1

THE GENERAL STRUCTURAL EQUATION MODEL

WITH LATENT VARIATES

I. Specification:

A full structural equation model with latent variables consists of two parts:

• a latent variable model (which specifies the relations among the latent variables
of substantive interest)

η = Βη + Γξ + ς

• a measurement model (which links the latent variables to observed variables);

y = Λ yη + ε
x = Λxξ + δ

where: η → m x 1 vector of endogenous latent variates in deviation form,

ξ → n x 1 vector of exogenous latent variates in deviation form,
Β → m x m matrix of coefficients showing the influence of endogenous
latent variates on each other,
Γ → m x n matrix of coefficients showing the influence of exogenous latent
variates on endogenous latent variates,
ζ → m x 1 vector of errors in equations,
y → p x 1 vector of observed measures (indicators) of the endogenous latent
variates in deviation form,
x → q x 1 vector of observed measures (indicators) of the exogenous latent
variates in deviation form,
Λ → p x m matrix of factor loadings for y;
y

Λx → q x n matrix of factor loadings for x;

ε → p x 1 vector of unique factors corresponding to y;
δ → q x 1 vector of unique factors corresponding to x;

assumptions:

(1) (I − Β) is nonsingular
(2) E(ς ) = 0, E (ε ) = 0, E (δ ) = 0
(3) Cov (ξ ,ς ') = 0, Cov (ξ ,ε ') = 0, Cov (ξ ,δ ') = 0
( 4) Cov (ς , ε ') = 0, Cov (ς , δ ') = 0, Cov (ε , δ ') = 0
 2

let
Cov(δ , δ ') = Θ δ Cov(ξ , ξ ') = Φ
ε
Cov(ε , ε ') = Θ Cov(ς ,ς ') = Ψ

then the implied variance-covariance matrix of the observed variables is:

 Σ yy Σ yx 
Σ ( Λx , Λy , Θ δ , Θ ε , Β, Γ , Φ , Ψ ) =  
 Σ xy Σ xx 

where:
Σ yy = Λy ( I − Β ) ( ΓΦΓ ' + Ψ ) ( I − Β ) ' Λy ' + Θ ε
−1 −1

Σ yx = Λy E ( ηξ ' ) Λx ' = Λy ( I − Β ) ΓΦΛx '

−1

Σ xy = Λx ΦΓ ' ( I − Β ) ' Λy '

−1

Σ xx = Λx Φ Λx ' + Θ δ

Figure 1:

ϕ 11
θ11δ
δ1 x1 1

ξ1
δ ψ11 ψ22 ψ33
θ22 δ2 x2 λ21 x
γ11
ϕ 21 ζ1 ζ2 ζ3
ϕ 22
δ
θ33 δ3 x3 1
γ12 β21 β32
ϕ 31 ξ2 η1 η2 η3
θ44δ
δ4 x4 λx42
1 λy41 1 λy62 1
ϕ 32
γ13 λy21 λy31 y5 y6 y7
θδ55 δ5 x5 1
λ63 x y1 y2 y3 y4
θδ66 δ6 x6 ξ3 ε5 ε6
θ77δ
δ7 x7 λ73 x
ε1 ε2 ε3 ε4
ϕ 33 θε55 θε66
θε11 θε22 θε33 θε44

[see Appendix A for the specification of this model]

 3

distinguish between the following two types of models:

• recursive models: Β is a lower triangular matrix and Ψ is diagonal (i.e., there are
no reciprocal paths, feedback loops, or correlated disturbances);

• nonrecursive models: Β is not lower triangular and/or Ψ is not diagonal (i.e.,

there are reciprocal paths, feedback loops, and/or correlated disturbances);

note that when

Λy = I m , Λx = I n , Θ δ = 0, Θ ε = 0

the specification of the model reduces to the usual structural equation model with
observed variables;

Why is it important to consider measurement error?

(i) consequences in bivariate correlation and regression:

consider the following model:

η = γ ξ +ζ
y =η + ε
x =ξ +δ

• correlation: from classical reliability theory we know that

ρ ( y, x )
ρ (η ,ξ ) =
Var (η ) Var (ξ )
Var ( y ) Var ( x )

since the denominator is smaller than or equal to one,

ρ (η ,ξ ) ≥ ρ ( y, x) ;

• regression: notice that Cov(η ,ξ ) =γ ϕ so that

Cov(η ,ξ )
γ=
ϕ
 4

what would happen if we mistakenly assumed that η and ξ were

measured without error by y and x; in this case we would regress y on
x and get

Cov( y, x) Cov(η , ξ )  ϕ 
γ *= = =γ  
Var ( x) Var (ξ ) +Var (δ ) ϕ + θ 

thus, for θ > 0, γ * <γ ;

it can also be shown that if we have only a sample of observations on y

and x, ρ ( y, x ) and γ * are inconsistent estimators of
ρ ( η, ξ ) and γ ;

