Due Friday, November 30, 2007 1:30pm
Although P5 was originally planned to be a checkpoint
only, it represents a significant effort on your part and we are short an
assignment. Therefore, we have decided
to grade your preparation for P5. It
will be worth 35 points.
The
usability testing that you will do in P5 is a hybrid of paper prototyping
(because you will simulate your system’s performance, perhaps manually) and
system evaluation (because it is near the end of your “implementation” phase).
P1. Usability feedback (in lab on Friday 11/30)
· Goal: Feedback on the effectiveness of your system
· Prepare two exercises that support the two scenarios you will present in P6. These should include a description of the user and their problem or goal, plus a list of tasks to perform, like the NameVoyager exercise used in the first lab. Tasks may for example include finding particular data value(s), ranking values, finding percentages, finding deviations or comparing to min/avg/max values, comparing particular values, describing trends, finding correlations, drawing conclusions, and so on.
· The testing will be performed like a paper prototype, but your “pages” will be fairly polished, digital versions of your storyboard. Each action by the user creates a new “page” or screen arrangement.
· The lab will be divided into four 25 minute testing slots (4 slots x 5 projects = 20 total trials). Each person in the class will act as a test participant during one of these slots (Marilyn will fill the remaining slots).
· For each slot, one project member will be absent being a participant for a different project. The remaining project members will drive the storyboard and observe their participant. After the tasks are completed, plan questions that will gather verbal feedback from the participant on their satisfaction and/or suggestions.
· Improvements can be made to your system design between participants to react to issues that arise during testing. (You may wish to do this only for serious issues.) Record all changes for potential discussion in P6 and P7.
· Post your participant materials (two test exercises with user description, goal and tasks, and follow-up questions) on your project website.
· Grading: Effectiveness of participant materials for assessing your system, level of preparation for the trials
Download Tableau 3.5 if desired. Stephen Few has written a very flattering review of Tableau 3.5 that mentions useful new features. Read it on Stephen’s blog. You must access Tableau 3.5 through our course web portal. See the Schedule page under the Oct 5 Tableau lab.
Data. You may add additional rows and/or columns to your data to support your scenarios and your tasks. This data can be simulated or approximated as needed to provide more richness. You have grappled with authentic data; now work to design for and utilize the data that you wish you had. For example, you may add more location or time-based data, or additional contextual data. Document your additions and report on them in P6 and P7.
Overviews. In order to demonstrate a full understanding of the topics in this course, we expect that your system will set the context for users by offering one or more visual overviews on the first screen that the user sees. Users will then navigate using appropriate interaction techniques to reach effective views of the information they seek.
Storyboards.
We expect you to create reasonably polished, digital versions of your
storyboards. Each action by the user creates a new “page” or screen
arrangement. You may change the pages/screen arrangements by hand, or use some
simple mechanism like clickable gifs, Javascript, or Flash. You could also arrange multiple windows on
the screen so as to have for example interactive Tableau, Treemap, or similar
windows in one or more regions, and other windows that appear to be control or
detail windows. The user could then use interactive features of Tableau or
similar when appropriate, and actions/transitions of other windows would be
simulated by the project members. You should practice whatever mechanism you
choose, to ensure that you can manipulate it successfully during your testing
periods. You may ask the user to look
away or close their eyes while you prepare the next screen. For dynamically arranged screens, you may
also wish to take a screenshot of each screen before changing it, to document
the sequence and to provide a version you can easily post on the web.
Note: One or two pre-tests with friends or roommates will significantly
increase your likelihood of successful trials in lab.
Tasks/testing.
Class lecture on Tuesday 11/27 will be guest speaker Mary Czerwinski of
Microsoft Research on Usability & Evaluation.
Reference materials about usability testing & paper
prototyping
Usability 101 from Jakob Nielsen's
Alertbox, August 25, 2003
This article
has several links in it. The crucial ones to follow are the ones
mentioned below. Others are optional.
Simple user testing (testing 5 users) from Jakob Nielsen's Alertbox, March 19, 2000
observational techniques (article from Brenda Laurel collected readings book)
Rules for Usability Test
Observers
Paper prototypes. Recall that our testing is a hybrid of paper prototyping and usability evaluation, so should in general be beyond the paper and post-its phase. Nevertheless, the following short articles offer insights about conducting simulated testing.
· Using Paper Prototypes to Manage Risk 01/01/1996
· Paper Prototypes: Still Our Favorite 05/01/1998
· Five Paper Prototyping Tips 03/01/2000
· Looking Back on 16 Years of Paper Prototyping 07/27/2005