Web Usability Training 10/31/2008











* Your preference isn’t the same as everyone else

* Lengths: In general use shorter pages for homepage and navigation pages and apage that need to be quickly browsed and read. Use longer pages to

– facilitate uninterrupted reading esp. content pages

– match the structure of a page counterpart

– simplify page maintenance (fewer Web page files to maintin)

– make pages more convenient to download and print

* Line length: if reading speed is most important, use longer line length (75-100). if acceptance of the site is most important use short line length (50 characters per line)
* Web style guide: “one of the fundamental principles of the Web however, is that users should be able to structure their own view…..So although leaving text free to fill the browser window may affect readability, following conventions may also affect the accessibility and legibility of your documents”
* White space: limit the amount of white space (area without text) on pages that are used for scanning searching, on content (text) apges use some white space to separate paragraphs, too much separation of items may require users to scroll unnecessarily.

SEE: http://www.webstyleguide.com/type/space.html

* Navigation: web site’s navigation scheme and features should allow users to find and access information effectively and efficiently. When possible, thi
* Dead end: dont make pages with out any navigation, never assume that a user will have a back button, don’t open a new window without letting the user close it with a link
* Group navigation: consistent grouped navigation improves users access and accomplishment rate
* Navigation: use clickable in page navigation for long pages, content headers should offer in page navigation, use “Return to Top” on lengthy pages.
* location: let the user know where they are, bread crumbs are a great way to keep the user informed about their location, differentiate navigation elements for the users location too
* Placement: user navigation times improve the primary navigation on the left, group secondary and tertiary menus together, users prefer all to be grouped, grouping all three increased right side menu performance to almost the same level
* To tab or not: only use tabs when they are descriptive, users dislike tabs that leave doubt, emulate software and real world tabs as closely as possible, avoid requiring the user to mouse over a tab to discover that it is a navigation element
* Navigation only: keep it short, never require a user to scroll on a navigation only page, users often to do not continue past a single screen full and may miss any navigation options displayed…
* Menu types: use sequential menus for simple forward-moving tasks, and use ‘simultaneous’ menu for tasks that would otherwise
* Sitemap: use site maps for sites that have many pages, site maps provide an overview of the site, they may be

– hierarchical view

– resemble a traditional table of contents

– a simple index

* Glosses/glossary: preview links with information to help the user find the correct link, never use to make up for poorly titled links, keep it close but dont mess up the main text.
* Breadcrumbs: uneducated users rarely benefit from them, users who have learned how to use them tend to experience a significant increase in take completion speed.
* Iterative design approach: build prototypes, do usability testing, continue until goals are met

– benefits: 30% increase in task completion, 25% decrease in time required

* Solicit tester comments: testing is more effective, “think aloud” comments, retrospective comments
* Evaluate before and after: when you redesign a site, test it before and after to ensure your goals are being met and to help define those goals.
* Usability priorities: rank and fix usability issues by :

expected | outcome

#1 easy | hard

#2 hard | easy

#3 easy | easy

#4 hard | hard

* Frequency vs severity: frequency is determined by number of users experiencing the problem, severity is the individual users problems, fix the most severe problems
* Right size test groups: testing with experts, testing with users ( graph looks like y=ln(x) )
* Use the right prototype:

paper – often used early, when computer cannot support idea, when members of the team lack skill

computer –

* Inspection evals: good for surface, bad for deep issues, often requires high numbers of evaluators, heuristic evaluations often misapplied
* Evaluator effect: it occurs when multi evals evaluated the same interface detect markedly different sets of problems
* Automatic evalutations: commercial tools are a great start, never a replacement of professionals and usability testing, will usually find slow pages missing links jargon accessibility
* Cognitive walkthrough: often exaggerates issues only about 25% of potential problems when compared with use testing

frequ misses about 13% of problems

* Lab vs. remote: almost identical task completion rate, time to complete, satisfaction scores
* Severity ratings: often misleading, different teams identify issues differently, heuristic…