Use of credibility heuristics in a social question-answering service

Matthews, P. (2015) Use of credibility heuristics in a social question-answering service. Information Research, 20 (1). ISSN 1368-1613 Available from:

[img] HTML
Available under License Creative Commons Attribution Non-commercial No Derivatives 3.0.


Publisher's URL:


Introduction.This study looked at the effect of community peripheral cues (specifically voting score and answerer's reputation) on the user's credibility rating of answers. Method. Students in technology and philosophy were asked to assess the credibility of answers to questions posted on a social question-answering platform. Through the use of a test interface, the availability of cues as to community rating and the reputation of the answer providers were manipulated. Analysis. The main analysis concerned the correlations between the students' credibility rating and the community votes the answers had received. Results. Findings showed that across subjects and knowledge levels, the availability of cues increased the level of agreement between the test subjects and the community aggregate vote. Agreement with answerer reputation was weaker but still significant. Furthermore, at low knowledge levels cue visibility increased confidence regarding credibility judgements. Conclusions. The findings are discussed within a framework of social epistemology, which provides a theoretical foundation for the assessment of knowledge systems. Some suggestions are provided for future work aimed at correcting over reliance on heuristics in credibility decisions.

Item Type:Article
Uncontrolled Keywords:question-answering, qa, sqa, credibility, heuristics, collaborative filtering, knowledge
Faculty/Department:Faculty of Environment and Technology > Department of Computer Science and Creative Technologies
ID Code:25215
Deposited By: Dr P. Matthews
Deposited On:16 Mar 2015 12:21
Last Modified:15 Nov 2016 21:21

Request a change to this item

Total Document Downloads in Past 12 Months

Document Downloads

Total Document Downloads

More statistics for this item...