BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
QY_Liu
Obsidian | Level 7

The dependent variable is Y(0,1), the independent variables are A(0,1), B(0,1,2) and C is continuous. 

Model 2.

proc logistic data=tmp descending;
class A(ref='0')/param=glm;
model Y=A C A*C/clodds=pl;;
run;

The output:

Capture.PNG

Q2. Why the p_values for Effects in Type 3 Analysis are the same as the p-values for Parameters in Analysis of Maximum Likelihood?

Thank you in advance!

1 ACCEPTED SOLUTION

Accepted Solutions
StatDave
SAS Super FREQ

In the Parameter Estimates table, the Wald chi-square statistic for a parameter is computed as (estimate/std.error)**2. In the Type 3 table, the statistics are computed as shown in the "Details: Testing Linear Hypotheses about the Regression Coefficients" section of the LOGISTIC documentation. When only one parameter (1 df) is involved, they are the same.

View solution in original post

6 REPLIES 6
PaigeMiller
Diamond | Level 26

Q1: Parameter — These are tests of each coefficient in the model. The specific null hypothesis being tested for each coefficient in the model is H0: coefficient=0. So when the p-value is less than some cutoff (usually 0.05 but you can pick other cutoffs) then the conclusion is that we reject the null hypothesis. Type III test of Effect — tests whether the effect(s) due to that term in the model is zero or not.

 

Q2: In the simple case where a parameter has one degree of freedom (or two levels), the Type III and maximum likelihood tests are supposed to be identical.

--
Paige Miller
QY_Liu
Obsidian | Level 7
Thank you, PaigeMiller.
I like to know how the chi-square statistic calculated in the two tables. Could you please suggest some reference?
StatDave
SAS Super FREQ

In the Parameter Estimates table, the Wald chi-square statistic for a parameter is computed as (estimate/std.error)**2. In the Type 3 table, the statistics are computed as shown in the "Details: Testing Linear Hypotheses about the Regression Coefficients" section of the LOGISTIC documentation. When only one parameter (1 df) is involved, they are the same.

QY_Liu
Obsidian | Level 7

The dependent variable is Y(0,1), the independent variables are A(0,1), B(0,1,2) and C is continuous. 

Model 1.

proc logistic data=tmp descending;
class A(ref='0') B(ref='0')/param=glm;
model Y=A B A*B/clodds=pl;
run;

The output:

QY_Liu_0-1607139782482.png

Q1. What's the hypothesis for Effect A, B and A*B in Type 3 Analysis and what's the hypothesis for Parameter A(1), B(1,2) and A*B(1*1, 1*2) in Analysis of Maximum Likelihood?

Thank you in advance!

SAS Innovate 2025: Call for Content

Are you ready for the spotlight? We're accepting content ideas for SAS Innovate 2025 to be held May 6-9 in Orlando, FL. The call is open until September 25. Read more here about why you should contribute and what is in it for you!

Submit your idea!

What is ANOVA?

ANOVA, or Analysis Of Variance, is used to compare the averages or means of two or more populations to better understand how they differ. Watch this tutorial for more.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 6 replies
  • 2899 views
  • 0 likes
  • 3 in conversation