Habuchi, Kitajima, & Takeuchi: "Comparison of Eye Movements in Searching for Easy-to-Find and Hard-to-Find Information in a Hierarchically Organized Information Structure"

Habuchi, Y., Kitajima, M., & Takeuchi, H. (2008). Comparison of Eye Movements in Searching for Easy-to-Find and Hard-to-Find Information in a Hierarchically Organized Information Structure. In K. Räihä & A. T. Duchowski (Eds.), ETRA 2008 - Proceedings of the Eye Tracking Research and Application Symposium. Presented at the ETRA 2008, Savannah, GA: ACM Press. 131-134.

 


 

Introduction

 

Eye Movements in Search of Hyperlinks to Go to Other Pages

 

F-shaped eye movements for websites. Cutrell et al. and Guan et al. found that adding info to search results significantly improved performance for info tasks but degraded for nav tasks (131).

 

Combining Web Usability Studies with Eye-Tracking Studies

 

argues against equating difficulty with eye-movement data.

 

Blackmon''s Cognitive Walkthrough for the Web (CWW). Uses Latent Semantic Analysis (LSA) semantic space. ID''s user probs "based on an estimate of the strength of information scent that the correct hyperlinks emit for a given search target relative to the strength of information scent that the rest of the hyperlinks, i.e., the wrong hyperlinks, emit" (131). "The purpose of this study is to investigate the relationship between task difficulty and eye movements by independently assessing task difficulty, applying a method developed for web usability studies" (131). Info search tasks. Four difficulties: 1) NP; 2) weak scent correct link problem; 3) unfamiliar correct link problem; 4) competing links nested under competing headings.

 

Types of Website Usability Problems Identified by CWW

 

Weak Scent Correct Link Problem

 

= correct link semantically unrelated to visitor''s goal (serious)

 

Unfamiliar Correct Link Problem

 

target visitors lack sufficient background knowledge to comprehend the link text and estimate its similarity to current goals.

 

Competing Links Nested Under Competing Headings Problem

 

heading semantically similar to visitor''s goal but associated links do not have link that leads to accomplishing visitor''s goal (likely serious)

 

Experiment

 

Task

 

Find object in simulated encyclopedia website.

 

Materials

 

32 items (16/16 Humanities/Sciences). Each selected for a usability problem type.

 

Participants

 

11: 4 males, 7 f; 21-45; normal/corrected to normal vision; regular web users.

 

Apparatus and Procedure

 

Tobii 1750 Eye-tracking system. Dependent variables: time to complete task, total number fixations, total duration of fixation.

 

Results

 

performance divided into two stages: from beginning until 1stclick (initial heading selection stage); rest of task (link selection stage) (133).

 

Initial Heading Selection Stage

 

One way ANOVA for total processing, total # fixations, total duration fixations. No dff among conditions

 

Link Selection Stage

 

1 way ANOVA for processing, total # fixation, duration of fixations

 

Total Processing Time and Total Number of Fixations

 

Significant direction between processing and # of fixations. Participants spent more time on pages with usability probs.

 

Total Duration of the Fixations

 

again, significant difference.

 

Discussion

 

difference not in initial stage but significant in link selection for pages with usability issues.

 

Conclusions

 

Issues worse on pages with weak scent or unfamiliar problem.