get.dprime: get dprime
In matsukik/mrsat: Multiple Response Speed Accuracy Tradeoff

Description Usage Arguments Value References Examples

A function that calculates dprimes for each timing bin

get.dprime(x, signal.list, noise.list, is.index = FALSE,
  subset = (x$SeqFlag == 0 & x$Experiment == 1), 
  binmax = 16, 
  dprime.correction=c("extreme", "loglinear", "none"),
  negative.dprime = FALSE,
  bintail = 0)

`x`	data frame usually obtained as an output from `get.bins`
`signal.list`	a list specifying for each condition, which "Condition" in `x` correspond to signal.
`noise.list`	a list specifying for each condition, which "Condition" in `x` correspond to noise.
`is.index`	a logical specifying whether the values in `signal.list` and `noise.list` should be treated as numeric indices of `x$Condition`. Default to `FALSE`.
`subset`	an optional vector specifying a subset of data to be used in the calculating dprimes.
`binmax`	an integer specifying maximum number of bins to be retained. All the data exceeding the limit will be discarded.
`dprime.correction`	the method to be used for calculating the dprime when hit rate or false alarm rate is zero or one. If `dprime.correction="extreme"` which is the default, zeros are replaced with `0.5/n` and ones are replaced with `(n-0.5)/n`, where n is the number of signal or noise trials. If `dprime.correction="loglinear"`, hit and false alarm rates are calculated by adding .5 to the numerator (the number of hit/false alarm trials) and 1 to the denominator (the number of signal/noise trials). If `dprime.correction="none"`, no adjustment will be applied.
`negative.dprime`	logical. If `FALSE` (default), the values of zero is used in place of negative dprime values.
`bintail`	an optional integer specifying the number of bins to be concatenated at the tail.

An object of class mrsat.data, which is a data frame containing following columns:

`bin`	timing bins
`lags`	means of time points at which the responses occured within each timing bin
`dprimes`	calculated as `dprime = qnorm(hit) - qnomr(fa)`
`condition`	values based on experimental conditions defined in the argument `scalelist` to which the dprime values belong to.

The output of get.dprime contains following additional columns:

`hit`	hit rate (correct response to correct stimuli) in each bin
`hit.correction`	a character indicating the type of correction performed when calculating hit rate.
`hit.denom`	total number of response to correct stimuli in each bin
`fa`	false alarm rate (correct response to incorrect stimuli) in each bin
`fa.correction`	a character indicating the type of correction performed when calculating false alarm rate.
`fa.denom`	total number of responses to incorrect stimuli in each bin

Stanislaw, H., & Todorov, N. (1999). Calculation of signal detection theory measures. Behavior Research Methods, Instruments, & Computers, 31, 137–149.

#need to define the scale.list
exp1.signal <- list(noint = c(1,3), 
                obrel = c(5,8), 
                obrelsub = c(11, 14))

exp1.noise <- list(noint = c(2,4), 
               obrel = c(6, 7, 9, 10), 
               obrelsub = c(12, 13, 15, 16))

#then load data and tag bins

data(Auditory_demo)
s01.bins <- get.bins(Auditory_demo, auditory=TRUE)

# finally, obtain dprime
s01.dp <- get.dprime(s01.bins$bins, 
  signal.list = exp1.signal, noise.list = exp1.noise, 
  is.index = TRUE, binmax=14)


plot(s01.dp)