Dear all, I am following this individual group meta-analysis example in "matadat", https://wviechtb.github.io/metadat/reference/dat.pritz1997.html Here when the raw proportion is used as the outcome measure, why is the upper bound of the prediction interval exceeding 1? which means more than a 100% success?? if yes! it does not make sense to me. Could someone please clarify how should I interpret the prediction interval in this example? Thank you, Tharaka ### random-effects model with raw proportionsdat <- escalc <https://wviechtb.github.io/metafor/reference/escalc.html>(measure="PR", xi=xi, ni=ni, data=dat)res <- rma <https://wviechtb.github.io/metafor/reference/rma.uni.html>(yi, vi, data=dat)predict <https://wviechtb.github.io/metafor/reference/predict.rma.html>(res)#> #> pred se ci.lb ci.ub pi.lb pi.ub #> 0.7968 0.0423 0.7138 0.8797 0.5306 1.0629
[R-meta] Use of raw proportion as the outcome measure for individual group meta analysis
2 messages · Tharaka S. Priyadarshana, Wolfgang Viechtbauer
Dear Tharaka, When meta-analyzing raw proportions, this can indeed happen. It simply means that the upper bound of the prediction interval is 1. When working with the logit transformed proportions, this cannot happen (after the back-transformation). Best, Wolfgang
-----Original Message----- From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces at r-project.org] On Behalf Of Tharaka S. Priyadarshana Sent: Friday, 08 July, 2022 13:29 To: r-sig-meta-analysis at r-project.org Subject: [R-meta] Use of raw proportion as the outcome measure for individual group meta analysis Dear all, I am following this individual group meta-analysis example in "matadat", https://wviechtb.github.io/metadat/reference/dat.pritz1997.html Here when the raw proportion is used as the outcome measure, why is the upper bound of the prediction interval exceeding 1? which means more than a 100% success?? if yes! it does not make sense to me. Could someone please clarify how should I interpret the prediction interval in this example? Thank you, Tharaka ### random-effects model with raw proportionsdat <- escalc <https://wviechtb.github.io/metafor/reference/escalc.html>(measure="PR", xi=xi, ni=ni, data=dat)res <- rma <https://wviechtb.github.io/metafor/reference/rma.uni.html>(yi, vi, data=dat)predict <https://wviechtb.github.io/metafor/reference/predict.rma.html>(res)#> #> pred se ci.lb ci.ub pi.lb pi.ub #> 0.7968 0.0423 0.7138 0.8797 0.5306 1.0629