This page contains information associated with our publication for InfoVis 2013, Perception of Average Value in Multiclass Scatterplots.
For the paper itself, click here.
For the talk slides, click here.
For information about the experimental stimuli and design (along with sample experiments), click here.
For data tables and other resources, click here.
Other questions and requests can be directed to the paper authors.
Experiments
We conducted a series of experiments for this paper. This website provides details about these experiments not included in the paper for reasons of space or relevancy, as well as information for reproducibility.
A sample version of the experiment run in this paper (with slightly fewer stimuli for the purpose of convenience) is kept running here.
Simplified versions of the experiments (without consent forms, training, and with participant feedback) will be available here, as well as the associated stimuli for the experiment sets.
Experiment Set One
The following is the stimuli set we used in our first round of experiments, in which encoding type was a between-subjects factor. The "short name" is used to refer to a particular encoding level both here and in the data tables.
Each file contains .png images with the following format:
| (R,P,T,F) | +/-(00-40) | -(00-02) |
| The image kind (Regular, Practice, Tutorial, or Flipped regular). | The difference between the first class and the second, in units of 4 pixels | The image index |
Experiment Set Two
The following is the stimuli set we used in our second round of experiments, in which encoding type was a two-level within-subjects factor. The "short name" is used to refer to a particular encoding level both here and in the data tables. Block 1 represents a "baseline" condition.
Each file contains two folders, Block 1 and Block 2, each of which contain .png images with the following format:
| (R,P,T,F) | +/-(00-40) | -(00-02) |
| The image kind (Regular, Practice, Tutorial, or Flipped regular). | The difference between the first class and the second, in units of 4 pixels | The image index |
Data
The raw data tables for our experiment sets are stored as two .csv files. For reasons of privacy, information which could be use to identify participants has been removed. Certain participants did not meet the quality standards required for inclusion in our analysis. These participants have been excised from the data tables.
Each csv contains a number of columns. Not every column is in every table. This table explains those headings:
| id | A pseudo-random anonymous participant idenitifier. |
| index | The question number of this particular trial for the participant. |
| img | What specific image did the participant see in this trial? |
| flipped | Was a point from the first group always the highest absolute point (0), or a point from the second group (1). |
| encoding | What encoding was being tested in this experiment? (See above for list of encoding abbreviations) |
| blockOrder | Which block did this participant see first. |
| block | Which block was the participant on when they answered the question. |
| delta | Difference between the first group and the second, in units of 4 pixels. |
| absDelta | Absolute value of delta. |
| sign | Is the sign of delta positive (1) or negative (0). |
| ans | Which group did the participant guess? (1=the first group or 0= the second group). |
| isRight | Did the participant guess correctly (1=yes or 0=no). |
Data for experiment set one (between-subjects test of encoding): between.csv (236 kB)
Data for experiment set two (within-subjects test of encoding): within.csv (1.5 MB)