predict.mlds {MLDS}

Predict method for MLDS Fits

Description

Predict values based on difference scale fit.

Usage

## S3 method for class 'mlds'
predict(object, newdata = NULL, type = "link", ...)

## S3 method for class 'mlbs'
predict(object, newdata = NULL, type = "link", ...)

Arguments

object

object of class ‘mlds’ or ‘mlbs’, typically from the output of mlds

newdata

A data.frame or object of class ‘mlbs.df’ with 4 columns or ‘mlds.df’ with 5 columns, corresponding to the response at each trial and the ranks of, respectively, the triple or quadruple of stimuli presented.

type

character indicating scale on which predictions should be made, “link”, the default, on the scale of the linear predictor or “response”, on the response scale.

...

When obj$method = “glm”, these are additional arguments passed to predict.glm.

Details

The newdata argument is needed principally for the 6-point test (see lik6pt), to extract the estimated probabilities for the subsets of the original data that form valid 6-point tests.

Value

A numeric vector of the predicted values either on the scale of the linear predictor or on the response scale.

Author(s)

Kenneth Knoblauch

References

Maloney, L. T. and Yang, J. N. (2003). Maximum likelihood difference scaling. Journal of Vision, 3(8):5, 573–585, doi:10.1167/3.8.5.

See Also

mlds, lik6pt

Examples

data(kk1)
kk1.mlds <- mlds(SwapOrder(kk1))
predict(kk1.mlds, type = "response")