Table 5.

Marginal Effects and Coefficients from Regression

	Adjusted rate / mean			Early retention (1–3)	Late retention (4–8)
Outcomes	Early retention (1–3)	Late retention (4–8)	Promotion	Marginal effects (p-value)	Marginal effects (p-value)
College Attendance (N=1,343)	(n=233)	(n=91)	(n=1,019)
2- or 4-year college attendance, %	29.2	7.7	40.6	−11.4% (.006)	−32.9% (.000)
2-year college attendance, %	18.6	7.0	29.9	−11.3% (.004)	−22.9% (.000)
4-year college attendance, %	11.7	1.0	17.4	−5.7% (.126)	−16.4% (.001)
Public Assistance Programs (N=1,288)	(n=233)	(n=89)	(n=966)
TANF, %	24.7	17.8	18.8	5.9% (.107)	−1.0% (.839)
Food Stamps, %	60.2	54.0	51.6	8.6% (.052)	2.4% (.693)
Medicaid, %	48.7	45.7	44.8	3.9% (.392)	0.9% (.881)
Any public assistance, %	67.0	66.5	60.3	6.7% (.119)	6.2% (.282)

Note. Reference group is the promotion group. The marginal effects are shown in percentage change for dichotomous outcomes. Results are adjusted for covariates: sociodemographic factors, CPC program participation, and early scholastic abilities, as listed in the measures section. Participants retained in both early and late were excluded from the analyses: 24 for college measures and 22 for public assistance measures. Sample sizes for early, late retention, and promotion groups are 215, 91, and 845 for 2-year college attendance, and 195, 86, and 754 for 4-year college attendance.