Results of the ANOVA indicated no statistically significant differences in the user performance between the three treatments. Thus, the user performance as measured in terms of time to complete the tasks did not support any one particular history mechanism as being faster than the others. Our original hypothesis stated that the subjects would not be able to complete the history-related tasks in a shorter period of time using PadPrints than Netscape or Internet Explorer. Our results supported this hypothesis.

Results from the subjective satisfaction questionnaire showed no statistically significant differences between the three treatments. Our original hypothesis stated that the users would be more satisfied with the history style of PadPrints. Although the results indicated no preference for a particular history style, some of the subjects' comments "Can I take this home with me" and "Wow, it's creating a tree" were suggestive that PadPrints was visually appealing and useful. It should be noted that the zooming feature of PadPrints was disabled throughout the experiment and that several subjects expressed a desire to make the thumbprints larger. It is our belief that enabling the zooming feature on PadPrints would have increased its subjective satisfaction.

4.3 History Style and Short Term Memory 

There were 29 out of 30 subjects who completed the memory recall test. The average percentage for correctness of the web page order was 22.2% for Internet Explorer subjects, 18.5% for PadPrints subjects and 10.5% for Netscape subjects. Differences in these percentages could be associated with the structure used in the history mechanism of Internet Explorer and PadPrints as compared to the structure provided by Netscape. The similarity between Internet Explorer and memory recall was the use of tool-tips to aid in dissemination of information such as URLs and page titles. PadPrints and memory recall were similar because they both used small graphical images to represent the pages. However, it should be noted that the page size was too small and the zooming feature of PadPrints was not active. The subjects depended on the URLs that were displayed in the status bar of the window rather than the graphical presentation in PadPrints. This may have accounted for the low performance of subjects using PadPrints on the memory recall test. A detailed breakdown of the results indicated that both Internet Explorer and PadPrints had 3 subjects, compared to 1 Netscape subject, who correctly answered 4 out of the first 5 web pages. These results suggest that the type of history style may have some affect on the short term memory of cognitive tasks. George Miller's theory states that people can rapidly recognize approximately 7 +- 2 pieces of information. Further research is needed in this area to determine if a relationship exists between web history styles and short term memory. We observed several problems with the memory recall test that we believe contributed to the lowering of the overall performance of our subjects.

The overall experiment went smoothly, but several unanticipated user reactions and circumstances did occur. The majority of our subjects seemed most disturbed by the limitation of the screen space and font size used on the 17" monitor. Most of the subjects verbally expressed concern about the small print on the pages and exhibited behavior to support their discomfort. Many subjects leaned toward the screen display, squinted their eyes and tryed to resize and adjust the multiple windows. For the most part, the environment was stable and quiet, but in several cases the environment was noisy and this distracted the users from concentrating on the tasks. The experiment was driven by two sets of tasks, one contained directions to build the web page history and the other questions to invoke revisitation of these pages. During both sets of tasks, some users expressed difficulty in understanding the semantics and others showed signs of frustration due to the number of tasks. The frustration level of the subjects seemed to increase when the second set, composed of questions, required searching through the history for a page with information to provide the answer. Since multiple windows were displayed on the screen, several subjects fumbled with the necessary coordination of actions. The required sequence of actions to manipulate the instrument program to display the questions was sometimes confusing and unexpected. Several subjects complained and showed physical reactions to the blanking out or covering up of the entire screen between each question. Several subjects actually moved away from the screen as if they were afraid of invoking the wrong action. We believe that more training on the use of the system would have eliminated some of these fears, but it also would have increased the length of the experiment. Following are the related comments collected after the experiment from the subjects:

• "The lettering in the history box was rather small and somewhat difficult to read."
• "The screens were too small and the interface screen kept moving. It was in the way."
• "If you had more screen real estate we could have enlarged the history window."
• "The number of tasks to perform were too many."
• "It tested my patience when I had to perform more than 15 tasks… I mean I got restless."
• "Have to stay too close to the screen to get some answers."
• "Experiment is too long."

 Several problems were noted with the memory recall test. The main issue was the graphical images were too small. The subjects wanted to read the contents of the pages to aid in the order selection, but most were pleased to see the tooltips containing titles. Another complaint was they were unable to see and easily change the previously selected page order. A mechanism to allow them to select numbers rather then check boxes would have improved the interface and allowed the subjects to better view their progress. The general feeling was the subjects felt lost after several page selections. Following are comments from the subjects that were recorded during the test:

• "I can't see anything."
• "What if I'm wrong"
• "I can hardy see anything."
• "God, this is tough."
• "Pictures don't help at all!"
• "This is too stressing."
• "Too difficult too remember."
• "Pictures are too small."
• "Tips were very useful."

Top | Credits | Introduction | Literature Review | Experiment | Results | Discussion | Conclusions | Acknowledgements | References | Appendices

Last modified: May 11, 1998