Table 3.

Teacher EA as mediator of relationships between student problem behavior on study outcomes

Baron & Kenny’s four criteria for mediation	B	SE	t-ratio	p	r
Effect of predictor on outcome
Student problem behavior → depression	0.15	0.02	7.16	.013	.31
Student problem behavior → burnout – EE	0.48	0.04	10.84	.001	.43
Student problem behavior → burnout – DEP	0.50	0.04	13.02	<.001	.50
Student problem behavior → burnout – PA	−0.31	0.03	−11.77	.002	.46
Effect of predictor on mediator
Student problem behavior → EA	−0.47	0.03	−17.39	<.001	.61
Effect of mediator on outcome
EA → depression	0.30	0.03	12.35	<.001	.48
EA → burnout – EE	0.80	0.05	15.09	<.001	.55
EA → burnout – DEP	0.63	0.05	12.72	<.001	.49
EA → burnout – PA	−0.40	0.03	−11.91	<.001	.46
Effect of predictor on outcome controlling for mediator
Student problem behavior → depression | EA	0.01	0.02	0.35	.126	.02
Student problem behavior → burnout – EE | EA	0.17	0.05	3.27	.002	.14
Student problem behavior → burnout – DEP | EA	0.32	0.05	6.86	<.001	.29
Student problem behavior → burnout – PA | EA	−0.19	0.03	−5.95	<.001	.25

Notes. B = beta coefficient; SE = standard error; pr = partial regression coefficient; | = controlling for; EE = emotional exhaustion, DEP = depersonalization, PA = personal accomplishment, EA = experiential avoidance