Certainty in categorical judgment of size


Eric J. Fimbel - Research results and data sets - home page


Citation: Fimbel, E.J. (2009) Certainty in categorical judgment of size. Available at http://visualcategories.googlepages.com/. Last retrieved : (mm/dd/yyyy)

Publication: Fimbel, E.J., Michaud, R., Martin, M. (2009) Certainty in Categorical Judgment of Size. PLoS ONE, 4(7), e6198. doi:10.1371/journal.pone.0006198 download draft (pdf)



Introduction

The structure of the categories of a 5-point scale was documented in absolute judgment of size. This page presents the raw data, processed data and charts of the individual and group categories.

The participants were 20 young healthy volunteers (age: mean=22.3 years, stdev = 1.7, years, 10 males, 1 left-handed). 

The stimuli were disks between 1 mm and 80 mm presented on a computer screen. Categories were rated by means of letters A..E (also called ratings). There were 10 blocks of 32 trials. The size of the disk followed a random uniform distribution in the range 8-260 pixels by steps of 4 pixels (64 values). The task was self-paced.

Participants clicked on the center of the screen, an interstimulus composed of concentric disks with shaded tones of gray appeared. After a delay, a blue disk appeared. Participants clicked on the letter (A..E) that represented most adequately the size of the disk according to their own judgment. Participants were only instructed that A was the smallest category and E was the largest. There was no feedback.

Figure 1. steps of a trial. From left to right: click to start; eraser ;stimulus + scale


Results 


The categories (also called membership functions) were obtained as follows. 1) The trials were grouped by Participant (1-20) x Rating rank (1-5). The membership function of participant P and rating R was the distribution of relative frequency, i.e., number of stimuli of size S rated R by participant / total number of stimuli of size S. 

trials (raw data)  (xls, 600k) 

categories (descriptive statistics)
(xls, 24 kb) 

membership functions (numerical data) 
(xls, 2 mb) 


Below are the individual membership functions, then the average membership functions (averaged across participants)



Figure 2. Individual membership functions. Horizontal: stimulus size. Vertical: membership function (relative frequencies are averaged on bins of 4 consecutive sizes, i.e. about 20 stimuli per bin).



Figure 3. Membership functions averaged across participants. Horizontal: stimulus size. Vertical: membership function (relative frequencies are averaged on bins of 4 consecutive sizes, i.e. about 20 stimuli per bin).


Software

(only if you like exotic languages). For Windows XP, written in Java (JRE 1.4.2 ), comments in French.

visualCategorization2004.zip 


Instructions

unzip anywhere

to execute a test : (demo, training, complete), unzip the file, go to the directory and run corresponding commands (runOne.bat, etc.)

to configure : you can change screen layout, parameters of legend, scale, stimuli etc in configuration files *.ini (they are commented).



Copyright: (c) 2009 E.J. Fimbel. This is open-access content distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. This data contain no sociodemographic or personal information about participants.