SAS Output

The MEANS Procedure

Analysis Variable : GPA
N	Mean	Std Dev	Minimum	Maximum
400	2.998925	0.3979409	1.9	4


The FREQ Procedure

Likelihood of Applying (1 = likely)
Lapply	Frequency	Percent	Cumulative	Cumulative
Lapply	Frequency	Percent	Frequency	Percent
0	220	55	220	55
1	180	45	400	100

APPLY	Frequency	Percent	Cumulative	Cumulative
APPLY	Frequency	Percent	Frequency	Percent
0	220	55	220	55
1	140	35	360	90
2	40	10	400	100

Parent Has Graduate Degree
parentGD	Frequency	Percent	Cumulative	Cumulative
parentGD	Frequency	Percent	Frequency	Percent
0	337	84.25	337	84.25
1	63	15.75	400	100

Student Attends Public University
PUBLIC	Frequency	Percent	Cumulative	Cumulative
PUBLIC	Frequency	Percent	Frequency	Percent
0	343	85.75	343	85.75
1	57	14.25	400	100


The GENMOD Procedure

Model Information
Data Set	WORK.GRADPLAN
Distribution	Binomial
Link Function	Logit
Dependent Variable	Lapply	Likelihood of Applying (1 = likely)

Number of Observations Read	400
Number of Observations Used	400
Number of Events	220
Number of Trials	400

Response Profile
Ordered	Lapply	Total
Value	Lapply	Frequency
1	0	220
2	1	180

PROC GENMOD is modeling the probability that Lapply='0'. One way to change this to model the probability that Lapply='1' is to specify the DESCENDING option in the PROC statement.

Parameter Information
Parameter	Effect
Prm1	Intercept

Iteration History For Parameter
Estimates
Iter	Ridge	LogLikelihood	Prm1
0	0	-275.27348	0.2197225
1	0	-275.25553	0.2006514
2	0	-275.25553	0.2006707

Criteria For Assessing Goodness Of Fit
Criterion	DF	Value	Value/DF
Log Likelihood		-275.2555
Full Log Likelihood		-275.2555
AIC (smaller is better)		552.5111
AICC (smaller is better)		552.5211
BIC (smaller is better)		556.5025

Last Evaluation Of The
Negative Of The Gradient
and Hessian
	Prm1
Gradient	-1.84E-09
Prm1	99

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter	DF	Estimate	Standard Error	Wald 95% Confidence Limits		Wald Chi-Square	Pr > ChiSq
Intercept	1	0.2007	0.1005	0.0037	0.3977	3.99	0.0459
Scale	0	1	0	1	1

	The scale parameter was held fixed.



The GENMOD Procedure

Model Information
Data Set	WORK.GRADPLAN
Distribution	Binomial
Link Function	Logit
Dependent Variable	Lapply	Likelihood of Applying (1 = likely)

Number of Observations Read	400
Number of Observations Used	400
Number of Events	180
Number of Trials	400

Response Profile
Ordered	Lapply	Total
Value	Lapply	Frequency
1	1	180
2	0	220

PROC GENMOD is modeling the probability that Lapply='1'.

Parameter Information
Parameter	Effect
Prm1	Intercept

Iteration History For Parameter
Estimates
Iter	Ridge	LogLikelihood	Prm1
0	0	-275.27348	-0.219722
1	0	-275.25553	-0.200651
2	0	-275.25553	-0.200671

Criteria For Assessing Goodness Of Fit
Criterion	DF	Value	Value/DF
Log Likelihood		-275.2555
Full Log Likelihood		-275.2555
AIC (smaller is better)		552.5111
AICC (smaller is better)		552.5211
BIC (smaller is better)		556.5025

Last Evaluation Of The
Negative Of The Gradient
and Hessian
	Prm1
Gradient	1.84E-09
Prm1	99

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter	DF	Estimate	Standard Error	Wald 95% Confidence Limits		Wald Chi-Square	Pr > ChiSq
Intercept	1	-0.2007	0.1005	-0.3977	-0.0037	3.99	0.0459
Scale	0	1	0	1	1

	The scale parameter was held fixed.



The GENMOD Procedure

Model Information
Data Set	WORK.GRADPLAN
Distribution	Binomial
Link Function	Logit
Dependent Variable	Lapply	Likelihood of Applying (1 = likely)

Number of Observations Read	400
Number of Observations Used	400
Number of Events	220
Number of Trials	400

Response Profile
Ordered	Lapply	Total
Value	Lapply	Frequency
1	0	220
2	1	180

PROC GENMOD is modeling the probability that Lapply='0'. One way to change this to model the probability that Lapply='1' is to specify the DESCENDING option in the PROC statement.

Parameter Information
Parameter	Effect
Prm1	Intercept
Prm2	parentGD
Prm3	GPA3
Prm4	PUBLIC

Iteration History For Parameter Estimates
Iter	Ridge	LogLikelihood	Prm1	Prm2	Prm3	Prm4
0	0	-265.00194	0.3650003	-1.113614	-0.567908	0.2070592
1	0	-264.9624	0.3381472	-1.059362	-0.548005	0.2004713
2	0	-264.9624	0.3382338	-1.059612	-0.548246	0.2005571
3	0	-264.9624	0.3382338	-1.059612	-0.548246	0.2005571

Criteria For Assessing Goodness Of Fit
Criterion	DF	Value	Value/DF
Log Likelihood		-264.9624
Full Log Likelihood		-264.9624
AIC (smaller is better)		537.9248
AICC (smaller is better)		538.0261
BIC (smaller is better)		553.8907

Last Evaluation Of The Negative Of The Gradient and Hessian
	Prm1	Prm2	Prm3	Prm4
Gradient	-9.77E-08	1.17E-07	1.56E-07	-1.92E-08
Prm1	93.971936	13.525946	0.4546087	13.199123
Prm2	13.525946	13.525946	1.9430187	2.828192
Prm3	0.4546087	1.9430187	14.4641	2.9910389
Prm4	13.199123	2.828192	2.9910389	13.199123

Algorithm converged.		0.583753215

Analysis Of Maximum Likelihood Parameter Estimates
Parameter	DF	Estimate	Standard Error	Wald 95% Confidence Limits		Wald Chi-Square	Pr > ChiSq
Intercept	1	0.3382	0.1187	0.1056	0.5709	8.12	0.0044
parentGD	1	-1.0596	0.2974	-1.6425	-0.4767	12.7	0.0004
GPA3	1	-0.5482	0.2724	-1.0822	-0.0143	4.05	0.0442
PUBLIC	1	0.2006	0.3053	-0.3979	0.799	0.43	0.5113
Scale	0	1	0	1	1

	The scale parameter was held fixed.
0.583753215		1.40242096	2.885216672		0.34659442	1.40242096
0.327084768		0.48607128	0.34659442

GPA3	Logit	Odds of 0	Prob = 0
1	-0.210	0.811	0.448
0	0.338	1.402	0.584
-1	0.886	2.426	0.708
-2	1.435	4.198	0.808

Public	Logit	Odds of 0	Prob = 0
1	0.539	1.714	0.632
0	0.338	1.402	0.584