(ii) consequences in multiple regression

in general, Γ * is an inconsistent estimator of Γ; the multiple correlation

coefficient is attenuated if explanatory variables are measured with error;

Identification:

(i) a necessary condition for identification is that the number of parameters to be

freely estimated not be greater than the number of distinct elements in the
variance-covariance matrix of y and x;

(ii) a sufficient condition for identification is the two-step rule:

• in the first step, ignore the particular structural specification of interest

and consider a confirmatory factor analysis model in which the factors
are allowed to correlate freely; identify the free parameters in
Λx , Λy , Θ δ , Θ ε , Φ , the variance-covariance matrix of η, and the
variance-covariance matrix of η and ξ (see the handout on
confirmatory factor analysis for details);

• if the model in step one is identified, proceed to the second step;

assume that η and ξ are directly observable and show that the
structural model is identified in this case:

• null B rule [sufficient but not necessary]: if B is zero, the

elements in Γ, Φ, and Ψ are identified;
 5

• recursive rule [sufficient but not necessary]: if B can be

written as a lower-triangular matrix and Ψ is diagonal, the
latent variable model is identified;

• rank condition: assumes that all elements of Ψ are freely

estimated;

• solve for the free parameters in Β, Γ, and Ψ in terms of the

“known” elements of the variance-covariance matrix of η and
ξ;

(b) Estimation:

the goal is to find values for the unknown parameters in Λx, Λy, Θδ, Θε, Β, Γ, Φ and Ψ,
based on S, such that the variance-covariance matrix Σ implied by the estimated
parameters will be as “close” as possible to the observed variance-covariance matrix S;

as in the estimation of confirmatory factor models, the most commonly used estimation
procedures are ULS, ML, and GLS (refer to the handout on factor analysis for details);

(c) Testing:

1. Global fit measures:

(a) χ 2 goodness-of-fit test:

test of the hypothesis that the specified model is correct against the
alternative that Σ is unconstrained;

(b) alternative fit indices:

we will discuss these in a separate handout;

 6

2. Local fit measures:

(a) parameter estimates and associated standard errors

(b) reliability and discriminant validity

(c) explained variation for each structural equation:

ψ
R 2ηi = 1 −  ii
Var ( ηi )

3. Model modification:

(a) modification indices

(b) residual analysis

 7

Appendix A: Specification of the model in Figure 1

 η1  0 0 0  η1   γ 11 γ 12 γ 13   ξ1   ς 1 
         
 η2  = β 21 0 0  η2  +  0 0 0  ξ 2  + ς 2 
 η3   0 β 32 0  η3   0 0 0  ξ 3  ς 3 

 x1   1 0 0  δ1 
x  λ 0 0  δ 
 2  21   2
x3   0 1 0   ξ1  δ 3 
       
x 4  =  0 λ 32 0  ξ 2  + δ 4 
 x5   0 0 1  ξ 3  δ 5 
     
x6   0 0 λ 63  δ 6 
x   0 λ 73  δ 
 7  0  7

 y1   1 0 0  ε1 
y  λ 0 0 ε 
 2  21   2
y3  λ 31 0 0  η1  ε 3 
       
 y 4  = λ 41 0 0  η2  +  ε 4 
 y5   0 1 0  η3  ε 5 
     
y6   0 λ 62 0 ε 6 
y   0 1 0
 7  0  

ϕ11  ψ 11 0 0 

Φ = ϕ 21 ϕ 22  
Ψ = 0 ψ 22 0 
   
ϕ 31 ϕ 32 ϕ 33   0 0 ψ 33 

Θδ = Diag [θ11 .... θ 77 ]

Θε = Diag [θ11 ... θ 66 0]

 8
EXPLAINING CONSUMERS’ USAGE OF COUPONS FOR GROCERY SHOPPING
(cf. Bagozzi, Baumgartner, and Yi, JCR 1992)

Procedure

Female staff members at two American universities completed two questionnaires that were sent to them via campus mail. The first
questionnaire contained measures of seven beliefs about the consequences of using coupons and corresponding evaluations, as well as measures of
attitude toward using coupons, behavioral intentions, and the personality variable of state-/action-orientation.
One week later a second questionnaire was sent to those people who had participated in the first wave of data collection. This
questionnaire assessed some of the same variables as in wave one as well as people’s self-reported coupon usage during the past week.
Specifically, subjects were presented with a table that had 21 product categories as its rows (e.g., cereal, juice drinks, paper towels, snack foods,
canned goods) and six sources of coupons as its columns (i.e., direct mail, newspapers, magazines, in or on packages, from store displays or flyers,
from relatives or friends). An additional row was included for “other” products so that respondents could indicate usage in categories not covered
by the 21 listed. Subjects were asked to state how many coupons they had used for each category and source combination.

