5. Conclusion

5.1 Impact for practitioners

Through our experiments we achieved to prove our hypothesis that mean task time in accessing specific information on the WEB is proportional to the the depth of the tree structure. Furthermore it seems that people don't like WEB structures of big depth. We believe that practitioners (eg. WEB designers) should keep in mind these observations while constructing homepages on the internet. Low depth designs of a taste of variety (eg. Model 4 and 5) seem to be preferred by users.

5.2 Suggestions for future researchers

Possible expansions of this projects might include:

• Investigating the impact of the number of links per page combined with depth and breadth variations.
• Investigating the impact of the complexity of questions on both access time and errors. Eventhough in this experiment care has been taken to pick questions (tasks) that were of relatively the same complexity it will be interesting to observe what happens as the complexity of the tasks is varied.
• Investigate the effect of larger WEB structures with more complex (realistic) look, and determining whether our results also apply to these structures.
• Carry experiment on larger numbers of subjects.
• An expansion of this test might be to run it both on site (as done) and on-line with remote subjects. This would get a bigger and more diverse group of users.

5.3 Refine the theory

Mean total task time in browsing on the WEB is proportional to the depth of the WEB design and people tend to prefer designs of low depth. On the other hand our experiment failed to prove our hypothesis that errors while browsing are proportional to the depth of the design. On the other hand it is proven that depth is not significant in the total task time.

Continue
Return to the Title Page for "Depth vs Breadth in the Arrangement of Web Links"