Recommendations

The research team proposes several recommendations for future consideration. The suggestions can be grouped according to interest including practitioners (the LOC), current and potential NDL users, and future experimenters.

Practitioners (LOC)

• The Learning Page could benefit from additional training material.
• Comprehensive examples of query formulations
• Descriptions about the differences between and the purposes of the search and browse features
• Online tutorials or modules
• Descriptions of filter categories (i.e., photographs, recordings, documents)
• Instruction or assistance should be provided to users as they navigate through the system.
• Helpful hints or examples of better queries when searching
• Feedback when the search is too vague or generates an extensive list of hits
• Clear identifiers for pointing and clicking
• Improve the cross reference or integration of the search and browse features.
• Modify the interface to facilitate navigation and satisfaction.
• Minimize need for scrolling
• Utilize descriptive, comprehensible, and specific titles
• Incorporate relevance feedback (particularly if need to preserve titles of documents)
• Color
• Confidence Score
• Position critical features like the Timeline at the top of the page
• Add a site map
• Schedule system changes at regular intervals and add messages like "server down" to manage user expectations about response time.

NDL Users

• Read the helpful suggestions made available on the Learning Page.
• Initially, utilize the browse feature even though search may be preferred to get acquainted with the system and content.
• Take advantage of the Timeline features, particularly for an orientation to the subject matter.
• Expect document titles that may not be relevant to the query search terms.

Future Experimenters

• Additional observations would increase the richness of the results.
• Measure the time it takes the user to construct each query (including user think time)
• Specifically observe users in the training period
• Count the number of keystrokes used while browsing
• Improve the training materials.
• Incorporate active learning or practice exercises instead of or in conjunction with a guided demonstration
• Demonstrate sophisticated search strategies like Boolean logic
• Videotape the experiment to ensure that the observations are consistently recorded.
• Conduct the experiment at a time when the system is normally not inundated with "traffic".
• Utilize subjects who are genuinely interested in history and are motivated to learn.

Continue

Return to the first page