1 of 1

A Replication of Visual Perception Studies with Tactile Representations of Data for Visually Impaired Users

�

Areen Khalaila Lane Harrison Nam Wook Kim Dylan Cashman

Brandeis University Worcester Polytechnic Institute Boston College Brandeis University

​

Contact Information �Areen Khalaila areenkh@brandeis.edu

Dylan Cashman dylancashman@brandeis.edu

Lane Harrison ltharrison@wpi.edu

Nam Wook Kim nam.wook.kim@bc.edu

​

​

​

​

Motivation

Method

Production of Graphics

Tactile Printing

User Testing

Data Analysis

Results

Main Findings

Future Work

26.74 sec

54 sec

Graph 2: Average Completion Time per Chart Judgement

Tactile Graphics

Heer & Bostock studys

The Heer & Bostock study, conducted via Mechanical Turk, involved sighted users performing similar tasks with visual charts

Participant Engagement in Tactile Graphic Interpretation

11 visually impaired participants were given tactile representations of bar, pie, bubble, and stacked bar charts, and were tasked with determining the percentage difference between comparative elements within the charts

How we quantified the accuracy of tactile graphic interpretation

We employed the mean log error calculation as used in the foundational Cleveland & McGill study

Graph 1: Proportional judgment results. Top: Results from our tactile study. Middle and Bottom: Estimated results from previous studies {Cleveland McGill 1984,Heer Bostock 2010}. Error bars indicate 95% confidence intervals. Detailed results will be made available in tabular form on OSF.

  • Performance on all chart types was not less accurate for visually impaired users than for sighted users (p>0.5).
  • Visually impaired users demonstrated an average completion time per chart of 26.74 seconds, notably quicker than the 54 seconds reported in Heer & Bostock’s MTurk study.

Future research will delve deeper into optimizing tactile graphic designs through comprehensive user engagement, leveraging feedback to refine interaction strategies

References

[1] W. S. Cleveland and R. McGill. Graphical perception: Theory, experimentation, and application to the development of graphical methods. Journal of the American Statistical Association.

[2] J. Heer and M. Bostock. Crowdsourcing graphical perception: Using mechanical turk to assess visualization design. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems.

[3] M. C. McDonnall and Z. Sui. Employment and unemployment rates of people who are blind or visually impaired.

[4] S. Tabrik. Neural mechanisms underlying cross-modal object categorization: Visual and tactile sense.

​

​

​

Graph 3: Average Charts Reviewed per Participant

We replicate the Cleveland and McGill (1984) graphical perception study with tactile graphics on swell-form paper.

​

  • Visually impaired people have higher unemployment and underemployment rates than the general population
  • Data science careers rely on visualization to communicate and explore trends in data
  • Tactile graphics displays and printers promise to make visualizations accessible to the visually impaired
  • It isn’t clear if visualizations designed for the visual perception system are accurately perceived by the tactile perception system