Table 2. Descriptive statistics, correlations among variables and measurement models in Study 2.

	M	SD	1	2	3	4	5	6	7	8	9	10
1. Illegitimate tasks	2.49	0.71	.88
2. Distributive justice	3.49	0.76	−.34	.82
3. Interactional and procedural justice	4.87	1.26	−.61	.34	.97
4. Resentment	2.97	1.51	.66	−.31	−.56	.93
5. Burnout	2.09	0.60	.62	−.34	−.53	.71	.86
6. Self-esteem	4.12	0.54	−.26	−.07	.05	−.30	−.34	.85
7. Sexa	0.56	0.50	.03	.09	.02	−.02	−.05	.05	–
8. Age	41.75	11.51	−.02	.09	−.14	.09	−.09	.05	.19	–
9. Hours workedb	91.49	16.18	.02	−.10	.01	−.01	−.04	.07	.34	.03	–
10. Job tenure	5.99	6.34	−.01	.11	−.10	.07	.04	−.04	.23	.56	.03	–
Measurement models	χ2		df		CFI		TLI		RMSEA		SRMR	AIC
Predictors model 1 (facets as parcels)	4.87		6		1		1		.000		.016	3046.00
Predictors model 2 (2nd order)	107.25		49		.962		.949		.073		.050	6007.68
Predictors model 3 (two variables)	106.94		48		.962		.948		.074		.050	6079.36
Predictors model 4 (one variable)	357.30		51		.802		.744		.164		.089	6323.72
Dependent variables model	19.16		17		.998		.997		.024		.025	3486.38

Note: Pearson correlations; Cronbach's α's are presented in the diagonal (given in italics). N = 224; r ≥ │.14│ p ≤ .05; r ≥ │.19│ = p ≤ .01 (two-tailed).

Predictor models contain all work characteristics (illegitimate tasks, distributive justice, and interactional & procedural justice); Model 1 = illegitimate tasks as single construct with the two subscales as indicators; Model 2 = illegitimate tasks as second-order construct, with unnecessary and unreasonable tasks as first-order constructs; Model 3 = unnecessary and unreasonable tasks as two separate constructs (r _unn−unr = .55, p ≤ .01); Model 4 = illegitimate tasks as single construct with the eight items as indicators; the dependent variables model contains feelings of resentment, burnout and self-esteem.

^a0 = f; 1 = m; ^bin percent of a full time equivalent (FTE).