The 6-point test evaluates the validity of the estimated difference scale. Given 6 values, a, b, c, a', b', c', on the stimulus scale, if the pair \((a, b) > (a', b')\) and \((b, c) > (b', c')\) then it must be that \((a, c) > (a', c')\), where the symbol \(>\) is taken here to mean “is judged more different than”. Given the observer's difference scale and \(\sigma\) estimate, the likelihood of the choices made is calculated based on the link function indicated in the ‘mlds’ object.
lik6pt(x, Six.Pts, ...)Returns the likelihood of the observer's responses for all of the 6-point conditions from a given data set. As currently implemented, it returns a 1x1 matrix.
an object of class 'mlds', typically created by mlds
a list of 3 data.frames, with names A, B, E. Each data.frame corresponds to a sample from a difference scaling experiment. The corresponding rows of the three data.frames yield the triples of trials that provide a 6-point test. The list can be constructed with the function GetSixPts.
currently unused.
Kenneth Knoblauch, based on C code by Laurence T. Maloney and J. N. Yang.
Maloney, L. T. and Yang, J. N. (2003). Maximum likelihood difference scaling. Journal of Vision, 3(8):5, 573--585, tools:::Rd_expr_doi("10.1167/3.8.5").
Knoblauch, K. and Maloney, L. T. (2008) MLDS: Maximum likelihood difference scaling in R. Journal of Statistical Software, 25:2, 1--26, tools:::Rd_expr_doi("10.18637/jss.v025.i02").
Get6pts, mlds,
simu.6pt
data(kk1)
x.df <- mlds(SwapOrder(kk1))
lik6pt(x.df, Get6pts(x.df, nrep = 1))
Run the code above in your browser using DataLab