Table 7.

Stepwise, logistic regression analysis for predicting women's sexual orientation based on gender-related scales in study 2 at first measurement.

	Step 1	Step 2	Step 3
Predictors:	B (SE) Exp(B)	B (SE) Exp(B)	B (SE) Exp(B)
GEPAQ-F	[0.50 (0.58) 1.66]	[0.39 (0.76) 1.48]	[0.72 (0.92) 2.05]
GEPAQ-M	[−0.37 (0.72) 0.69]	[−0.86 (0.87) 0.43]	[−0.95 (0.94) 0.39]
GRB-F		3.01 (0.97) 20.30	3.69 (1.38) 40.15
GRB-M		−1.13 (0.56) −0.52	[−1.43 (0.80) 0.24]
TMF			[0.35 (0.66) 1.42]
CGRB			[−0.90 (0.58) 0.41]
	${\chi }_{\left(2\right)}^2$ = 1.01, p = 0.605, R2 = 0.03, 56%	${\chi }_{\left(2\right)}^2$ = 18.68, p < 0.001, R2 = 0.47, 73%	${\chi }_{\left(2\right)}^2$ = 2.99, p = 0.224, R2 = 0.53, 78%

Chosen method was “Forward: Wald” in each block. R² means Nagelkerke's R². Percentage values refer to correctly classified lesbian and straight women. All correlations are statistically significant at α = 0.05 except for those in brackets. TMF, Traditional Masculinity-Femininity; GEPAQ, German Extended Personality Attributes Questionnaire; GRB, Gender-Role Behavior; and CGRB, Childhood Gender-Role Behavior. Endings indicate masculinity (−M) und femininity (−F) scales. Higher scores indicate higher masculinity on masculinity scales, higher femininity on femininity scales and TMF, and higher gender-conformity on CGRB.