Quick Search: Usability Study, Round 1 (2005)

We conducted an initial usability test evaluation of the NCSU Libraries Quick Search tool.

Insights

Expectations

When confronted with a single search box to search the library, the participants had different expectations about what would be searched.  Some felt that a library search is exclusively about searching for books.  Others expanded this to include articles and journals.  When confronted with a search tool like this that includes FAQs and information about the library, most of the users were (pleasantly) surprised.  One participant was utterly confused by this and didn’t know why someone would want to even search library webpages.

Library Web Pages Section a Center of Gravity

A commonly observed tendency was for the participants to first look for their answer in the Library Web Pages results space before looking at the other sections of the search results page (even though the other sections were better starting points).  In most cases the participants would eventually migrate to Frequently Asked Questions, Search the Collection or Browse Subjects when they did not find what they wanted in Library Web Pages. 

Query Reformulation

Participants would search for a term… briefly look at the results and if they didn’t get their answer they would try a variant of the search term they entered.  The reformulated query was typically broader… participants would begin with a very specific search (sometimes using as many as 6 or 7 query terms), then move to a less specific search to try to find a relevant hit.

Book and Journal Title Results

It is clear that the participants expect the search tool to display book titles and journal titles in the results set so they are no more than one click away.  That the search tool integrates information about the library is appealing, but clearly limited if it excludes book results.  It should be noted that the test was negatively biased on this issue because the participants were instructed to located books and journal titles using the Quick Search tool.  In some cases their instinct was to go to the homepage to find a catalog link or a journal finder link but because the web site tested was in development this was not possible.  A post-launch usability test should be performed that doesn’t instruct the user to use a specific tool to find their information.

Article Search

We expected users to have major trouble locating the article search links.  Surprisingly 4 out of 6 participants were able to locate the “Quick Article Search” or “Advanced Article Search” links, though typically only after the Library Web Pages failed to yield an answer for them.  The surprisingly high number may be due to the fact that at the beginning of the usability test they were asked to describe what the links under “Search the Collection”.

Best Bets… Great, When Found

Several participants really appreciated the direct Best Bet links to the library map and LexisNexis. Two participants even noticed that it appeared even though they misspelled the name.  But there were also two participants that didn’t noticed the link in the yellow box even though it was exactly what they were looking for. In these cases they looked at the first few results in the Library Web Pages section.

Team and Partners

• Josh Boyer
• Tito Sierra

Project Links

• Script read to participants before the usability test (MS Word)

Reports

• Report discussing the website's performance for each type of task and each individual task, with suggested changes (MS Word)

Last updated: June 6, 2009