first time with the exogenous latent variable included (yielding
R2included) and the second time with the exogenous latent variable
excluded (yielding R2excluded).
The results of calculating the f 2 for the reputation model example
endogenous variables are shown in a later slide.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 28
from Sage, Prentice-Hall, SmartPLS, and session presenters.
Example: Calculation of f2 Effect Size
As indicated in the previous slide, to compute the f2 value of
a selected endogenous latent construct, we need the R2
included and R2excluded values. The R2included results from the
overall model estimation were previously shown (Exhibit 6.15).
The R2excluded value is obtained from a model re-estimation
after deleting a specific predecessor of that endogenous latent
variable. For example, the endogenous latent variable CUSL
has an original R2 value of 0.562 (R2included). If CUSA is
deleted from the path model and the model is re-estimated, the
R2 of CUSL has a value of only 0.385 (R2excluded). These two
values are the inputs for computing the f2 effect size of CUSA on
CUSL. The formula is shown below:

29
 Example: Calculation of f2 for other endogenous constructs
To get the f2 values you need to run the full model first and
determine the R2 for the endogenous construct you want to
evaluate. Next, eliminate one path pointing at the construct
you are looking at (simply right click on a construct and delete
the predictor construct), and re-run the model. The R2 with the
construct/path will be lower since the predictor construct was
removed. Now enter the R2 included for the selected
construct (based on the full model) and the R2 excluded
(based on the reduced model where one path/construct has
been deleted) into the formula on the previous slide.
The same procedure is followed for the q2 but instead of
entering the R2 (excluded and included), you use blindfolding
to get the Q2 values for the full model (included) and the
reduced model (construct/path deleted).
All rights reserved ©. Cannot be reproduced or distributed without express written permission from 30
Prentice-Hall, McGraw-Hill, SmartPLS, and session presenters.
 Summary of Results – Path Coefficients, f2 and q2
COMP CUSL
Path f2 effect q2 effect Path f2 effect q2 effect
Coefficient size size Coefficient size size
ATTR 0.0861 0.011 0.028
COMP 0.0057 0.000 0.101
CSOR 0.0589 0.005 0.004
CUSA 0.5050 0.404 0.229
LIKE 0.3440 0.139 0.081
PERF 0.2955 0.076 0.042
QUAL 0.4297 0.144 0.062

Example interpretation of f2: Look under the f2 column for CUSL. Note
the 0.404 is the f2 effect size for the predictive value of CUSA on CUSL. The
0.404 indicates that CUSA has a large effect in producing the R2 for CUSL. In
contrast, the 0.139 is the f2 effect size for the predictive value of LIKE on
CUSL. The 0.139 indicates that LIKE has close to a medium effect in
producing the R2 for CUSL. 31
 Reputation Model Results
– Path Coefficients –

Guidelines for assessing

ƒ2 values:
Value Effect Size
0.02 = small
0.15
All rights reserved ©. Cannot be reproduced or distributed without express written permission = medium
32
from Sage, Prentice-Hall, SmartPLS, and session presenters. 0.35 = large
 Summary of Results – Path Coefficients, f2 and q2
CUSA LIKE

Path f2 effect q2 effect Path f2 effect q2 effect

Coefficient size size Coefficient size size

ATTR 0.1671 0.029 0.019

COMP 0.1455 0.018 0.001

CSOR 0.1784 0.036 0.029

CUSA

LIKE 0.4357 0.161 0.140

PERF 0.1170 0.011 0.007

QUAL 0.3800 0.095 0.053

33
Step 5: Blindfolding and Predictive Relevance – Q 2

In addition to evaluating the magnitude of the R² values as a criterion of

predictive accuracy, researchers should also examine the Q² value – which
is an indicator of the model’s predictive relevance. The Q² measure
applies a sample re-use technique that omits part of the data matrix and
uses the model estimates to predict the omitted part. Specifically, when a
PLS-SEM model exhibits predictive relevance, it accurately predicts the data
points of the indicators in reflective measurement models of multi-item as
well as single-item endogenous constructs (the procedure does not apply to
formative endogenous constructs).
For SEM models, Q² values larger than zero for a specific reflective
endogenous latent variable indicate the path model’s predictive relevance
for a particular construct. Q² values of zero or below indicate a lack of
predictive relevance. As a relative measure of predictive relevance, values
of 0.02, 0.15, and 0.35 indicate that an exogenous construct has a small,
medium, or large predictive relevance for a selected endogenous construct.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 34
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Blindfolding and Predictive Relevance – Q 2
The Q ² value can be calculated by using two different approaches.
The cross-validated redundancy approach uses the path model
estimates of both the structural model (scores of the antecedent
constructs) and the measurement model (target endogenous constructs).
An alternative method is the cross-validated communality
approach. This method uses only the construct scores estimated for the
target endogenous construct (without including the structural model
information) to predict the omitted data points.
We recommend using the cross-validated redundancy as a measure
of Q2 since it includes the key element of the path model, the structural
model, to predict eliminated data points.
When you run the blindfolding option for cross-validated redundancy,
all constructs in your SEM model are shown (see next slide). You can
select multiple latent variables in the dialog box, but you need to run this
routine for one reflective target construct at a time – one after the other
until all have been tested. Finally, this option is only used with
endogenous constructs that are measured reflectively.
All rights reserved ©. Cannot be reproduced or distributed without express written permission 35
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 SmartPLS Predictive Relevance – Blindfolding
Redundancy vs. Communality?

