# Fixed effect model
for (i in studies) {
	delta[i, 1] <- 0
	for (k in 2:na[i]) {
		delta[i, k] <- d[t[i, 1], t[i, k]]
	}
}
