Influence of graph presentation format on user's decision making behavior
Dong-gook Kim
Dalton State College
Dr. Dong-gook “DK” Kim is an associate professor of management at Dalton State College in Dalton, GA. He teaches statistics, quantitative methods, data mining, simulation, optimization and other operations management courses. His research interests include human decision making, data mining, and student learning (pedagogy). He has published articles in Journal of Education for Business, Issues in Information Systems, Journal of Business and Social Science among others. He received a Ph.D. degree in Decision Sciences from Georgia State University in 2008. He became a Certified Analytics Professional (CAP) in spring 2014.
Chongwoo Park
Augusta University
Dr. ChongWoo Park is an associate professor of Management Information Systems at Augusta University. His teaching interests include database design & systems, systems analysis and design, information systems, business intelligence and systems, enterprise process integration, and computer concepts. His research work has been published in journals including Decision Sciences, Journal of the Association for Information Systems, IEEE Transactions on Engineering Management, Journal of Systems and Software, and Issues in Information Systems. His research interests include IT/IS adoption, IT use in education, IT project management, information seeking behavior, cloud computing, analytics, and behavioral side of security. Prior to his academic career, he worked as an IT consultant in South Korea.
Abstract
Graphs provide rich pattern information that are generally more understandable than tabular forms. However, individuals can be often misled by distorted graphs (e.g., Pennington and Tuttle 2009), and there are many forms of... [ view full abstract ]
Graphs provide rich pattern information that are generally more understandable than tabular forms. However, individuals can be often misled by distorted graphs (e.g., Pennington and Tuttle 2009), and there are many forms of distortion. For example, the Y axis of a graph does not begin from 0 or adding depth shading to make 2 dimensional (2D) graphs to look like 3 dimensional (3D) ones. We call the latter form of distortion pseudo 3D (P3D) graphs, which has 3D effect on 2D graphs. That is, the data set for the graph is 2D in nature, and P3D graphs, therefore, do not carry any more information than their 2D counterparts. P3D graphs can affect users’ perception often in a negative way and, therefore, are not generally recommended (e.g., Tufte, 2001). However, the use of such graphs is prevalent in schools, advertisement, business, Web, and so on. This can be attributed partly to the ease of producing such graphs in one of the most popular graphing software—Microsoft Excel (Su, 2008). Past studies found the poor user performance of P3D graphs (e.g., Carswell, Frankenberger & Bernhard, 1991; Tversky & Schiano, 1998). Kelton et al. (2010) suggest multiple representations (e.g., graphs and tables) to mitigate the negative effects of distorted graphs, such as P3D graphs. In this paper, we study the decision making behaviors of subjects when they are given P3D graphs vs. 2D graphs in experiments. In the first experiment, we test if subjects using P3D graphs are to interpret the underlying pattern differently from those using 2D graphs. In the second experiment, subjects can purchase additional information, which is a table of the data used to draw the graph. We test if subjects in the P3D group is more likely to purchase the second piece of the information than those in the 2D group. A subject will go through multiple iterations of the same task with different data in each time. Along with their responses, we measure the speed of their responses in the experiments. In both experiments, we design the task that prefers graphs to tables according to the cognitive fit theory (Vessey, 1991). Students in regional universities in the southeastern US will be participants of the experiments that will be performed in computer labs. An important implication is whether or not the use of pseudo 3D graphs is justified, especially when people may prefer P3D graphs if they try to impress others (Tractinsky & Meyer, 1999).
Authors
-
Dong-gook Kim
(Dalton State College)
-
Chongwoo Park
(Augusta University)
Topic Area
Topics: Analytics, Business Intelligence, Data Mining, & Statistics
Session
AS2 » Multi-Server Queueing Systems/Data Mining/Graph Presentation (16:30 - Thursday, 23rd February, Wraggborough)
Paper
Abstract_submitted_1-26-17.pdf
Presentation Files
The presenter has not uploaded any presentation files.