The Project
Project Home
The Process
The Data
The Analysis
Forum Topic Scripts
Pre-Test Script
JW Data Archive
participation guide
Data Workbooks
Anne Giffen
Supawadee
Edward Davis
Shiv Pal
Selden Richardson
The Analysis
Scenarios/Claims
Metaphors
PD/Active Theory
|
CS 5724 Models and Theories of Human Computer Interaction Dr. Carroll
The way we worked with the project was by breaking it down into sequential tasks (by
time) that would happen during the course of the semester. This is also evident in the dates that are
available for the data collection stages on the Data and Products page.
The following was the breakdown of the project into stages (which we call the process for our project):
- Project Scope Identification
- Identification of the CHI methodologies to employ for analysis
- Data Collection
- Participating in the online, collaborative session
- Archiving the data
- Finalize CHI methodologies
- Analyse the data using the selected methodologies
- Web-site design and construction (pervasive and concurrent activity)
As we were going to work on William Paul's dissertation , we had to first understand
his work and then try to identify and isolate the part that we could use for an HCI project. This involved
discussions with William and a presentation by him on his work. We went through the web-site
for his work and some of the documents that he identified as helpful for us to get "in-touch". We then
identified the part that we would work on. This was the first test event for his proposed electronic workshop
(the electronic charette, william george paul). The choice was based on the following facts:
- It was an activity spread through 3 months. This suited us just fine as
then we would have enough time for analysis and web-site design.
- It was an interactive activity, so it would involve all the team members to
actively get a feel of the workshop.
-
It was rich enough to permit analysis by a variety of HCI methods.
At the time our knowledge of the HCI methods was limited to a few that
analyse/help to design at the lower-level end of the spectrum. And we had a vague idea of some of the
other methods that we would be learning later and that could help us in evaluating the current work. We
identified the following methods that we thought were most applicable for a participatory, internet related
project:
-
Scenario based design
-
Claims Analysis of the scenarios
-
Metaphor based design/evaluation
-
Participatory Design and Activity Theory
-
GOMS/TAG for the participation guide and web site
The team members joined the group of participants who were taking part in the event
of the first testing of the charette. This involved data collection in a staged fashion over a period of about
2 months in various ways. The following were the main data collection methods that we decided to use for
our project:
-
The EC Forum- this is an online form that allows users to enter comments, suggestions
or raise questions regarding a topic that is being discussed there at the time. This topic is updated every
two weeks or so to allow for enough participation at user discretion and at the same time allow discussion
of a good number of topics to cover most aspects for the collaborative event. Scripts of all comments from
the form for the various topics, including some related to feedback from the collaborative sessions, are
archived and available from the Data and Products page.
-
The Collaborative Event- initially we wanted to have two rounds of this event. The first
was aimed at users getting familiar with the tools (NetMeeting and DataBeam) and their use in a concurrent,
collaborative environment. And then a final event where a similar session is undertaken after sometime. But
then we decided to have collaborative sessions for each of the participants on a one-on-one basis
with William. The latter approach, which we finally followed had the added advantage that we had more
data to work with ... one from each session rather than just two sessions that was proposed earlier. The drop
side was that we did not have a session with more than two participants. Anyway, the collaborative sessions
went as scheduled and the chat scripts from NetMeeting and the image workbook from DataBeam for each
of the sessions was archived and is available as data for our analysis as well as for later work from the Data
and Products page.
As we learnt more of our project, and also of the various other
methods that HCI provides for analysis, we concluded that GOMS/TAG was too low level for our work. So
we eliminated them from our list of proposed analysis techniques and decide to go with the following:
-
Scenario based design
-
Claims Analysis of the scenarios
-
Metaphor based design/evaluation
-
Participatory Design and Activity Theory
We planned to complete our analysis by the third week of November as the last
scheduled collaborative session was for the first week of November.
We wanted to treat the web-site design and construction as an ongoing task
so that we don't face a crunch in terms of time towards the end of the semester when all would be pressed
for time. Besides that would also give us time for a more thorough analysis once the data-collection process
was over.
|