Multiple Comparison in SAS

[JUMMY]

I have a data of this nature. 

data have;
  	input diet : $1. @;
			do i=1 to 7;
			input  cholesterol @;
		output;
	end;
drop i;
datalines;
A: 29 32 26 15 34 15 23
B: 11 30 34 13 27 24 21
C: 22 12 20 7 3 27 4
D: 16 21 13 6 0 9 25
E: 6 2 10 -3 8 1 -4
;
run;

I did run the Tukey Test using the code below:

proc glm data=have;
  class diet;
  model cholesterol=diet;
  means diet / lines tukey;
run;

And I got this output:

Can someone explain what this output mean? What do the blue and red lines stand for? And which diet is superior compared to the rest?

[PaigeMiller]

---

@JUMMY wrote:

I have a data of this nature. 

data have;
  	input diet : $1. @;
			do i=1 to 7;
			input  cholesterol @;
		output;
	end;
drop i;
datalines;
A: 29 32 26 15 34 15 23
B: 11 30 34 13 27 24 21
C: 22 12 20 7 3 27 4
D: 16 21 13 6 0 9 25
E: 6 2 10 -3 8 1 -4
;
run;

I did run the Tukey Test using the code below:

proc glm data=have;
  class diet;
  model cholesterol=diet;
  means diet / lines tukey;
run;

And I got this output:

Can someone explain what this output mean? What do the blue and red lines stand for?

---

The blue line crosses C D E, so these are not statistically different. The red line crosses A B C D, so these are not statisitically different. (Or to phrase it opposite, E is different than A and B with 95% confidence, but no other difference is statistically significant.)

And which diet is superior compared to the rest?

I don't know. What in the data tells you that one diet is better than another diet?

[JUMMY]

@PaigeMiller , what of a case where there are three lines involved? What does the blue, red and green lines means in the diagram below?

[PaigeMiller]

These lines are interpreted the same way as two lines.