Measures

(1) beliefs: perceived likelihood of the following consequences of using coupons (rated on 7-point unlikely-likely scales):
• inconveniences:
o searching for, gathering, and organizing coupons takes much time and effort;
o planning the use of and actually redeeming coupons in the supermarket takes much time and effort;
• rewards:
o using coupons saves much money on the grocery bill;
o using coupons leads to feelings of being a thrifty shopper;
• encumbrances:
o in order to obtain coupons one has to subscribe to extra newspapers, magazines, etc.;
o in order to take advantage of coupon offers one has to purchase nonpreferred brands;
o in order to take advantage of coupon offers one has to shop at multiple supermarkets;
(2) evaluations: how each of the seven consequences of using coupons makes the respondent feel, rated on 7-point good-bad scales;
(3) Aact: attitude toward using coupons for shopping in the supermarket during the upcoming week (assessed on four semantic differential scales,
i.e., unpleasant-pleasant, bad-good, foolish-wise, and unfavorable-favorable); measured twice (week 1, week 2);
(4) BI: behavioral intentions to use coupons for shopping in the supermarket during the upcoming week (measured with a 7-point unlikely-likely
scale assessing intentions to use coupons and an 11-point no chance-certain scale asking about plans to use coupons);
(5) actual coupon usage: the total number of coupons used across product categories and sources; a square root transformation was used to
normalize the variable;
 9

SIMPLIS SPECIFICATION:

Title
A general structural equation model (explaining coupon usage)
Observed Variables
id be1 be2 be3 be4 be5 be6 be7 aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1
Raw Data from File=d:\EDEN\sem.dat
Latent Variables
INCONV REWARDS ENCUMBR AACT BI BH
Sample Size 250
Relationships
be1 = 1*INCONV
be2 = INCONV
be3 = 1*REWARDS
be4 = REWARDS
be5 = 1*ENCUMBR
be6 = ENCUMBR
be7 = ENCUMBR
aa1t1 = 1*AACT
aa2t1 = AACT
aa3t1 = AACT
aa4t1 = AACT
bi1 = 1*BI
bi2 = BI
bh1 = 1*BH
AACT = INCONV REWARDS ENCUMBR
BI = AACT
BH = BI
Set the Error Variance of bh1 to zero
Options sc rs mi wp
Path Diagram
End of Problem
 10

Covariance Matrix

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1 be1 be2 be3
-------- -------- -------- -------- -------- -------- -------- -------- -------- --------
aa1t1 1.86
aa2t1 1.20 1.71
aa3t1 1.05 1.01 1.61
aa4t1 1.33 1.38 1.09 2.03
bi1 1.20 1.34 1.06 1.31 3.93
bi2 1.37 1.47 1.28 1.45 3.23 3.77
bh1 0.71 0.70 0.61 0.84 1.48 1.57 2.13
be1 -0.80 -0.68 -0.48 -0.64 -0.99 -1.14 -0.36 2.66
be2 -0.86 -0.69 -0.51 -0.65 -1.03 -1.10 -0.30 2.07 2.64
be3 0.50 0.69 0.54 0.72 0.73 0.94 0.74 -0.16 -0.16 1.78
be4 0.42 0.60 0.54 0.61 0.75 0.81 0.58 -0.16 -0.19 1.09
be5 -0.27 -0.27 -0.20 -0.28 -0.28 -0.48 -0.33 0.70 0.54 -0.20
be6 -0.69 -0.47 -0.58 -0.62 -0.76 -1.04 -0.79 1.11 1.05 -0.78
be7 -0.71 -0.52 -0.58 -0.61 -0.96 -0.96 -0.54 1.11 1.10 -0.26

Covariance Matrix

be4 be5 be6 be7

-------- -------- -------- --------
be4 1.87
be5 -0.02 3.67
be6 -0.46 1.57 4.48
be7 -0.17 1.33 2.24 3.87

Total Variance = 38.006 Generalized Variance = 383.934

Largest Eigenvalue = 14.391 Smallest Eigenvalue = 0.452

Condition Number = 5.643

 11

LISREL Estimates (Maximum Likelihood)

