Author Topic: Q statistic - homogeneity of studies  (Read 25452 times)

0 Members and 1 Guest are viewing this topic.

Offline StaceyRH

  • Newbie
  • *
  • Posts: 3
  • Karma: 0
Re: Q statistic - homogeneity of studies
« Reply #6 on: 25 March, 2008, 10:18:55 AM »
I hate to bother you but I was wondering how this issue is coming along.  I'm getting close to needing to run the meta-analysis for my thesis project.  Thanks!
Stacey

Offline StaceyRH

  • Newbie
  • *
  • Posts: 3
  • Karma: 0
Re: Q statistic - homogeneity of studies
« Reply #5 on: 17 March, 2008, 08:41:36 AM »
Thanks for working on this issue!!  I copied the heterogeneity section from the example meta-analysis for the version that I have (which I think is the most recent).  The output does appear to be "zero" for all of the tests. 

HETEROGENEITY
Mean Cohen's d Weighted By Variance: 0
Q Statistic Based On Cohen's d Using Inverse Variance as Weighting: 0, DF: 0
Q Statistic Significance Level (Using Chi Squared Test): Two Tailed p = 0, One Tailed p = 0
I Squared (Higgins & Thompson, 2002) For Cohen's d Calculations Using Inverse Variance as Weighting: 0% of Total Study Variation Due to Heterogeneity.

Thanks again!
Stacey

Offline Dev

  • Administrator
  • Sr. Member
  • *****
  • Posts: 60
  • Karma: 2
  • Forum Administrator
    • ClinTools
Re: Q statistic - homogeneity of studies
« Reply #4 on: 17 March, 2008, 08:25:29 AM »
Question here....I was trying to use the test for heterogeneity also and tried to use the link below to get it up and running but there seems to be an error.  When I run the test for heterogeneity in the meta-analysis, no values are given.  The "test" is there and appears to run but no output.  Am I doing something wrong here??  Thanks!
Stacey :)


Hi Stacey,

We're getting there - just testing before release. The version you have - does it give a value of "0" (zero)?

Thanks for your patience,

Dev
I am logged on, therefore I am.

Offline StaceyRH

  • Newbie
  • *
  • Posts: 3
  • Karma: 0
Re: Q statistic - homogeneity of studies
« Reply #3 on: 19 February, 2008, 07:33:16 PM »
Question here....I was trying to use the test for heterogeneity also and tried to use the link below to get it up and running but there seems to be an error.  When I run the test for heterogeneity in the meta-analysis, no values are given.  The "test" is there and appears to run but no output.  Am I doing something wrong here??  Thanks!
Stacey :)

Just a quick update - the Beta version now includes two tests for heterogeneity - the Q statistic and I-Squared - when conducting a meta-analysis. This is now available for download at: http://www.clintools.com/download/ClinTools40-Beta.exe the statistics are in the advanced printout ("save Meta-analysis To Notepad" button).  ;D
Best,
Dev



Offline Dev

  • Administrator
  • Sr. Member
  • *****
  • Posts: 60
  • Karma: 2
  • Forum Administrator
    • ClinTools
Re: Q statistic - homogeneity of studies
« Reply #2 on: 09 January, 2007, 04:49:29 AM »
Just a quick update - the Beta version now includes two tests for heterogeneity - the Q statistic and I-Squared - when conducting a meta-analysis. This is now available for download at: http://www.clintools.com/download/ClinTools40-Beta.exe the statistics are in the advanced printout ("save Meta-analysis To Notepad" button).  ;D
Best,
Dev

I am logged on, therefore I am.

Offline Dev

  • Administrator
  • Sr. Member
  • *****
  • Posts: 60
  • Karma: 2
  • Forum Administrator
    • ClinTools
Q statistic - homogeneity of studies
« Reply #1 on: 14 November, 2006, 11:02:28 PM »
We've had some recent requests for a Q statistic to be given for meta-analyses and this is just to let you know that we're onto it and this will be built into the next version.
Grant
I am logged on, therefore I am.