2-way ANOVA with interaction not giving P-value

[Sil_R]

Hello, I am trying to run proc GLM considering disease severity on 102 samples and 3 environments. It is ok when I am considering only the variables by itself, but SAS does not give me p-values when I try the interaction between them. Why is that? Is it because my model is saturated? How can I have that interaction analysed?

proc glm data=sev;
class line env;
model sev = line env line*env;
run;

Thank you.

[Rick_SAS]

Please show the output.  

Also, please use PROC FREQ to provide the answers for

1. How many levels for LINE?

2. How many joint levels for ENV*LINE?

Since you didn't specify the SOLUTION option, I assume you are asking about p-values in the Type 1 and Type III model ANOVA tables.

As the following example shows, PROC GLM does give p-values for those tables when the data are nondegenerate:

proc glm data=sashelp.bweight plots=none;
class boy MomEdLevel;
model weight = boy MomEdLevel boy*MomEdLevel;
run;

[Sil_R]

Hello,

Thank you for your answer.

Proc Freq gaves 3 levels for line and total of 306 env*line. I am looking for Type III model anova tables.

SAS Output

The SAS System

The GLM Procedure

Class Level Information Class Levels Values line env

102	1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102
3	1 2 3

Number of Observations Read Number of Observations Used

306

305

---

The SAS System

The GLM Procedure

Dependent Variable: sev

Source DF Sum of Squares Mean Square F Value Pr > F Model Error Corrected Total

304	67164.16118	220.93474	.	.
0	0.00000	.
304	67164.16118

R-Square Coeff Var Root MSE sev Mean

1.000000

.

.

34.64164

Source DF Type I SS Mean Square F Value Pr > F line env line*env

101	30175.75618	298.76986	.	.
2	5588.61953	2794.30976	.	.
201	31399.78547	156.21784	.	.

Source DF Type III SS Mean Square F Value Pr > F line env line*env

101	30089.07898	297.91167	.	.
2	5588.61953	2794.30976	.	.
201	31399.78547	156.21784	.	.

[Rick_SAS · Accepted solution]

You only have 305 valid observations but ENV has 102 unique levels. Thus the model with the interaction term contains more parameters than there are observations.

Some options are 

1. Get more data

2. Do not model interactions

3. Consider whether you can model ENV as a continuous variable

[Sil_R]

Thank you very much.