Measurement Equations

aa1t1 = 1.000*AACT, Errorvar.= 0.679 , R² = 0.635

Standerr (0.0748)
Z-values 9.076
P-values 0.000

aa2t1 = 1.036*AACT, Errorvar.= 0.445 , R² = 0.740

Standerr (0.0691) (0.0577)
Z-values 14.997 7.715
P-values 0.000 0.000

aa3t1 = 0.849*AACT, Errorvar.= 0.761 , R² = 0.528

Standerr (0.0698) (0.0773)
Z-values 12.169 9.841
P-values 0.000 0.000

aa4t1 = 1.105*AACT, Errorvar.= 0.594 , R² = 0.708

Standerr (0.0756) (0.0723)
Z-values 14.608 8.221
P-values 0.000 0.000

bi1 = 1.000*BI, Errorvar.= 0.969 , R² = 0.753

Standerr (0.137)
Z-values 7.051
P-values 0.000

bi2 = 1.090*BI, Errorvar.= 0.248 , R² = 0.934

Standerr (0.0575) (0.127)
Z-values 18.950 1.950
P-values 0.000 0.051

bh1 = 1.000*BH, R² = 1.000

be1 = 1.000*INCONV, Errorvar.= 0.559 , R² = 0.790

Standerr (0.168)
Z-values 3.323
P-values 0.001

be2 = 0.983*INCONV, Errorvar.= 0.611 , R² = 0.769

Standerr (0.0866) (0.164)
Z-values 11.347 3.716
P-values 0.000 0.000

be3 = 1.000*REWARDS, Errorvar.= 0.451 , R² = 0.746

Standerr (0.177)
Z-values 2.553
P-values 0.011

be4 = 0.824*REWARDS, Errorvar.= 0.964 , R² = 0.483

Standerr (0.119) (0.145)
Z-values 6.906 6.641
P-values 0.000 0.000
 12

be5 = 1.000*ENCUMBR, Errorvar.= 2.785 , R² = 0.241

Standerr (0.279)
Z-values 9.992
P-values 0.000

be6 = 1.726*ENCUMBR, Errorvar.= 1.851 , R² = 0.587

Standerr (0.273) (0.336)
Z-values 6.311 5.501
P-values 0.000 0.000

be7 = 1.483*ENCUMBR, Errorvar.= 1.921 , R² = 0.503

Standerr (0.235) (0.279)
Z-values 6.308 6.888
P-values 0.000 0.000

Structural Equations

AACT = - 0.277*INCONV + 0.438*REWARDS - 0.0498*ENCUMBR, Errorvar.= 0.686 , R² = 0.418

Standerr (0.0579) (0.0808) (0.0969) (0.105)
Z-values -4.780 5.427 -0.513 6.535
P-values 0.000 0.000 0.608 0.000

BI = 1.103*AACT, Errorvar.= 1.528 , R² = 0.484

Standerr (0.110) (0.197)
Z-values 10.063 7.741
P-values 0.000 0.000

BH = 0.492*BI, Errorvar.= 1.412 , R² = 0.337

Standerr (0.0486) (0.131)
Z-values 10.122 10.797
P-values 0.000 0.000

NOTE: R² for Structural Equations are Hayduk's (2006) Blocked-Error R²

Reduced Form Equations

AACT = - 0.277*INCONV + 0.438*REWARDS - 0.0498*ENCUMBR, Errorvar.= 0.686, R² = 0.418

Standerr (0.0579) (0.0808) (0.0969)
Z-values -4.780 5.427 -0.513
P-values 0.000 0.000 0.608

BI = - 0.305*INCONV + 0.484*REWARDS - 0.0549*ENCUMBR, Errorvar.= 2.362, R² = 0.202

Standerr (0.0671) (0.0949) (0.107)
Z-values -4.550 5.097 -0.513
P-values 0.000 0.000 0.608

BH = - 0.150*INCONV + 0.238*REWARDS - 0.0270*ENCUMBR, Errorvar.= 1.984, R² = 0.0682

Standerr (0.0352) (0.0506) (0.0527)
Z-values -4.264 4.703 -0.513
P-values 0.000 0.000 0.608
 13

Covariance Matrix of Independent Variables

INCONV REWARDS ENCUMBR

-------- -------- --------
INCONV 2.105
(0.282)
7.454

REWARDS -0.175 1.328

(0.127) (0.230)
-1.373 5.764

ENCUMBR 0.674 -0.297 0.884

(0.147) (0.100) (0.247)
4.590 -2.983 3.579

Covariance Matrix of Latent Variables

AACT BI BH INCONV REWARDS ENCUMBR

-------- -------- -------- -------- -------- --------
AACT 1.179
BI 1.300 2.962
BH 0.640 1.457 2.129
INCONV -0.693 -0.764 -0.376 2.105
REWARDS 0.645 0.712 0.350 -0.175 1.328
ENCUMBR -0.361 -0.398 -0.196 0.674 -0.297 0.884

