An example of a test report
• I worked with our web team and created a card sorting test in Optimal Workshop
• The test was launched on Sep 19 2018 and closed on Oct 02 2018
• The latest response was received on Sep 21 2018
• 51 (68%) people completed your study out of a total of 75 participants. 24 abandoned
• Harvard's completion rate is 67%
• Above 60% is a fair conversion rate for surveys
• Median time: 2. 57 min
• Only 3 emails that completed are spam - and not counted here
• Takeaway: we should not have included FAQs as an option for a main nav item. FAQs should be part of the 'in-context" content help.Our users placed the following into FAQs (see below on possible solutions)
• # of times each card was sorted into a corresponding category.
• The darker color the stronger association with that category users have.
• Lighter blues indicate the uncertainty in placement.
Problematic labels are illustrated by the Popular placement matrix (below).
Percentage of users not understanding terminology:
• DocXpress - 21% (anecdotal evidence - from Chris’ talk “But we’re not doctors, I thought “DocXpress” means only ‘Doctors’ can use it" )
• Medical History Center - 17% . (Harvard has 60% did not understand what the "Center for the History of Medicine" means but they have a different audience)
• Course Reserves - 15%
• Health Statistics, Library Teams, Bassett Anatomy Collection - 11%
