psychophy: Psychophy processes psychophysic raw data to fitted value

In Cogitos/pphy: Fit psychophysic data, extract slopes and PSS and draw plots by subjects and by average.

Description

This function processes psychophysic raw data to compute descriptive data by condition and by subjects; fit the data by subject; and extract the slopes and PSS index.

Usage

psychophy(data, wid = "subject_nr", stim = NULL, resp = "correct",
  vars = NULL, axnames = NULL)

Arguments

data	A data frame in long format (one row per record).
wid	A string indicating the column name of the subjects or observations. Defaults to "subject_nr".
stim	A string indicating the column name of the physical stimulation (e.g. luminance, sound intensity...).
resp	A string indicating the column name of the response as 0 and 1. The function will remove all values below 0 and above 1. Defaults to "correct".
vars	A vector of string indicating the column name of the dependent variables.
axnames	A vector of length 2 with the strings to use as y label and x label in the graphs. The function automatically add "Proportion of " for the y label. Defaults to NULL.

Details

The fitting is done using the fitPPCurve function of the present package which use the 'modelfree' package to fit the data locally (see http://personalpages.manchester.ac.uk/staff/d.h.foster/software-modelfree/latest/index.html)

Value

Return a list with : Means_per_subjects: a data frame of means by condition and by subjects, a data frame Descript_data: a data frame of descriptive data by condition Fit: a data frame of fitted values (slopes and PSS) Graphs: a list of two graphics, by subjects (BySubj) and global.

Author(s)

Guillaume T. Vallet [e-mail]

Maintainer: Guillaume T. Vallet [e-mail]

See Also

See vignette("locglmfit", package = "modelfree") for more details on the fit function used.

Examples

### Generate fake date to use the function ---------------------------------
# Levels of the physical stimulation
data = rbind(
   data.frame(subject=1, stim = 1:10, cond='c1',
       corr = c(0, 0, 3, 5, 10, 14, 16, 18, 19, 20),
       trials = c(20, 20, 19, 20, 18, 19, 20, 20, 19, 20)),
   data.frame(subject=2, stim = 1:10, cond='c1',
       corr = c(1, 0, 2, 6, 11, 15, 15, 19, 20, 20),
       trials = c(19, 20, 20, 20, 19, 20, 18, 20, 19, 20)),
   data.frame(subject=3, stim = 1:10, cond='c1',
       corr = c(0, 2, 3, 6, 10, 16, 15, 18, 19, 20),
       trials = c(20, 20, 19, 20, 20, 20, 18, 20, 18, 20)),
   data.frame(subject=1, stim = 1:10, cond='c2',
       corr = c(0, 1, 4, 6, 11, 15, 17, 18, 19, 20),
       trials = c(20, 20, 19, 20, 18, 19, 20, 20, 19, 20)),
   data.frame(subject=2, stim = 1:10, cond='c2',
       corr = c(0, 0, 3, 8, 13, 16, 17, 19, 20, 20),
       trials = c(19, 20, 20, 20, 19, 20, 18, 20, 19, 20)),
   data.frame(subject=3, stim = 1:10, cond='c2',
       corr = c(0, 1, 3, 7, 12, 14, 16, 17, 19, 20),
       trials = c(20, 20, 19, 20, 20, 20, 18, 20, 18, 20))
   )

### Fitting the curve with the modelfree adapation -------------------------
fitted = psychophy(data, wid='subject', stim='stim', resp='corr',
               vars='cond')
fitted

Cogitos/pphy documentation built on May 6, 2019, 12:21 p.m.