Table 3.

Results of mediation testing using bootstrapping analyses (MEDIATE macro; Preacher and Hayes 2008) (N=55)

	B	SE	95 % Confidence intervals
Hypothesized mediation relationship (indirect effect) tested
Role model (vs control) → perceived identity compatibility between gender and pre-med → sense of belonging in pre-med	.28	.18	.01 and .71
Perceived identity compatibility between gender and pre-med → sense of belonging in pre-med → interest in being pre-med	.28	.13	.07 and .56
Perceived identity compatibility between gender and pre-med → sense of belonging in pre-med → interest in pursuing a career as a physician	.28	.14	.06 and .58
Reverse mediation relationship (indirect effect) tested
Role model (vs control) → sense of belonging in pre-med → perceived identity compatibility between gender and pre-med	.24	.17	−.02 and .64
Perceived identity compatibility between gender and pre-med → interest in being pre-med → sense of belonging in pre-med	.12	.07	.01 and .28
Perceived identity compatibility between gender and pre-med → interest in pursuing a career as a physician → sense of belonging in pre-med	.08	.06	−.02 and .21

Arrows indicate direction of effects for each mediation relationship (indirect effect) tested