2 x 2 factorial with a control

[AgReseach7]

Design: 2 x 2 factorial plus a control.

Fed all hogs a basal diet and then fed 1 of 4 treatment supplements. For the control group, we just fed additional basal diet vs. any form of "supplement."

factor 1: type of feed ingredient (ROUGH), which is either cottonseed hulls or juniper

factor 2: percentage of that ingredient in the diet (CONC), which is 20 or 40% of diet

I am interested of course, in the interaction, but am also interested in seeing if any of the 4 trt groups are different than the control group.

Questions

do I analyze the factorial 1st, (using "ROUGH" and "CONC" and ROUGH*CONC in my model) and then re-analyze to see if there are any differences between trt & CNTL?

Or, do I just do an "all-pairwise comparison" and use "TRT" in my model statement (not "ROUGH and CONC")? Or various contrast statements?

OR?

Ex of data

ID			TRT	ROUGH	CONC	DATE	DAY	Period	BWkg
5			20JUN	JUN	20	6-1	19	1	19.5
8			20JUN	JUN	20	6-1	19	1	12.7
9			20JUN	JUN	20	6-1	19	1	25.0
16			20JUN	JUN	20	6-1	19	1	29.9
30			20JUN	JUN	20	6-1	19	1	17.2
5			20JUN	JUN	20	6-10	28	2	22.2
8			20JUN	JUN	20	6-10	28	2	14.1
9			20JUN	JUN	20	6-10	28	2	25.9
16			20JUN	JUN	20	6-10	28	2	33.6
30			20JUN	JUN	20	6-10	28	2	19.1
5			20JUN	JUN	20	6-17	35	3	26.8
8			20JUN	JUN	20	6-17	35	3	16.8
9			20JUN	JUN	20	6-17	35	3	29.5
16			20JUN	JUN	20	6-17	35	3	37.7
30			20JUN	JUN	20	6-17	35	3	21.8
6			20CSH	CSH	20	6-1	19	1	12.7
13			20CSH	CSH	20	6-1	19	1	40.8
15			20CSH	CSH	20	6-1	19	1	34.9
28			20CSH	CSH	20	6-1	19	1	17.7
34			20CSH	CSH	20	6-1	19	1	15.4
6			20CSH	CSH	20	6-10	28	2	14.5
13			20CSH	CSH	20	6-10	28	2	43.1
15			20CSH	CSH	20	6-10	28	2	40.8
28			20CSH	CSH	20	6-10	28	2	20.4
34			20CSH	CSH	20	6-10	28	2	18.6
6			20CSH	CSH	20	6-17	35	3	15.0
13			20CSH	CSH	20	6-17	35	3	46.7
15			20CSH	CSH	20	6-17	35	3	42.2
28			20CSH	CSH	20	6-17	35	3	22.7
34			20CSH	CSH	20	6-17	35	3	20.9