In the realm of experimental behavior analysis, most often in timingrelated research, the Full Width at Half Maximum (FWHM) analysis servesas a metric for quantifying the width of response peaks, particularly inGaussian-like distributions observed across experimental conditions.FWHM provides a tangible measure to assess the precision, variability,and sensitivity of timing behavior.
Within timing experiments, where datasets often exhibit Gaussian-likedistributions with linear components, FWHM analysis becomes a veryuseful analysis tool. By quantifying peak width, researchers can discernnuanced differences between conditions, assert precision and accuracy,validate theoretical models, and enhance the interpretability of timingdata.
The FWHM computes the width of the range of data points in the X axisin which the Y axis values are at least half of the peak value ormaximum height. Thefwhm function takes two parameters:
x a numeric vector of values from a densitydistribution.y a numeric vector of probabilities or raw datapoints.Returning theFull Width at Half Maximum value and the pointin the\(X\) axis at which the maxvalue occurred, as well as\(x1\) and\(x2\) corresponding to the firstintersection with the FWHM to the left and right of the peak.
Lets load a data set from a Peak Procedure session and extract theFWHM and peak values:
data("gauss_example_2",package ="YEAB")# In this case the IF value is 30s.peak_data<- gauss_example_2den<-density(peak_data$Response_Average)fwhm_values<-fwhm(den$x, den$y)fwhm_values## $fwhm## [1] 1.049185## ## $peak## [1] 0.9022719## ## $x1## [1] 0.4246578## ## $x2## [1] 1.473843