Log-likelihood Values

Estimated Model Saturated Model

--------------- ---------------
Number of free parameters(t) 35 105
-2ln(L) 5081.621 4987.618
AIC (Akaike, 1974)* 5151.621 5197.618
BIC (Schwarz, 1978)* 5274.873 5567.371

*LISREL uses AIC= 2t - 2ln(L) and BIC = tln(N)- 2ln(L)

 14

Goodness of Fit Statistics

Degrees of Freedom for (C1)-(C2) 70

Maximum Likelihood Ratio Chi-Square (C1) 94.003 (P = 0.0294)
Browne's (1984) ADF Chi-Square (C2_NT) 92.969 (P = 0.0346)
Estimated Non-centrality Parameter (NCP) 24.003
90 Percent Confidence Interval for NCP (2.698 ; 53.380)

Minimum Fit Function Value 0.376

Population Discrepancy Function Value (F0) 0.0960
90 Percent Confidence Interval for F0 (0.0108 ; 0.214)
Root Mean Square Error of Approximation (RMSEA) 0.0370
90 Percent Confidence Interval for RMSEA (0.0124 ; 0.0552)
P-Value for Test of Close Fit (RMSEA < 0.05) 0.871

Expected Cross-Validation Index (ECVI) 0.656

90 Percent Confidence Interval for ECVI (0.571 ; 0.774)
ECVI for Saturated Model 0.840
ECVI for Independence Model 12.158

Chi-Square for Independence Model (91 df) 3011.470

Normed Fit Index (NFI) 0.969

Non-Normed Fit Index (NNFI) 0.989
Parsimony Normed Fit Index (PNFI) 0.745
Comparative Fit Index (CFI) 0.992
Incremental Fit Index (IFI) 0.992
Relative Fit Index (RFI) 0.959

Critical N (CN) 267.011

Root Mean Square Residual (RMR) 0.133

Standardized RMR 0.0494
Goodness of Fit Index (GFI) 0.950
Adjusted Goodness of Fit Index (AGFI) 0.924
Parsimony Goodness of Fit Index (PGFI) 0.633
 15

Summary Statistics for Standardized Residuals

Smallest Standardized Residual = -3.193

Median Standardized Residual = 0.000
Largest Standardized Residual = 3.729

Stemleaf Plot

- 3|2
- 2|76
- 2|4310
- 1|9987766555
- 1|2110
- 0|99888776655
- 0|443222210000000000000000
0|11112222333444
0|55556667888889
1|0012223444
1|55666689
2|0
2|7
3|
3|7
Largest Negative Standardized Residuals
Residual for be3 and aa1t1 -2.672
Residual for be6 and be3 -3.193
Largest Positive Standardized Residuals
Residual for be3 and bh1 3.729
Residual for be4 and bh1 2.680

Modification Indices and Expected Change

Modification Indices for LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 - - 0.638 0.025
aa2t1 - - 0.011 0.683
aa3t1 - - 0.750 0.000
aa4t1 - - 3.738 0.634
bi1 1.654 - - 0.120
bi2 0.117 - - 1.663
bh1 2.332 - - - -

Expected Change for LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 - - -0.043 -0.007
aa2t1 - - 0.005 -0.032
aa3t1 - - 0.047 0.000
aa4t1 - - -0.104 0.034
bi1 -0.184 - - 0.025
bi2 0.055 - - -0.098
bh1 0.173 - - - -
 16

Standardized Expected Change for LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 - - -0.074 -0.010
aa2t1 - - 0.009 -0.046
aa3t1 - - 0.080 0.000
aa4t1 - - -0.179 0.050
bi1 -0.200 - - 0.036
bi2 0.059 - - -0.142
bh1 0.188 - - - -

Completely Standardized Expected Change for LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 - - -0.054 -0.007
aa2t1 - - 0.007 -0.035
aa3t1 - - 0.063 0.000
aa4t1 - - -0.126 0.035
bi1 -0.101 - - 0.018
bi2 0.031 - - -0.073
bh1 0.129 - - - -

Modification Indices for LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 - - 0.043 0.292
be2 - - 0.043 0.292
be3 0.109 - - 0.942
be4 0.110 - - 0.946
be5 0.136 1.343 - -
be6 1.737 7.662 - -
be7 2.621 4.272 - -

