Start Date
10-12-2017 12:00 AM
Description
The Information Systems discipline, as well as countless others, rely on surveys to explore hypotheses and answer research questions. All research methods have strengths and weaknesses. A significant threat to the validity of surveys are response biases – i.e., the tendency of people to respond to questions on some basis other than the question content. We propose that by monitoring how a respondent answers an online survey—through monitoring Human-Computer Interaction movement indicators, specifically mouse-cursor movements—can provide novel information that can be used to statistically detect, understand, and control for many different types of response biases. We explore this proposition in an experiment that examined the relationship between intentions and one’s actual behavior overtime. By moderating the relationship between intentions and behavior using respondents’ mouse-cursor movements, we approximately doubled the r-squared of our conceptual model. The results suggest that some response biases influence mouse-cursor movements when completing an online survey.
Recommended Citation
Jenkins, Jeffrey L.; Valacich, Joseph S.; and Williams, Parker, "Human-Computer Interaction Movement Indicators of Response Biases in Online Surveys" (2017). ICIS 2017 Proceedings. 12.
https://aisel.aisnet.org/icis2017/ResearchMethods/Presentations/12
Human-Computer Interaction Movement Indicators of Response Biases in Online Surveys
The Information Systems discipline, as well as countless others, rely on surveys to explore hypotheses and answer research questions. All research methods have strengths and weaknesses. A significant threat to the validity of surveys are response biases – i.e., the tendency of people to respond to questions on some basis other than the question content. We propose that by monitoring how a respondent answers an online survey—through monitoring Human-Computer Interaction movement indicators, specifically mouse-cursor movements—can provide novel information that can be used to statistically detect, understand, and control for many different types of response biases. We explore this proposition in an experiment that examined the relationship between intentions and one’s actual behavior overtime. By moderating the relationship between intentions and behavior using respondents’ mouse-cursor movements, we approximately doubled the r-squared of our conceptual model. The results suggest that some response biases influence mouse-cursor movements when completing an online survey.