R code for a blogpost

Blog 20150821

# Cumulative meta-analysis coronary heart disease
pval = c(0.1495,
         0.0364,
         0.0821,
         0.0301,
         0.0067,
         0.0296,
         0.0678,
         0.0476,
         0.0417,
         0.0154,
         0.0500,
         0.0290,
         0.0392,
         0.0231,
         0.0225,
         0.0297,
         0.0160,
         0.0064,
         0.0329,
         0.0115,
         0.0125,
         0.0159)
alpha = .05

pval.sig <- pval[pval <= alpha]
pval.sig.adj <- pval.sig / alpha

fisher <- -2*sum(log(pval.sig.adj))
df = length(pval.sig.adj) * 2
fisher
df
pchisq(fisher, df = df)

# Cumulative meta-analysis stroke
pval = c(0.1616,
         0.2540,
         0.2053,
         0.0340,
         0.0229,
         0.0314,
         0.0197,
         0.0199,
         0.0077,
         0.0017,
         0.0042,
         0.0012,
         0.0025,
         0.0022)
alpha = .05

pval.sig <- pval[pval <= alpha]
pval.sig.adj <- pval.sig / alpha

fisher <- -2*sum(log(pval.sig.adj))
df = length(pval.sig.adj) * 2
fisher
df
pchisq(fisher, df = df)