Research Method

Assignment 2
Syeda Areeba
U23200046

Measurement Model:

Factor Loadings:
None of the items in the dataset have the factor loadings less than the recommended value of 0.50 in the
below table, therefore we exclude none of the items.

COMM COMP INT PLN PM POM POP PS RM TM USE

COMM1 0.831
COMM2 0.820
COMM3 0.825
COMM4 0.819
COMM5 0.792
COMM6
COMP1 0.801
COMP2 0.886
COMP3
 INT1 0.882
INT2 0.856
INT3 0.797
PLN1 0.721
PLN2 0.799
PLN3 0.808
PLN4 0.876
PLN5
POM1
POM2
POM3 0.848
POM4 0.753
POM5 0.765
POP1
POP2 0.866
POP3
POP4 0.763
RM1 0.737
RM2 0.795
RM3 0.833
RM4 0.775
RM5 0.806
RM6 0.767
USE1 0.723
USE10 0.772
USE2 0.812
USE3 0.783
USE4 0.777
USE5 0.765
USE6 0.767
USE7 0.726
USE8
USE9

Indicator Multicollinearity:
In the below table we see that the value of VIF (indicator of multicollinearity) is below 5 and therefore
satisfies the criteria of VIF being below 5.

VIF
COMM1 2.349
COMM2 2.531
COMM2 2.344
COMM3 2.401
COMM3 2.536
COMM4 2.484
COMM4 2.748
COMM5 1.774
COMM5 2.134
COMM6 1.854
COMP1 1.594
COMP1 1.457
COMP2 1.783
COMP2 1.448
COMP3 1.875
COMP3 1.419
INT1 2.151
INT1 1.755
INT2 1.868
INT2 2.276
INT3 1.626
INT3 2.113
PLN1 2.110
PLN1 1.978
PLN2 2.418
PLN2 1.883
PLN3 2.826
PLN3 2.428
PLN4 1.772
PLN4 2.275
PLN5 1.895
PLN5 2.149
POM1 2.016
POM1 1.802
POM2 1.344
POM2 1.276
POM3 2.482
POM3 2.396
POM4 2.211
POM4 2.140
POM5 1.759
POM5 1.740
POP1 1.465
POP1 1.492
POP2 2.116
POP2 1.986
POP3 1.560
POP3 1.399
POP4 1.477
 POP4 1.561
RM1 2.014
RM1 1.914
RM2 2.807
RM2 2.425
RM3 2.459
RM3 2.150
RM4 2.312
RM4 2.206
RM5 2.352
RM5 3.056
RM6 1.837
RM6 1.540
USE1 2.436
USE1 2.839
USE10 3.363
USE10 3.231
USE2 3.297
USE2 3.192
USE3 3.036
USE3 2.877
USE4 2.842
USE4 2.623
USE5 2.926
USE5 2.679
USE6 3.115
USE6 2.930
USE7 2.096
USE7 2.323
USE8 2.584
USE8 2.774
USE9 2.179
USE9 2.508

Reliability Analysis:
In the below table, we see that the values of Cronbach Alpha and Composite Reliability has the range of
0.912 - 0.763 & 0.776-0.918 respectively thereby satisfying the criteria of the values being over the
required threshold of 0.70

Composite
Cronbach' reliability
s alpha (rho_a)
PM 0.801 0.817
POM 0.786 0.821
 POP 0.763 0.776
PS 0.910 0.913
TM 0.912 0.918

Construct Validity:
Convergent Validity:
The values of AVE in the below table are over 0.50 for POM and POP and the rest are below, hence the
factors PM, PS and TM have a convergent validity issue.

Average
variance
extracted
(AVE)
PM 0.395
POM 0.551
POP 0.587
PS 0.412
TM 0.446

Discriminant Validity:
Fornell and Larcker Criterion:
This criterion ensures that each construct’s square root of the AVE is greater than its correlations with
other constructs. But PM (0.629) does not meet the criteria since the AVE values are lower than its
correlation showing validity issues.

PM POM POP PS TM
PM 0.629
POM 0.869 0.742
POP 0.798 0.396 0.766
PS 0.734 0.663 0.552 0.642
TM 0.346 0.217 0.347 0.417 0.668

Hetrotrait-Monotrait Ratio – HTMT

This ratio shows the discriminant validity by seeing the values below 0.85 or 0.9. The POM with Pm
exceeds by 1.088 and POP with PM (1.018) suggesting a lack of discriminant validity between the
constructs.

PM POM POP PS TM
PM
POM 1.088
POP 1.018 0.497
PS 0.856 0.787 0.653
 TM 0.534 0.449 0.457 0.518

Second Measurement Model

Path Co-efficient:

PM PS TM
0.59
PM 7
PS
0.50 0.22
TM 4 2

VIF:
Shows the range of 1.186 to1.708 which is below 5 and therefore is low in multicollinearity among
variables.

VIF
COM 1.51
M 5
COM 1.31
P 5
1.42
INT 2
1.39
PLN 2
1.18
POM 6
1.18
POP 6
1.70
RM 8
1.32
USE 1
 Outer T statistics P
weights (|O/STDEV|) values
PM ->
PS 0.597 9.097 0.000
TM ->
PM 0.504 7.917 0.000
TM ->
PS 0.222 2.697 0.004
Since all the P-values are shown under the criterion of 0.05, therefore the relationships are meaningful
and statistically significant in nature.
Prediction-Oriented Segmentation:
POS analysis shows group 1 with 24 items and group 2 with 158 items which shows differing effects in
the subgroups.

Group1 Group2
Number 24.000 158.000