Table 1. Selected measured covariate standardized biases between treatment and comparison groups pre- and post- propensity score (PS) adjustment.

	Standardized Bias * Pre-PS adjustment	Standardized Bias Post-PS adjustment **
Race
white	0.27	0.05
black	0.29	0.05
other	0.09	0.03
Gender
male	0.63	0.004
female	0.63	0.004
Family income
low	0.01	0.09
middle	0.10	0.09
high	0.06	0.03
missing	0.04	0.06
Alcohol problem use
no	0.85	0.11
yes	0.80	0.14
missing	0.13	0.06
Other illegal drug use
no	0.18	0.03
yes	0.16	0.07
missing	0.08	0.04
Daily cigarette user
no	1.15	0.08
yes	1.16	0.08
missing	0.11	0.01
Free lunch
no	0.09	0.15
yes	0.09	0.13
missing	0.02	0.05
Parent monitoring
none	0.02	0.08
low	0.02	0.09
moderate	0.02	0.08
high	0.06	0.14
missing	0.04	0.10
Anxiety symptoms
low	0.06	0.11
moderate	0.16	0.01
high	0.01	0.08
missing	0.12	0.06
Depression symptoms
low	0.03	0.00
moderate	0.12	0.02
high	0.06	0.07
missing	0.12	0.06
Behavior problems
none	0.46	0.02
low	0.05	0.02
moderate	0.01	0.10
high	0.21	0.06
higher	0.14	0.08
missing	0.09	0.08
Concentration problems
none	0.21	0.08
lower	0.16	0.12
low	0.07	0.05
moderate	0.04	0.05
high	0.18	0.10
higher	0.11	0.10
missing	0.09	0.08
Shyness
none	0.11	0.01
low	0.04	0.11
moderate	0.02	0.06
high	0.03	0.05
higher	0.04	0.002
missing	0.09	0.08
Intervention in class
none	0.04	0.08
good behavior game	0.04	0.11
mastery learning	0.01	0.01
Intervention in school
none	0.01	0.10
good behavior game	0.00	0.21
mastery learning	0.01	0.11

^*Standardized Bias = Covariate Effect Size = Difference in mean covariate value divided by the standard deviation of treated group

^**Resulting standardized bias after the chosen average treatment on the treated PS technique (generalized boosted modeling estimation paired with weighing by the odds application)