Expected Change for LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 - - 0.014 0.083
be2 - - -0.014 -0.081
be3 0.022 - - -0.119
be4 -0.018 - - 0.098
be5 -0.038 0.134 - -
be6 -0.181 -0.365 - -
be7 0.193 0.242 - -
 17

Standardized Expected Change for LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 - - 0.017 0.078
be2 - - -0.016 -0.076
be3 0.032 - - -0.112
be4 -0.026 - - 0.092
be5 -0.056 0.154 - -
be6 -0.262 -0.421 - -
be7 0.280 0.279 - -

Completely Standardized Expected Change for LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 - - 0.010 0.048
be2 - - -0.010 -0.047
be3 0.024 - - -0.084
be4 -0.019 - - 0.068
be5 -0.029 0.081 - -
be6 -0.124 -0.199 - -
be7 0.142 0.142 - -

The Modification Indices Suggest to Add the

Path to from Decrease in Chi-Square New Estimate
AACT BI 10.7 -0.28

Modification Indices for BETA

AACT BI BH
-------- -------- --------
AACT - - 10.749 1.515
BI - - - - 2.314
BH 2.332 - - - -

Expected Change for BETA

AACT BI BH
-------- -------- --------
AACT - - -0.278 -0.074
BI - - - - -0.169
BH 0.173 - - - -

Standardized Expected Change for BETA

AACT BI BH
-------- -------- --------
AACT - - -0.149 -0.047
BI - - - - -0.067
BH 0.109 - - - -

The Modification Indices Suggest to Add the

Path to from Decrease in Chi-Square New Estimate
BH REWARDS 12.7 0.29
 18

Modification Indices for GAMMA

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT - - - - - -
BI 5.558 3.063 5.144
BH 1.611 12.655 2.777

Expected Change for GAMMA

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT - - - - - -
BI -0.172 0.181 -0.266
BH 0.076 0.288 -0.163

Standardized Expected Change for GAMMA

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT - - - - - -
BI -0.145 0.121 -0.145
BH 0.075 0.228 -0.105

No Non-Zero Modification Indices for PHI

The Modification Indices Suggest to Add an Error Covariance

Between and Decrease in Chi-Square New Estimate
BI AACT 11.0 -0.43

Modification Indices for PSI

AACT BI BH
-------- -------- --------
AACT - -
BI 10.977 - -
BH 0.009 2.333 - -

Expected Change for PSI

AACT BI BH
-------- -------- --------
AACT - -
BI -0.434 - -
BH -0.009 -0.240 - -

Standardized Expected Change for PSI

AACT BI BH
-------- -------- --------
AACT - -
BI -0.233 - -
BH -0.006 -0.096 - -

The Modification Indices Suggest to Add an Error Covariance

Between and Decrease in Chi-Square New Estimate
be6 be3 8.5 -0.34
 19

Modification Indices for THETA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
aa1t1 - -
aa2t1 1.262 - -
aa3t1 1.619 0.728 - -
aa4t1 1.046 3.374 0.375 - -
bi1 0.263 0.627 1.709 0.006 - -
bi2 0.020 0.058 3.112 3.558 2.005 - -
bh1 0.033 0.918 0.153 2.970 0.120 1.670 - -

Expected Change for THETA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
aa1t1 - -
aa2t1 -0.063 - -
aa3t1 0.072 -0.045 - -
aa4t1 0.062 0.109 -0.035 - -
bi1 -0.032 0.043 -0.083 0.005 - -
bi2 -0.008 -0.012 0.100 -0.104 0.457 - -
bh1 0.013 -0.057 -0.028 0.116 0.035 -0.138 - -

Completely Standardized Expected Change for THETA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
aa1t1 - -
aa2t1 -0.035 - -
aa3t1 0.042 -0.027 - -
aa4t1 0.032 0.058 -0.020 - -
bi1 -0.012 0.017 -0.033 0.002 - -
bi2 -0.003 -0.005 0.041 -0.038 0.119 - -
bh1 0.006 -0.030 -0.015 0.056 0.012 -0.049 - -

Modification Indices for THETA-DELTA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
be1 0.023 0.114 0.797 0.517 0.519 2.214 0.062
be2 1.619 0.029 0.056 0.422 1.294 0.059 1.757
be3 2.062 0.292 0.982 0.118 3.408 2.254 4.922
be4 1.229 0.122 1.657 0.167 1.884 0.563 0.039
be5 0.330 0.434 0.321 0.000 1.189 0.536 0.054
be6 0.020 3.252 0.349 0.002 2.655 0.831 4.351
be7 1.058 0.326 0.785 0.000 3.489 1.020 0.163
 20

