Help using Base SAS procedures

P value significant but means are not significant

Accepted Solution Solved
Reply
Contributor
Posts: 31
Accepted Solution

P value significant but means are not significant

Hi All,

Please I am working on a 2 year data using the Proc mixed procedure. After the analysis (with no error in my SAS log), the P value associated with F value showed a significant difference but my mean values (after Tukey Kramer's adjustement) are not significant.

 

please how do I explain this.

 

Please find tattached the result (in excel)


Accepted Solutions
Solution
‎10-19-2017 10:39 AM
Esteemed Advisor
Posts: 5,529

Re: P value significant but means are not significant

The way to put this is something like: "Overall F-test shows significant variation between means, but multiple testing cannot identify significantly different pairs."

PG

View solution in original post


All Replies
Respected Advisor
Posts: 3,018

Re: P value significant but means are not significant

Do not use MS Office files to share information, they are a security risk, and most people (including me) will not open them. Use PDF or TXT or GIF

 

The overall p value, associated with the F-test, does not have to agree with Tukey-Kramer. There is a lot of literature on the differences between multiple comparison of means procedures, perhaps you want to read up on this subject.

--
Paige Miller
Contributor
Posts: 31

Re: P value significant but means are not significant

Posted in reply to PaigeMiller

Noted paige miller

Solution
‎10-19-2017 10:39 AM
Esteemed Advisor
Posts: 5,529

Re: P value significant but means are not significant

The way to put this is something like: "Overall F-test shows significant variation between means, but multiple testing cannot identify significantly different pairs."

PG
Contributor
Posts: 31

Re: P value significant but means are not significant

Much appreciated PG. I will report it as you have advised.

☑ This topic is solved.

Need further help from the community? Please ask a new question.

Discussion stats
  • 4 replies
  • 306 views
  • 5 likes
  • 3 in conversation