| kk {MLDS} | R Documentation |
Three data sets for one subject in a sample difference scaling experiment. The stimuli were scatterplots of bivariate Gaussian samples with different correlations. Either the function runQuadExperiment or runTriadExperiment was used to collect the data.
data(kk1)
data(kk2)
data(kk3)
data(kktriad)
Four data frames, (kk1, kk2, kk3, kktriad) with 330, or in the case of kktriad 165, observations each with the following 5 (4 for kktriad) components.
respS1S2S3S4kktriad does not contain this component as it is from an experiment with triads rather than quadruples
The kk1-3 datasets were generated on three separate days using the function
runQuadExperiment with DisplayOneQuad and DefineMyScale to define the stimuli and display them, anti-respectively. The experiments were run on a Macintosh Pro with a 15 inch screen. The observer was seated about 40 cm from the screen. The kktriad data set was generated from one run using the function runTriadExperiment with DisplayOneTriad to control the display.
Knoblauch, K. and Maloney, L. T. (2008) MLDS: Maximum likelihood difference scaling in R. Journal of Statistical Software, 25:2, 1–26, http://www.jstatsoft.org/v25/i02.
data(kk1) plot(mlds(SwapOrder(kk1))) # Fit and plot difference scale for first data set kk1, # using quadruples of stimuli data(kktriad) plot(mlds(kktriad), type = "b") # Fit and plot experimental data # using triples of stimuli