Analysis of regression model using proc gen procedure with cloglog function

[pammers]

I am new to SAS. I am struggling to interpret a model using cloglog function.

Issue 1: when I take the exp of the coefficient when it is negative, it becomes .55. I get that means X has a 55% less change of .....

But then I take the exp of a positive coefficient and I get a ridiculously large positive influence. What am I not getting? (I have only provided stump here. Am I not supposed to take the exp of the beta coefficient?

Estimate

Pr > ChiSq

baseline (ref)	1	0	.		0	.		0	.		0	.
spellnumber	2	1.1271	<.0001	3.0866921	1.1531	0.0007	3.1679985	1.2616	<.0001	3.531066677	1.4663	<.0001	4.333173
spellnumber	3	1.9147	<.0001	6.78490302	1.9457	0.0071	6.9985291	2.0752	<.0001	7.966139528	2.5292	<.0001	12.54347
spellnumber	4	2.6179	<.0001	13.7069088	3.1014	0.0023	22.22905	2.8168	<.0001	16.72325056	3.2615	<.0001	26.08864
spellnumber (ref)	1	0	.		0	.		0	.		0	.
sex	2	0.1481	0.0005	1.15962885	0.4984	0.0315	1.6460854	0.216	<.0001	1.241102379	0.2489	0.0061	1.282614
sex (ref)	1	0	.		0	.		0	.		0	.
education	1	0.5606	<.0001	1.75172322	1.1016	0.0019	3.0089765	0.7236	<.0001	2.061842499	0.7273	<.0001	2.069485
education	2	1.2517	<.0001	3.49628159	2.0141	<.0001	7.4939798	1.5609	<.0001	4.763106112	1.813	<.0001	6.128806
education (ref)	0	0	.		0	.		0	.		0	.
married	1	0.4439	<.0001	1.5587746	1.6139	<.0001	5.0223603	0.4739	<.0001	1.606246354	0.6419	0.0003	1.900088
married	0	0	.		0	.		0	.		0	.
children	1	-0.5877	<.0001	0.5556037	0.0939	0.8309	1.0984499	-0.6037	0.0005	1	-0.2866	0.26	0.750812
children (ref)	0	0	.		0	.		0	.		0	.
agegroup	2	-0.6833	<.0001	0.50494791	-0.821	0.0006	0.4399914	-0.8013	<.0001	0.448745216	-0.7238	<.0001	0.484906
agegroup	3	-1.2909	<.0001	0.27502315	-1.8697	<.0001	0.1541699	-1.4736	<.0001	0.229099242	-1.3621	<.0001	0.256122
agegroup (ref)	1	0	.		0	.		0	.		0	.
healthreason	1	0.2153	<.0001	1.24023391	0.2871	0.1785	1.3325575	0.3079	<.0001	1.360564926	0.3247	0.0002	1.383615
healthreason (ref)	3	0	.		0	.	0	0	.		0	.