Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Computing effect size for a clustered randomized control trial
From
Steve Samuels <[email protected]>
To
[email protected]
Subject
Re: st: Computing effect size for a clustered randomized control trial
Date
Mon, 3 May 2010 21:59:12 -0400
Joe's description of his design is vague. He did not say whether the
unit of randomization will be the classroom or thestudent (with
classrooms as experimental blocks); nor whether the intervention will
be applied to students in groups or individually.
Steve
On Mon, May 3, 2010 at 9:30 PM, Michael Norman Mitchell
<[email protected]> wrote:
> Greetings
>
> I concur with Scott, that Optimal Design is very useful in this kind of
> situation. I would especially recommend taking time with the user manual for
> this program. The program is super well designed and easy to use, and I
> think it is very tempting just to use the program without the documentation.
> However, I feel that is cheating oneself because the manual is exceptional
> in the way that it explains the underlying issues and how to conceptualize
> them in terms of using their software. I think the trickiest bit, in your
> case, will be the combination of the clustering and the repeated measures. I
> think you will get the most mileage out of conceptualizing the "pre test" as
> a level 1 covariate and I believe the optimal design software will permit
> this. However, as the manual describes, it is "level 2" covariates that will
> have the biggest impact on power. If the software does not support level 1
> covariates in such a model, I don't think it will alter the power estimate
> that much.
>
> I hope that is useful additional information.
>
> Michael N. Mitchell
> See the Stata tidbit of the week at...
> http://www.MichaelNormanMitchell.com
>
> On 2010-05-03 6.08 PM, Scott Baldwin wrote:
>>
>> I'm not aware of any built in or user-written method in Stata for
>> computing power in cluster randomized trials. You can try the Optimal
>> Design software
>> (http://sitemaker.umich.edu/group-based/optimal_design_software) or
>> the formulae in David Murray's book "Design and analysis of group
>> randomized trials."
>>
>> Best,
>> Scott
>>
>>
>> On Mon, May 3, 2010 at 5:23 PM, Joe McCrary<[email protected]> wrote:
>>
>>>
>>> I am designing a randomized control trial, where we are going to
>>> select about 12 students per classroom and randomize them into 3
>>> groups, a control and 2 treatment groups. Is there a way I can compute
>>> the number of classrooms needed for the study using the following
>>> criteria:
>>>
>>> Desired power=0.8
>>> Estimated effect size = 0.2
>>> alpha = 0.0167
>>> 1 pre-test measure
>>> 1 post-test measure
>>> approx correlation between the two measures = 0.25
>>>
>>> --
>>> Joe McCrary
>>> Senior Research Associate
>>> WestEd-Atlanta
>>> 260 Peachtree St., Suite 2200
>>> Atlanta, GA 30303
>>>
>>> ph: 541-782-8626
>>> fax: 404-393-3797
>>>
>>> Skype: joe.mccrary
>>> Google talk; joe.mccrary
>>> **********************
>>>
>>> "No problem can be solved from the same level of consciousness that
>>> created it."
>>> --Albert Einstein
>>> *
>>> * For searches and help try:
>>> * http://www.stata.com/help.cgi?search
>>> * http://www.stata.com/support/statalist/faq
>>> * http://www.ats.ucla.edu/stat/stata/
>>>
>>>
>>
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/statalist/faq
>> * http://www.ats.ucla.edu/stat/stata/
>>
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
--
Steven Samuels
[email protected]
18 Cantine's Island
Saugerties NY 12477
USA
Voice: 845-246-0774
Fax: 206-202-4783
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/