Hi all, I hope to do a metaanalysis of literature to establish the prevailing level of a pollutant detected in seafoods obtained from a specified region.
I wish to obtain a mean ± SEM as well as a confidence interval at 95% for the level of prevalence.
I initially wanted to calculate the pooled variances across each study in order to be able to merge the average but I realized that most of the available data did not report enough information to be able to compute the individual variance for each study.
Most of the available literature on my topic of interest reported mean with or without SD , some reported the raw values.
My question is what is the right way to merge the available data in such a way as to accomplish my objective.
Also within each study, the levels of the pollutant of interest is often reported for different types of seafood. am interested in the general prevalence of the pollutant across all the seafood for my region of interest. For each publication, I intend to sum up the mean values of all the types of seafood reported in it, then obtain a single mean value. after I will then merge this mean value with those of other publications inorder to arrive at an overall mean and confidence interval. Am wondering if this is acceptable?
Alternatively, I am hoping to extract the minimum and maximum values reported from each publication, then I will use the mid point value from all the publications to compute the mean ±SEM, as well as the confidence interval.
What is the right way to do this, and what assumptions would be involved?
Thanks