Table 5. Impact of PFL on emotional and behavioral functioning—Treatment effects by gender.

	Boysa					Girlsb
CBCL Cutoff Scores	Log.p-value1	IPW adj. Log.p-value2	Perm p-value3	IPW adj. perm p-value4	ES	Log. p-value1	IPW adj. Log. p-value2	Perm. p-value3	IPW adj. perm p-value4	ES
Internalizing Problems Cutoff	0.207	0.173	0.268	0.208	0.15	0.412	0.335	0.388	0.443	0.09
Externalizing Problems Cutoff	0.058	0.071	0.078	0.098	0.24	0.245	0.367	0.254	0.254	0.12
Total Problems Cutoff	0.005**	0.023**	0.005**	0.001**	0.33	0.492	0.325	0.447	0.083	0.07

Note:

^a n = 70 (intervention 41; control 29),

^b N = 94 (intervention 40; control 54).

Perm = permutation test. ‘ES’ = Cramer’s Phi effect size.

¹ two-tailed p-value from a logistic regression. For boys Total Problems cutoff a LPM model was fitted rather than a logistic regression as treatment status was a perfect predictor of being in the cutoff category.

² two-tailed p-value from a logistic regression (again a LPM model was used for boys Total Problems cutoff) applying inverse probability weights.

³ two-tailed p-value from a permutation test with 100,000 replications.

⁴ two-tailed p-value from a permutation test with 100,000 replications applying inverse probability weights.

* p <.05,

** p <.01 level.