The third collection path: respondents rate every alternative on every criterion with ordinary matrix items, one item per criterion with the alternatives as its rows, and the decision matrix is the per-cell aggregate. No new item type is needed. Per-cell counts and standard deviations are kept so the report can show how firm each cell is.
Usage
sframe_rated_matrix(data, instrument, items, statistic = c("mean", "median"))Arguments
- data
A data frame of responses.
- instrument
An
sframeinstrument declaring every id initems.- items
Character vector of
"matrix"item ids, one per criterion, in the intended criterion order. Every item must declare the samematrix_items(the alternatives) in the same order.- statistic
"mean"or"median".
Value
A list with matrix (alternatives x criteria, with dimnames), n,
sd, alternatives, criteria, and statistic.