This visualization displays four variables (or dimensions) with a two dimensional plot. The scatterplot displays reaction time (RT) differences between cochlear implant users and normal hearing participants in a lexical decision task. During lexical decision participants judge whether a sound stimulus they hear is a word or not (like `trud' in English).
The four dimensions are differences in RT (1) between normal hearing participants and participants with a cochlear implant displayed as a function of neighborhood density (2) and word frequency (3). The size of the dot reflects the size of the difference in RT. The colors of the dots distinguish words (blue) and non-words (orange)(4). The color range can be extended to another dimension as well. Some of the reaction times differences were negative, meaning that RT of cochlear implant user were shorter than the RT of normal hearing participants. I added two buttons to switch from a representation with two colors (words vs. non-words) to a representation with four colors (words vs. non-words for positive and negative differences respectively). On the right side of the plot a `size' legend helps recognizing the size and range of the differences. Moreover, the plot is interactive. Hovering the mouse on a dot displays the name of the (non-)word, the size of the difference in milliseconds (ms), the neighborhood density and frequency (log-scaled).
This post describes in detail the creation of the scatterplot.
Click on one of the text labels below to toggle between a display with two or four colors. The four colors display adds an additional dimension by which positive vs. negative differences allow to identify for which item one group was faster or slower than the other.
The graph displayed on this page shows the difference between words and non-words on a smaller scale displaying them for each pair of word and non-word but also for each single participant.