Cross-validated redundancy
LV1
Step 1: MV 1
The scores of the endogenous LV(s) LV3 MV 2
are estimated using the scores of the
exogenous LVs MV 3
LV2

LV1
MV 1
Step 2:
Newly estimated LV scores are used LV3 MV 2
to estimate the missing MV data MV 3
LV2

Cross-validated communality Only step 2.

 .

Running Blindfolding to obtain Q2

for Endogenous Construct CUSL
Note that only CUSL is checked.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 37
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Blindfolding Results for Endogenous Construct CUSL

Q2 value for CUSL

Full Base Model

The result column is at the top right

corner (1 – SSE/SSO). For our path
model the predictive relevance Q2 of
CUSL has a value of 0.4178, which
indicates the model has large predictive
relevance for this construct.
When blindfolding is run for all
endogenous latent constructs in the model
they all have Q2 values considerably
All rights reserved ©. Cannot be reproduced or distributed without express written permission 38
above
from zero, as shownSmartPLS,
Sage, Prentice-Hall, on theand next slide.
session presenters.
 Reputation Model – R2 and Q2 Measures

The table above shows that all Q2 values are considerably above zero, thus
providing support for the reputation model’s predictive relevance for the four
endogenous constructs.
 Computation of q 2
The final assessment addresses the calculation of the q² effect
sizes. The calculation of q2 for the CUSL construct of the reputation
model is shown below. To compute the q² value of a selected
endogenous latent variable, you need the Q2included and Q2excluded
values. The Q2included results for all endogenous constructs from the
overall model estimation are available from a previous slide. The
Q2excluded value is obtained from a model re-estimation after deleting a
specific predecessor of that endogenous latent variable. For example,
the endogenous latent variable CUSL has an original Q² value of 0.418
(Q2included). If CUSA is deleted from the path model and the model is
re-estimated, the Q² of CUSL has a value of only 0.285 (Q2excluded ).
These two values are the inputs for computing the q² effect size of
CUSA. on CUSL, as shown below:

All rights reserved ©. Cannot be reproduced or distributed without express written permission 40
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 PLS Algorithm Results
Base Model with CUSA Removed
.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 41
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Blindfolding Results for
CUSL with full Base Model
Q2 value for CUSL
with full model
.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 42
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Blindfolding Results for CUSL
Reduced Model with CUSA Removed Q2 value for CUSL
with CUSA Removed
.

All rights reserved ©. Cannot be reproduced or distributed without express written permission 43
from Sage, Prentice-Hall, SmartPLS, and session presenters.
 Summary of Results – Path Coefficients, f2 and q2
COMP CUSL
Path f2 effect q2 effect Path f2 effect q2 effect
Coefficient size size Coefficient size size
ATTR 0.0861 0.011 0.028
COMP 0.0057 0.000 0.101
CSOR 0.0589 0.005 0.004
CUSA 0.5050 0.404 0.229
LIKE 0.3440 0.139 0.081
PERF 0.2955 0.076 0.042
QUAL 0.4297 0.144 0.062

Example interpretation of q2: Look under the q2 column for CUSL. Note
the 0.229 is the q2 effect size for the predictive relevance of CUSA on CUSL.
The 0.229 indicates that CUSA has a medium effect in producing the Q2
(predictive relevance) for CUSL. In contrast, the 0.081 is the q2 effect size for
the predictive relevance of LIKE on CUSL. The 0.081 indicates that LIKE has a
small effect in producing the Q2 for CUSL. 44
 Reputation Model Results
– Path Coefficients –

Guidelines for assessing

q2 values:
Value Effect Size
0.02 = small
0.15
All rights reserved ©. Cannot be reproduced or distributed without express written permission = medium
45
from Sage, Prentice-Hall, SmartPLS, and session presenters. 0.35 = large
 Summary of Results – Path Coefficients, f2 and q2
CUSA LIKE

Path f2 effect q2 effect Path f2 effect q2 effect

Coefficient size size Coefficient size size

ATTR 0.1671 0.029 0.019

COMP 0.1455 0.018 0.001

CSOR 0.1784 0.036 0.029

CUSA

LIKE 0.4357 0.161 0.140

PERF Guidelines for assessing

0.1170 0.011 0.007
q2 values associated with
QUAL predictive relevance (Q2): 0.3800 0.095 0.053
Value Effect Size
0.02 = small
0.15 = medium 46
0.35 = large
.
 .

All rights reserved ©. Cannot be reproduced or distributed without express written permission 48
from Sage, Prentice-Hall, SmartPLS, and session presenters.