#------------------------------------------------------------------------------
#' hybridRecursive trimming procedure.
#'
#' \code{hybridRecursive} takes a data frame of RT data and returns trimmed rt
#' data. The returned value is the average returned from the nonRecursive
#' and the modifiedRecursive procedures as described in van Selst &
#' Jolicoeur (1994).
#'
#' @param data A data frame with columns containing: participant identification
#' number ('pptVar'); condition identification, if applicable ('condVar');
#' response time data ('rtVar'); and accuracy ('accVar'). The RT can be in
#' seconds (e.g., 0.654) or milliseconds (e.g., 654). Typically, "condition"
#' will consist of strings. Accuracy must be coded as 1 for correct and 0 for
#' error responses.
#' @param minRT The lower criteria for acceptable response time. Must be in
#' the same form as rt column in data frame (e.g., in seconds OR milliseconds).
#' All RTs below this value are removed before proceeding with SD trimming.
#' @param pptVar The quoted name of the column in the data that identifies
#' participants.
#' @param condVar The quoted name of the column in the data that includes the
#' conditions.
#' @param rtVar The quoted name of the column in the data containing reaction
#' times.
#' @param accVar The quoted name of the column in the data containing accuracy,
#' coded as 0 or 1 for incorrect and correct trial, respectively.
#' @param omitErrors If set to TRUE, error trials will be removed before
#' conducting trimming procedure. Final data returned will not be influenced
#' by errors in this case.
#' @param digits How many decimal places to round to after trimming?
#'
#' @references Van Selst, M. & Jolicoeur, P. (1994). A solution to the effect
#' of sample size on outlier elimination. \emph{Quarterly Journal of Experimental
#' Psychology, 47} (A), 631-650.
#'
#' @examples
#' # load the example data that ships with trimr
#' data(exampleData)
#'
#' # perform the trimming, returning mean RT
#' trimmedData <- hybridRecursive(data = exampleData, minRT = 150)
#'
#' @importFrom stats sd
#'
#' @export
hybridRecursive <- function(data,
minRT,
pptVar = "participant",
condVar = "condition",
rtVar = "rt",
accVar = "accuracy",
omitErrors = TRUE,
digits = 3) {
# remove errors if the user has asked for it
if(omitErrors == TRUE){
trimmedData <- data[data[[accVar]] == 1, ]
} else {
trimmedData <- data
}
# get the list of participant numbers
participant <- unique(data[[pptVar]])
# get the list of experimental conditions
conditionList <- unique(data[, condVar])
# trim the data
trimmedData <- trimmedData[trimmedData[[rtVar]] > minRT, ]
# ready the final data set
# make a df here to preserve ppt column
finalData <- as.data.frame(matrix(0, nrow = length(participant),
ncol = length(conditionList)))
# give the columns the condition names
colnames(finalData) <- conditionList
# add the participant column
finalData <- cbind(participant, finalData)
# intialise looping variable for subjects
i <- 1
# loop over all subjects
for(currSub in participant){
# intialise looping variable for conditions. It starts at 2 because the
# first column in the data file containing condition information is the
# second one.
j <- 2
# loop over all conditions
for(currCond in conditionList){
# get the relevant data
tempData <- trimmedData[trimmedData[[pptVar]] == currSub &
trimmedData[[condVar]] == currCond, ]
# get the nonRecursive mean
nonR <- nonRecursiveTrim(tempData[[rtVar]], returnType = "mean")
# get the modifiedRecursive mean
modR <- modifiedRecursiveTrim(tempData[[rtVar]], returnType = "mean")
# find the average, and add to the data frame
finalData[i, j] <- round(mean(c(nonR, modR)), digits = digits)
# update condition loop counter
j <- j + 1
}
# update participant loop counter
i <- i + 1
}
return(finalData)
}
#------------------------------------------------------------------------------
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.