SPSS Statistics

SPSS Statistics

Your hub for statistical analysis, data management, and data documentation. Connect, learn, and share with your peers! 

 View Only
Back to discussions
Expand all | Collapse all

Meta-analysis based on p values

  • 1.  Meta-analysis based on p values

    Posted Mon January 30, 2023 02:41 PM
    I am user of SPSS 29.0 and I have the following question. How can I perform Meta-analysis when the means of the outcome in the active and control group are known,  but the standard deviation of the outcome in the control group is uknown?. The p value of the comparison between the active and control group of each study is known.
    Thank you in advance

    Kind Regards

    George Christou

    ------------------------------
    Georgios Christou
    ------------------------------


  • 2.  RE: Meta-analysis based on p values

    Posted Mon January 30, 2023 02:46 PM
    From a statistician:

     

    "Meta-analysis is all about effect sizes per se. To estimate the effect size of each study, we must know or be able to estimate the standard deviation. Just given means and if we do not have original data available to estimate the standard deviations, we cannot do meta-analysis.

    "However, the customer mentioned that he had p-values, which made the situation interesting. Suppose for each study, the p-value is associated with a corresponding t-test, he may reversely find the T test statistic and then the pooled standard deviation (with the study size also known). With that and the means, the effect size could be estimated for a study. Then, I believe the meta-analysis could be done."



    ------------------------------
    Rick Marcantonio
    Quality Assurance
    IBM
    ------------------------------

    Original Message


Related Content