Detection of Confirmation and Distinction Biases in Visual Analytics Systems

Loading...
Thumbnail Image
Date
2019
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
Cognitive bias is a systematic error that introduces drifts and distortions in the human judgment in terms of visual decomposition in the direction of the dominant instance. It has a significant role in decision-making process by means of evaluation of data visualizations. This paper elaborates on the experimental depiction of two cognitive bias types, namely Distinction Bias and Confirmation Bias, through the examination of cognate visual experimentations. The main goal of this implementation is to indicate the existence of cognitive bias in visual analytics systems through the adjustment of data visualization and crowdsourcing in terms of confirmation and distinction biases. Two distinct surveys that include biased and unbiased data visualizations which are related to a given data set were established in order to detect and measure the level of existence of introduced bias types. Practice of crowdsourcing which is provided by Amazon Mechanical Turk have been used for experimentation purposes through prepared surveys. Results statistically indicate that both distinction and confirmation biases has substantial effect and prominent significance on decision-making process.
Description

        
@inproceedings{
10.2312:trvis.20191185
, booktitle = {
EuroVis Workshop on Trustworthy Visualization (TrustVis)
}, editor = {
Kosara, Robert and Lawonn, Kai and Linsen, Lars and Smit, Noeska
}, title = {{
Detection of Confirmation and Distinction Biases in Visual Analytics Systems
}}, author = {
Nalcaci, Atilla Alpay
and
Girgin, Dilara
and
Balki, Semih
and
Talay, Fatih
and
Boz, Hasan Alp
and
Balcisoy, Selim
}, year = {
2019
}, publisher = {
The Eurographics Association
}, ISBN = {
978-3-03868-091-8
}, DOI = {
10.2312/trvis.20191185
} }
Citation
Collections