Examining the factor loadings reveals several below the 0.700 level – these
include several that have, by now, become ‘usual suspects’, specifically, ML4,
ML5, ME5, ME7, JP3 and JP4. Additionally, in the Job Satisfaction factor, JS2, 3
and 4 are not loading very strongly (at 0.599. 0.456 and 0.610 respectively).
Construct^ Item^ Standardised factor loading est imates
ML ME JS AOC JP DSB
Motivational
Leadership
ML1 .891
ML2 .932
ML3 .898
ML4 .623
ML5 .623
Work Meaning
ME1 .873
ME3 .856
ME5. 611
ME6 .809
ME7.^282
Job Satisfaction
JS1 .774
JS2. 599
JS3. 456
JS4. 610
JS5 .732
JS6 .864
Affective
Organisational
Commitment
AOC1 (^) .827
AOC2 (^) .847
AOC 4 .886
Job Performance
JP1 .851
JP 2 .856
JP3. 556
JP4 .495
Discretionary
Service Behaviour
DSB1 .803
DSB2 .771
DSB3 .675
DSB4 .693
Model fit
statistics
χ^2 = 648.622; d.f. = 309; sig = 0.000
RMSEA = 0.072 (0.080; 0.064; pclose = 0.000)
CFI = 0.903
SRMR = 0.0696
CN (0.05) = 115
Table 7-14 Estimates for CFA 3:1
The standardised residual covariance matrix was examined and all of these low-
loading items were also associated with standardised residual covariance values
above greater than the ± 1.96 or ± 2.58 values that signal large volumes of
unmeasured (error) variances.
Accordingly, with the exception of JP4 (for the usual substantive reasons) all of
these low-loading indicators were removed and the model re-estimated. Table