BookmarkSubscribeRSS Feed
psyscience
Calcite | Level 5

Hi, I have been using proc surveylogistic to analyze a binary outcome variable in a public dataset which requires sample weights. My syntax runs fine when I incorporate the weights and add my variables to the model. However, I actually wanted to build a sequential (hierarchical) model by which I first added covariates, then added my predictor in a second step to see what it adds to the model (in terms of R2) over and above the covariates. The way my model stands now I just see their effect all together which is not good from my theoretical standpoint.

 

After reading through the website, I see that proc logistic has an option / SEQ for accomplishing this type of model; however, I cannot use proc logistic because of my need for weighting. 

 

Is there a way to conduct a sequential logistic regression using surveylogistic? My only thought is to run separate regressions and just note the change in R2. But as a student, who is very new to statistics and to SAS, I don't know if this is proper to do? Are there other suggestions? Forgive me if the answer is obvious! TYIA.

1 REPLY 1
psyscience
Calcite | Level 5
As a follow up, after doing additional research I now seem to understand that adding my covariates and predictors into the model all at once will indeed "control for" the covariates. So the more specific problem I now have is that I would still like to get a p value and R2 for the separate parts of my model,
if possible. For example, in the past I have been able to use SPSS to get a change in R2 for the second set of variables and a p value associated with it.

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

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
  • 1 reply
  • 1305 views
  • 0 likes
  • 1 in conversation