leadership and motivation in hospitality

Construct Item Standardised factor loading es timates

ML ME JP DSB

Motivational

Leadership

ML1 .904 (^)
ML2 .953 (^)
ML3 .878 (^)
Work Meaning
ME1 (^) .892 (^)
ME3 .871 (^)
ME6 .778 (^)
Job Performance
JP1 .872 (^)
JP2 .854 (^)
JP4 .458 (^)
Discretionary
Service Behaviour
DSB1 (^) .845
DSB2 .773
DSB3.^614
Model fit statistics
χ^2 = 69.398; d.f. = 48; sig = 0.023
RMSEA = 0.046 (0.068; 0.018; pclose = 0.592)
CFI = 0.985
SRMR = 0.0433
CN (0.05) = 200

Table 7-11 Estimates for CFA 2:2

The model fit statistics have improved considerably, only the χ^2 p value is not
satisfactory at <0.05. Examining the factor loadings flags DSB3 as a potential
problem with a moderate loading of 0.614 and an inspection of the standardised
residual covariance matrix reveals that DSB3 shares a large (2.697) SRC value
(with JP4). Once again, retaining JP4 for substantive reasons leads to the
removal of DSB3. The model is re-estimated once again and the estimates are
described in Table 7 - 12.

All factor loadings are now satisfactory (although JP4 is, as usual, only of
moderate strength), all factor covariances are statistically significant and positive,
the model fit statistics are all satisfactory and the requirements for convergent
and discriminant validity are all met (AVEs >0.5; CRs > 0.7; and the highest
squared multiple correlation value is 0.307). Multivariate normality is assessed
and the C.R. value was estimated at 14.5; the subsequent bootstrapped
estimates indicated that the parameters are robust to multivariate non-normality
(see Appendix IV).

The substantive implications for the removal of items ML4, ML5, JP3, DSB3 and
DSB4 were described during the development of Model 1. The development of
the Model 2 has also seen the removal of ME5 and ME7 from the Work Meaning
factor – how have these modifications affected the substantive content of Work
Meaning?