Dear friends,
There is Mean and SD of two parameters such as blood glucose and serum Insulin , which have been extracted from about 14 articles in order to a meta analysis.
Now,I have considered to obtain a ratio between these parameters.
- How should I obtain or estimate the Standar deviation of these two parameters which have not been determined in articles?
I would be pleased If some one answer me.
You can estimate the standard deviation of the ratio using propagation of errors. See the link below and.see the "Multiplication or Division" row in Table 1.
http://chemwiki.ucdavis.edu/Core/Analytical_Chemistry/Quantifying_Nature/Significant_Digits/Propagation_of_Error
Check this site as well.it explains in a simple way.
http://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/2-mean-and-standard-deviation
Many people assume they have to impute standard deviations when it is not really necessary. If the data you are extracting consists of a precise P value (eg. 0.0029) and the sample size (which is surely reported), you can then use the TINV function in Excel to calculate the t-statistic. You can then use t and N to calculate a standardized mean difference (SMD) effect size, either Hedges's g or Cohen's d. For details, please see Supplemental Appendix of my 2008 NEJM article on antidepressants (link provided).
http://www.nejm.org/action/showSupplements?doi=10.1056%2FNEJMsa065779
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis