topic Re: Question about Stability Analysis (ANCOVA) and the pooled MSE in Discussions
https://community.jmp.com/t5/Discussions/Question-about-Stability-Analysis-ANCOVA-and-the-pooled-MSE/m-p/51202#M29066
<P>Use of the pooled MSE is not necessarily not appropriate in the case of different slopes and different intercepts, it is just not the way it was implemented in the stability platform in JMP. As was previously stated, "Our development team chose to match Chow’s codes. I also believe this method was implemented to maintain consistency with results from a SAS macro called STAB that was written by FDA researchers."</P>
<P> </P>
<P>However, if the analyst would prefer to analyze this data with a pooled MSE, this is the approach that the Fit Model platform, Standard Least Squares personality uses. You can see an example of an ANCOVA using the Fit Model platform in the JMP Fitting Linear Models book: <A href="http://www.jmp.com/support/help/13-2/Analysis_of_Covariance_with_Unequal_Slopes_Examp.shtml#535610" target="_self">Analysis of Covariance with Unequal Slopes Example</A> .</P>Tue, 13 Feb 2018 20:49:00 GMTsusan_walsh12018-02-13T20:49:00ZQuestion about Stability Analysis (ANCOVA) and the pooled MSE
https://community.jmp.com/t5/Discussions/Question-about-Stability-Analysis-ANCOVA-and-the-pooled-MSE/m-p/28695#M19131
<P><FONT face="arial,helvetica,sans-serif">Good day: I have a question about the Stability Analysis platform in JMP.</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif"> For those of you who are not familiar with Stability Analysis (in the <EM>Analyze/Reliability and </EM></FONT></P><P><FONT face="arial,helvetica,sans-serif"><EM> Survival/Degradation</EM> menu) in the Pharmaceutical Industry, in its basic form it is ANCOVA with one </FONT></P><P><FONT face="arial,helvetica,sans-serif"> categorical factor (batch) and one covariate (time). From a model selection perspective, and starting with the </FONT></P><P><FONT face="arial,helvetica,sans-serif"> full model, the idea is to test the interaction term (batch*time)…if the p-value is <0.25, then the final model </FONT></P><P><FONT face="arial,helvetica,sans-serif"> used to estimate shelf life is “Model 1” (includes the terms batch, time, and batch*time) as described below. </FONT></P><P><FONT face="arial,helvetica,sans-serif"> Alpha is set at 0.25 (a regulatory requirement) for batch related terms to increase power to detect differences </FONT></P><P><FONT face="arial,helvetica,sans-serif"> between batches.</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">In “help” I see the following.</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">“When Model 1 (different slopes and different intercepts) is used for estimating the expiration date, the MSE (mean squared error) is not pooled across batches. Prediction intervals are computed for each batch using individual mean squared errors, and the interval that crosses the specification limit first is used to estimate the expiration date.” My question is this (I posed this question to JMP as well); for Model 1 (as defined above) why not use the pooled MSE?</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">JMP's response was: "Our development team chose to match Chow’s codes. I also believe this method was implemented to maintain consistency with results from a SAS macro called STAB that was written by FDA researchers."</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">In JMP help, they reference: Chow, Shein-Chung. (2007), Statistical Design and Analysis of Stability Studies.</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">I do not have a copy of the referenced text, and the FDA macros mentioned are decades old now.</FONT></P><P> </P><P><FONT face="arial,helvetica,sans-serif">Assuming homogeneity of variance, can anyone enlighten me on why the pooled MSE is not appropriate in this case (but it is in other two models considered: Common Intercept/Common Slope, Different Intercepts/Common Slope)?</FONT></P><P> </P><P><FONT color="#000000" face="arial,helvetica,sans-serif">I'm using JMP v13 (and it was this way in previous versions as well).</FONT></P><P><FONT color="#000000" face="arial,helvetica,sans-serif">Kind Regards</FONT></P><P> </P>Thu, 03 Nov 2016 09:47:21 GMThttps://community.jmp.com/t5/Discussions/Question-about-Stability-Analysis-ANCOVA-and-the-pooled-MSE/m-p/28695#M19131MRB38552016-11-03T09:47:21ZRe: Question about Stability Analysis (ANCOVA) and the pooled MSE
https://community.jmp.com/t5/Discussions/Question-about-Stability-Analysis-ANCOVA-and-the-pooled-MSE/m-p/51202#M29066
<P>Use of the pooled MSE is not necessarily not appropriate in the case of different slopes and different intercepts, it is just not the way it was implemented in the stability platform in JMP. As was previously stated, "Our development team chose to match Chow’s codes. I also believe this method was implemented to maintain consistency with results from a SAS macro called STAB that was written by FDA researchers."</P>
<P> </P>
<P>However, if the analyst would prefer to analyze this data with a pooled MSE, this is the approach that the Fit Model platform, Standard Least Squares personality uses. You can see an example of an ANCOVA using the Fit Model platform in the JMP Fitting Linear Models book: <A href="http://www.jmp.com/support/help/13-2/Analysis_of_Covariance_with_Unequal_Slopes_Examp.shtml#535610" target="_self">Analysis of Covariance with Unequal Slopes Example</A> .</P>Tue, 13 Feb 2018 20:49:00 GMThttps://community.jmp.com/t5/Discussions/Question-about-Stability-Analysis-ANCOVA-and-the-pooled-MSE/m-p/51202#M29066susan_walsh12018-02-13T20:49:00Z