Hi Ruru's, score.items in package psych is very handy for scoring test items. It has the structure score.items(keys,items). For instance: score.items(c(1,1,1),data.frame(a1=rep(1,5),a2=rep(1,5),a3=rep(1,5)))$scores correctly gives 1 on each case. But if key <> -1,0,1 the following happens score.items(c(0.1,0.1,0.1),data.frame(a1=rep(1,5),a2=rep(1,5),a3=rep(1,5)))$scores gives 10 on each case. It appears that the inverse of key is the relative weight of that item. My questions: is this correct? Is this an undocumented feature? Can i trust on it? Frans [[alternative HTML version deleted]]