[R] gsDesign Pocock & OBF boundary

[array chip]

Sorry it didn 't work again. I am on yahoo mail, and just found a switch to change from Rick text to Plain text, so here it goes again:


I am learning to use the gsDesign package. 

I have a question about Pocock and OBF boundary. As far as I can understand, these 2 boundaries require equal spacing between interim analyses (maybe this is not correct?). But looks like I can still use gsDesign to run an analysis based on unequal spacing: 

> library(gsDesign)
> gsDesign(k=2,test.type=2,timing=c(0.75,1),alpha=0.05,sfu='Pocock') 


Symmetric two-sided group sequential design with 
90 % power and 5 % Type I Error. 
Spending computations assume trial stops 
if a bound is crossed. 

         Sample 
         Size 
Analysis Ratio*  Z   Nominal p  Spend 
       1  0.796  1.82   0.0346 0.0346 
       2  1.061  1.82   0.0346 0.0154 
   Total                      0.0500 

++ alpha spending: 
Pocock boundary. 
* Sample size ratio compared to fixed design with no interim 
  

Can anyone share some light whether the above analysis is still valid? Or for unequal spacing, I have to use Lan-Demet’s error spending function approximations? Thank you,




________________________________
From: Jeff Newmiller <jdnewmil at dcn.davis.ca.us>

oject.org>; Berend Hasselman <bhh at xs4all.nl> 
Cc: R-help Mailing List <r-help at r-project.org>
Sent: Sunday, September 24, 2017 12:41 AM
Subject: Re: [R] gsDesign Pocock & OBF boundary



Still failed. 

The first secret is in your email program settings, to use Plain Text format (at least for emails you send to this mailing list).

The second secret tool to use is the reprex package to let you verify that your code example will do on our computers what it is doing on your computer before you send it to us. That will also involve giving us some sample data or referencing some data already available to us in a relevant package. See [1], [2] and [3] for more discussion of how to succeed at communicating on the Internet regarding R. 

[1] http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example

[2] http://adv-r.had.co.nz/Reproducibility.html

[3] https://cran.r-project.org/web/packages/reprex/index.html (read the vignette) 
-- 
Sent from my phone. Please excuse my brevity.


On September 23, 2017 9:53:05 PM PDT, array chip via R-help <r-help at r-project.org> wrote:
>Sorry for messed up text. Here it goes again:
>I am learning to use the gsDesign package.
>I have a question about Pocock and OBF boundary. As far as I can
>understand, these 2 boundaries require equal spacing between interim
>analyses (maybe this is not correct?). But looks like I can still use
>gsDesign to run an analysis based on unequal spacing: 
>> gsDesign(k=2,test.type=2,timing=c(0.75,1),alpha=0.05,sfu='Pocock')
>Symmetric two-sided group sequential design with 90 %power and 5 % Type
>I Error.Spending computations assume trial stops if a bound is
>crossed.          Sample          Size  Analysis Ratio*  Z  Nominal p 
>Spend        1  0.796 1.82    0.0346 0.0346        2  1.061 1.82   
>0.0346 0.0154    Total                      0.0500 ++alpha spending:
>Pocock boundary.*Sample size ratio compared to fixed design with no
>interim 
>Can anyone share some light whether the above analysis is still valid?
>Or for unequal spacing, I have to use Lan-Demet’s error spending
>function approximations? Thank you,
>
>
>
>      From: Berend Hasselman <bhh at xs4all.nl>

>Cc: R-help Mailing List <r-help at r-project.org>
> Sent: Friday, September 22, 2017 11:46 PM
> Subject: Re: [R] gsDesign Pocock & OBF boundary
>  
>
>> On 23 Sep 2017, at 01:32, array chip via R-help
><r-help at r-project.org> wrote:
>> 
>> Hi, 
>> 
>> I am learning to use your gsDesign package! I have a question about
>Pocock and OBF boundary. As far as Iunderstand, these 2 boundaries
>require equal spacing between interim analyses(maybe this is not
>correct?). But I can still use gsDesign to run an analysisbased on
>unequal spacing:
>gsDesign(k=2,test.type=2,timing=c(0.75,1),alpha=0.05,sfu='Pocock')Symmetrictwo-sided
>group sequential design with90 %power and 5 % Type I
>Error.Spendingcomputations assume trial stops if a bound is crossed.   
>      Sample          Size  AnalysisRatio*  Z  Nominal p  Spend       
>1  0.796 1.82    0.0346 0.0346        2  1.061 1.82    0.0346 0.0154   
>Total                      0.0500  ++alpha
>spending:Pocockboundary.*Sample size ratio compared to fixed design
>with no interim Can anyone share some light whether the above analysis
>is stillvalid? Or for unequal spacing, I have to use Lan-Demet’s error
>spendingfunction approximations? Thank you,
>>     [[alternative HTML version deleted]]
>> 
>
>Your example code is a complete mess.
>Do NOT post in html. This is a plain text mailing list.
>Read the Posting Guide (https://www.r-project.org/posting-guide.html).
>
>Berend Hasselman]
>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
>
>  
>    [[alternative HTML version deleted]]
>
>______________________________________________
>R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>https://stat.ethz.ch/mailman/listinfo/r-help
>PLEASE do read the posting guide
>http://www.R-project.org/posting-guide.html
>and provide commented, minimal, self-contained, reproducible code.