Expected Change for THETA-DELTA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
be1 0.009 -0.017 0.053 0.041 0.049 -0.090 0.019
be2 -0.074 0.009 0.014 0.037 -0.077 0.015 0.101
be3 -0.085 0.028 -0.059 0.020 -0.127 0.092 0.172
be4 -0.070 0.019 0.082 -0.025 0.101 -0.049 0.016
be5 0.057 -0.057 0.058 -0.002 0.129 -0.077 -0.031
be6 -0.013 0.151 -0.058 0.004 0.186 -0.093 -0.268
be7 -0.094 0.045 -0.083 -0.001 -0.203 0.097 0.049

Completely Standardized Expected Change for THETA-DELTA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
be1 0.004 -0.008 0.025 0.018 0.015 -0.028 0.008
be2 -0.033 0.004 0.007 0.016 -0.024 0.005 0.043
be3 -0.047 0.016 -0.035 0.011 -0.048 0.036 0.088
be4 -0.037 0.011 0.047 -0.013 0.037 -0.019 0.008
be5 0.022 -0.023 0.024 -0.001 0.034 -0.021 -0.011
be6 -0.005 0.055 -0.022 0.001 0.044 -0.023 -0.087
be7 -0.035 0.018 -0.033 0.000 -0.052 0.026 0.017

Modification Indices for THETA-DELTA

be1 be2 be3 be4 be5 be6 be7

-------- -------- -------- -------- -------- -------- --------
be1 - -
be2 - - - -
be3 0.000 0.129 - -
be4 0.001 0.251 - - - -
be5 0.961 1.390 0.004 1.156 - -
be6 0.001 0.080 8.458 0.140 0.292 - -
be7 0.031 0.314 4.162 0.103 0.057 0.854 - -

Expected Change for THETA-DELTA

be1 be2 be3 be4 be5 be6 be7

-------- -------- -------- -------- -------- -------- --------
be1 - -
be2 - - - -
be3 0.001 0.024 - -
be4 0.002 -0.035 - - - -
be5 0.110 -0.132 -0.007 0.129 - -
be6 0.004 -0.033 -0.340 0.044 0.150 - -
be7 -0.019 0.060 0.220 0.036 0.057 -0.451 - -
 21

Completely Standardized Expected Change for THETA-DELTA

be1 be2 be3 be4 be5 be6 be7

-------- -------- -------- -------- -------- -------- --------
be1 - -
be2 - - - -
be3 0.001 0.011 - -
be4 0.001 -0.016 - - - -
be5 0.035 -0.042 -0.003 0.049 - -
be6 0.001 -0.010 -0.120 0.015 0.037 - -
be7 -0.006 0.019 0.084 0.013 0.015 -0.108 - -

Maximum Modification Index is 12.65 for Element ( 3, 2) of GAMMA

A general structural equation model (explaining coupon usage)

Standardized Solution

LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 1.086 - - - -
aa2t1 1.124 - - - -
aa3t1 0.922 - - - -
aa4t1 1.199 - - - -
bi1 - - 1.721 - -
bi2 - - 1.875 - -
bh1 - - - - 1.459

LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 1.451 - - - -
be2 1.426 - - - -
be3 - - 1.152 - -
be4 - - 0.949 - -
be5 - - - - 0.940
be6 - - - - 1.623
be7 - - - - 1.395

BETA

AACT BI BH
-------- -------- --------
AACT - - - - - -
BI 0.696 - - - -
BH - - 0.580 - -

GAMMA

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT -0.370 0.465 -0.043
BI - - - - - -
BH - - - - - -
 22

Correlation Matrix of ETA and KSI

AACT BI BH INCONV REWARDS ENCUMBR

-------- -------- -------- -------- -------- --------
AACT 1.000
BI 0.696 1.000
BH 0.404 0.580 1.000
INCONV -0.440 -0.306 -0.178 1.000
REWARDS 0.516 0.359 0.208 -0.105 1.000
ENCUMBR -0.354 -0.246 -0.143 0.494 -0.274 1.000

PSI
Note: This matrix is diagonal.

