This study aims to investigate the effects of event type (concepts represented by the graph) in graph comprehension with three graph types (line, bar, area) and two graph designs (linear, round) by means of two different task types (trend assessment, discrete comparison). A novel round graph type was designed for that purpose. Five hypotheses were investigated: H1: Graph type affects comparison strategies / H2:
Event type affects comparison strategies / H3: Graph design affects comparison strategies / H4: Graph design and event type interact / H5: Task type affects comparison strategies. As a method to collect data on subjects' / graph perception and comprehension, behavioral (recollected values, word preferences in the description task) and eye-tracking data (scan paths, gaze length, number of fixation, fixation duration and number of transitions) were collected. As an outcome of this thesis, while the event type and the task type seemed to affect the graph comprehension, the effect of graph type, the graph design and interaction between graph design and event type were partially observed. These results point out that although round and linear graph designs are informationally equivalent, the round graphs are computationally better suited than linear graphs for the interpretation of cyclic concepts. However, grasping trend information for the linear events and making discrete comparisons were achieved with the same effort in both graph designs. This result is not trivial at all, given the fact that participants were not familiar with the round graph design and were confronted with them in this experiment for the first time.
| Identifer | oai:union.ndltd.org:METU/oai:etd.lib.metu.edu.tr:http://etd.lib.metu.edu.tr/upload/12611699/index.pdf |
| Date | 01 February 2010 |
| Creators | Alacam, Ozge |
| Contributors | Hohenberger, Annette |
| Publisher | METU |
| Source Sets | Middle East Technical Univ. |
| Language | English |
| Detected Language | English |
| Type | M.S. Thesis |
| Format | text/pdf |
| Rights | To liberate the content for public access |
Page generated in 0.0017 seconds