Linear Diagram Design Study

This page contains questions, data and experimental software for a Linear Diagram Design Study.

The work here is described in this publication:

The consumer version of the software. See below for a version that will allow reproduction of the study diagrams.

There are 7 studies, 6 testing at a specific design feature and the final testing the combined design. They had either 2 or 3 conditions. Each study was conducted in broadly the same way using Amazon MTurk. MTurk workers were restricted to those from the USA. Participants were tested for attention. Data from non-attentive partipants was discarded. Each (attentive) participant was awarded $1 for contributing to the research. A pilot of 10 subjects per condition was conducted, adjustments made depending on the pilot, then the main study of 100 participants per condition was conducted. The data contained in the subsequent study pages is filtered to remove non-attentive participants.

The studies used a between participant design. Participants took one of the tests, and could only take one test. Participants were barred from taking more than one of any of the studies, by recording Worker Ids. Participants who attempted to restart after starting a test but had not finished were not allowed to do so. For each different type the questions are the same, except for some variation in the text of training questions.

The first pages were training, and were always presented in the same order. There were 2 inattentive participant test questions. The text of these requires the participant to click on the image rather than the radio button and "Next page" button. The intention is to ensure the participant reads the questions rather than clicks as quickly as possible through them. They were always on pages 8 and 14. These two pages also afforded a rest point, as the participants were told they were not timed. The 12 data generating questions were presented in a random order.

The visualizations were generated from source data derived from SNAP Twitter ego networks. The sets names were changed, and some reduction in large data sets was performed.

The studies (in chronological order):

  1. line segments
  2. colour
  3. guide-lines
  4. set-order
  5. line width
  6. orientation
  7. final comparison of designs

This is the software used to generate the diagrams. It is set up so the guided options are the default.

The testing code is modified from that of INRIA, Paris, http://www.aviz.fr/bayes. The paper describing their study is:
Micallef, L.; Dragicevic, P.; Fekete, J.Assessing the Effect of Visualizations on Bayesian Reasoning through Crowdsourcing. Visualization and Computer Graphics, IEEE Transactions on , vol.18, no.12, pp.2536,2545, Dec. 2012,