leadership and motivation in hospitality

remove one of these (DSB3 or JP4) items and re-estimate the model. For the
same substantive rationale described above, JP4 is retained. Removing DSB3
does not radically alter the face validity of the DSB construct as DSB2 and DSB3
both deal with very similar aspects of extra effort directed towards high quality
customer service.

7.2.10 Measurement model: second respecification (CFA 1:3)

The measurement model is re-specified once again (without DSB3) and the
estimates are described in Table 7 - 6.

Factor loadings for ML and JP remain almost unchanged (movements of 0.001 to
ML2, JP2 and JP4). For DSB1 the factor loading has improved considerably (from
0.839 to 0.946) and DSB2 has dropped by a similar amount (from 0.776 to
0.686). All of the factor loadings are now satisfactory (JP4 remains lower than
ideal but is retained on substantive grounds).

Construct Item Standardised factor loading estimates
ML JP DSB

Motivational
Leadership

ML1 .903^

ML2 .958^

ML3^ .873

Job Performance

JP1 .868
JP2 .858
JP4^ .458
Discretionary
Service Behaviour

DSB1 .946^
DSB2^ .686^
Average variance
extracted (AVE) .767^ .618^ .733^
Construct
reliability (CR) .908^ .820^ .842^

Model fit statistics

χ^2 = 15.979; d.f. = 17; sig = 0.525

RMSEA = 0.000 (0.059; 0.000; pclose = 0.899)

CFI = 1.000

SRMR = 0.0314

CN (0.05) = 366

Table 7-6 Estimates for CFA 1:3

The fit measures for the model are all satisfactory:

 model χ^2 is non-significant at 0.525 (indicating that the observed covariance

estimates are not significantly different from those implied by the theoretical

model);