COMP6703 : eScience Project III ArtServe on Rubens

1
COMP6703 eScience Project IIIArtServe on Rubens

• Emy Elyanee binti Mustapha
• u4160964_at_anu.edu.au
• Supervisor Peter Stradzins
• Client Professor Michael Greenhalgh

2
Background

• It is established in January 1994 by Professor
  Michael Greenhalgh, who is the client for this
  project
• A website containing mainly visual images (photos
  from around the world)
• ArtServe has attracted many users up to 201153 in
  July 2005

3
Background - Continue

• Problems are lack of proper database, lack of
  proper search tool within the website, poor
  usability
• Previous semester, an eScience student has chosen
  this project and solves the problems
• For this semester, I did the usability testing
  for the website and implement the necessary
  changes based on the findings and recommendations
  made by the participants.

4
Requirements

• Installation
• Usability Testing
• Implementation based on Test Results

5
Installation

• Apache
• For usability study purpose, an ArtServe folder
  was created beneath the eScience website.
• PHP
• Automatically used the same php configuration
  file as eScience website.
• Trouble with coding fix the errors
• MySQL
• All the necessary tables were created and a small
  portion of the real website data was transferred
  to the eScience machine.

6
Usability Testing

• Cognitive Walkthrough
• Cognitive Walkthrough is a usability inspection
  method that focuses on evaluating a design for
  ease of learning, particularly by exploration.
  (Nielsen Mack, 1994)
• Heuristics Evaluation
• Heuristic evaluation is a usability engineering
  method for finding the usability problems in a
  user interface design so that they can be
  attended to as part of an iterative design
  process. (Nielsen Mack, 1994)
• User Testing
• Since Cognitive Walkthrough and Heuristic
  Evaluation do not include real users, I decide to
  perform a small scale User Testing to get the
  test results for the improved website.

7
Cognitive Walkthrough

• Purpose
• To provide the developer with a list of findings
  and recommendations on how to improve the
  usability of ArtServe.
• Guidance for the next usability test conducted to
  ensure the right test method and task list is
  prepared.
• Problem Statement and Test Objectives
• To gain quantitative and qualitative data
  addressing specific concerns about ArtServes
  usability. The focus of the project will be task
  oriented and directed toward how the user
  subjectively responds to the issues listed on the
  task list.
• User Profile
• Available students around ANU.

8
Cognitive Walkthrough Test Methodology

• 1.   Participant greeting and pre-test
  questionnaire
• 2.   Briefing
• Participants will each receive a scripted,
  verbal introduction and orientation which will
  explain the purpose and goals of the test. 
• 3.   Walkthrough
• The performance test consists of a series of
  tasks that the participants will carry out while
  being observed. During the performance test the
  monitor will make notes on elapsed time and
  participant errors.
• 4.   Participant Debriefing
• Each participant will be debriefed by the test
  monitor.

9
Cognitive Walkthrough Evaluation Measures

• Quantitative Data
• - Time required to complete task
• - Count of incomplete tasks
• Qualitative Data
• - Participant comments and/or mannerisms
• - Usefulness of the navigational terms used
• - Stated preference

10
Heuristic Walkthrough

• User Profile
• EScience students, who are experienced computer
  users with basic experience with usability
  testing.

11
Heuristic Evaluation Test Methodology

• The same as Cognitive Walkthrough except for
• 3.   Actual evaluation
• The evaluator will evaluate the system as
  individual and will do this independently. It is
  expected to be completed in 30 minutes for each
  evaluator. There are two phases as written by
  Jakob Nielsen which are
• a) By stepping through the pre-specified usage
  scenario
• b) By performing a more detailed analysis of
  individual dialogue elements.
• 5. Severity Rating
• Since not all usability problems can be solved,
  therefore a severity rating is done at the end to
  prioritize which usability problem is crucial.

12
Heuristic Evaluation Evaluation Measures

• The evaluation measures will be subjective
  observations. The subjective observations
  include
• Participant comments and/or mannerisms
• Suggestion for improvements
• Major/minor problems
• Severity ratings

13
User Testing

• Purpose
• To get feedback from real users who are in the
  future going to use the website.
• User Profile
• eScience students
• Faculty of Art students
• Repeating participants
• Administrator primarily, administrator is
  Professor Michael Greenhalgh. However, a small
  group of eScience students might be recruited to
  act as administrator to the website.

14
User Testing Test Methodology

• The same as Cognitive Walkthrough except for
• 3. Actual Testing
• Participants complete the task by completing all
  the task from the user guide.

15
User Testing Evaluation Measures

• The evaluation measures will be subjective
  observations. The subjective observations
  include
• Participant comments and/or mannerisms
• Suggestion for improvements

16
Test ResultsFindings and Recommendation (1)

• No index page for Administrator
• Recommendation Create a an index page with a
  simple menu to help admin to navigate the
  website.
• Every function opens a new window, which confused
  users into thinking they are still in the
  previous page and keep clicking the back button
  on the browser.
• Recommendation Every function does not opened a
  new window.
• Content placement and labelling is not
  consistent.
• Recommendation Ensure all links are labelled
  consistently and accurately describe the content
  of the page to which they are directed.

17
Test Results Findings and Recommendation (2)

• Pages often do not have content titles
• Recommendation Add titles to pages that are
  displayed in the content frame. The titles should
  also match the link that points to the page.
  There should be a consistent labelling scheme f
  or links. The sitemap should be updated to
  reflect site content.
• No site map for naïve users.
• Recommendation Ensure the sitemap reflects
  hierarchy of the site with recognisable titles.
• Titles that look like links
• Recommendation Instead of underlining the
  title, it can be highlighted using different
  colours or increase the font size to emphasize
  it.

18
Improvement based on Test Results

• For ease of navigation, a main page complete with
  menu is created for the administrator.
• User guide is included online as part of the
  website to help both visitors and administrator.
• Appropriate feedback for actions taken for edit
  album, edit metadata, delete album, block album
  and unblock album.
• No multiple windows except for the help window.

19
For ease of navigation, a main page complete
with menu is created for the administrator.
20
Administrators Main Page
21
Visitors Main Page
22
User guide is included online as part of the
website to help both visitors and administrator.
23
Visitors User Guide
24
Administrators User Guide
25
Appropriate feedback for actions taken for
contact us, edit album, edit metadata, delete
album, block album and unblock album.
26
Contact Us Form
27
Contact Us feedback
28
Edit Album Form
29
Edit Album feedback
30
Edit Metadata Form
31
Edit Metadata feedback
32
Delete, Unblocked, Blocked Album Form
33
Delete Album feedback
34
Unblocked Album feedback
35
Blocked Album feedback
36
No multiple windows except for the help
window.
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
Conclusion (1)

• 4 main improvements
• For ease of navigation, a main page complete with
  menu is created for the administrator.
• User guide is included online as part of the
  website to help both visitors and administrator.
• Appropriate feedback for actions taken for edit
  album, edit metadata, delete album, block album
  and unblock album.
• No multiple windows except for the help window.
• 3 sets of usability testing
• Cognitive Walkthrough
• Heuristic Evaluation
• User Testing

41
Conclusion (2)

• Background differences effected the test results
  eScience and other students
• The test monitor (myself) should be an
  independent person without any interest to the
  test results.
• The test results for User Testing show that the
  improved website is more accessible and better
  usability, therefore the participants are more
  aware of other minor usability issues colours,
  text