AACT BI BH
-------- -------- --------
0.582 0.516 0.663

Regression Matrix ETA on KSI (Standardized)

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT -0.370 0.465 -0.043
BI -0.257 0.324 -0.030
BH -0.149 0.188 -0.017

A general structural equation model (explaining coupon usage)

Completely Standardized Solution

LAMBDA-Y

AACT BI BH
-------- -------- --------
aa1t1 0.797 - - - -
aa2t1 0.860 - - - -
aa3t1 0.727 - - - -
aa4t1 0.841 - - - -
bi1 - - 0.868 - -
bi2 - - 0.966 - -
bh1 - - - - 1.000

LAMBDA-X

INCONV REWARDS ENCUMBR

-------- -------- --------
be1 0.889 - - - -
be2 0.877 - - - -
be3 - - 0.864 - -
be4 - - 0.695 - -
be5 - - - - 0.491
be6 - - - - 0.766
be7 - - - - 0.709
 23

BETA

AACT BI BH
-------- -------- --------
AACT - - - - - -
BI 0.696 - - - -
BH - - 0.580 - -

GAMMA

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT -0.370 0.465 -0.043
BI - - - - - -
BH - - - - - -

Correlation Matrix of ETA and KSI

AACT BI BH INCONV REWARDS ENCUMBR

-------- -------- -------- -------- -------- --------
AACT 1.000
BI 0.696 1.000
BH 0.404 0.580 1.000
INCONV -0.440 -0.306 -0.178 1.000
REWARDS 0.516 0.359 0.208 -0.105 1.000
ENCUMBR -0.354 -0.246 -0.143 0.494 -0.274 1.000

PSI
Note: This matrix is diagonal.

AACT BI BH
-------- -------- --------
0.582 0.516 0.663

THETA-EPS

aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh1

-------- -------- -------- -------- -------- -------- --------
0.365 0.260 0.472 0.292 0.247 0.066 - -

W_A_R_N_I_N_G: THETA-EPS is not positive definite

THETA-DELTA

be1 be2 be3 be4 be5 be6 be7

-------- -------- -------- -------- -------- -------- --------
0.210 0.231 0.254 0.517 0.759 0.413 0.497

Regression Matrix ETA on KSI (Standardized)

INCONV REWARDS ENCUMBR

-------- -------- --------
AACT -0.370 0.465 -0.043
BI -0.257 0.324 -0.030
BH -0.149 0.188 -0.017

Time used 0.031 seconds

 24

LISREL SPECIFICATION:

A general structural equation model (explaining coupon usage)

DA NI=15 NO=0
LA
id be1 be2 be3 be4 be5 be6 be7 aa1t1 aa2t1 aa3t1 aa4t1 bi1 bi2 bh
ra fi=d:\EDEN\sem.dat
se
9 10 11 12 13 14 15 2 3 4 5 6 7 8/
MO nx=7 nk=3 ny=7 ne=3 td=di,fr te=di,fr ga=fu,fi be=fu,fi ph=fr ps=di,fr
va 1 lx 1 1 lx 3 2 lx 5 3
fr lx 2 1 lx 4 2 lx 6 3 lx 7 3
va 1 ly 1 1 ly 5 2 ly 7 3
fr ly 2 1 ly 3 1 ly 4 1 ly 6 2
fi te 7 7
pa ga
1 1 1
0 0 0
0 0 0
pa be
0 0 0
1 0 0
0 1 0
lk
inconv rewards encumbr
le
aact bi bh
ou sc
 25

LOCAL FIT INDICES FOR THE MEASUREMENT MODEL

standardized individual- composite reliability

parameter
construct parameter parameter z-value item (average variance
estimate
estimate reliability extracted)

inconveniences .88 (.78)

λ x
11 1.00 0.89 -- 0.79
λx21 0.98 0.88 11.32 .77
θδ11 0.56 0.21 3.32 --
θδ22 0.61 0.23 3.71 --
rewards .76 (.61)
λx32 1.00 0.86 -- 0.75
λ x
42 0.82 0.70 6.89 0.48
δ
θ 33 0.45 0.25 2.55 --
δ
θ 44 0.96 0.52 6.63 --
encumbrances .70 (.45)
λx53 1.00 0.49 -- 0.24
λx63 1.73 0.77 6.30 0.59
λx73 1.48 0.71 6.30 0.50
θδ55 2.78 0.76 9.97 --
θδ66 1.85 0.41 5.49 --
θδ77 1.92 0.50 6.87 --
attitudes .88 (.66)
λ y
11 1.00 .80 -- 0.64
λ y
21 1.04 0.86 14.97 0.74
λ y
31 0.85 0.73 12.14 0.53
λ y
41 1.10 0.84 14.58 0.71
θε11 0.68 0.37 9.06 --
θε22 0.44 0.26 7.70 --
θε33 0.76 0.47 9.82 --
θε44 0.59 0.29 8.20 --
intentions .91 (.84)
λy42 1.00 0.87 -- 0.75
λy52 1.09 0.97 18.91 0.93
θε44 0.97 0.25 7.04 --
ε
θ 55 0.25 0.07 1.95 --
behavior
λy63 1.00 1.00 -- 1.00
θε66 0.00 0.00 -- --