Week 1 ...
This week was my orientation week. Irene
Cooperstein, another DMP student, and I arrived at Corvallis on Saturday
(06/12/2004) and stayed at Laura Beckwith's and Ledah Casburn's (postgraduate
students at OSU) house. Ledah was a DMP participant. She gave us a tour
around Corvallis and the OSU campus. We then met Dr. Burnett, my mentor,
and some of her research team members . I was next introduced to our lab
and the end-user
software engineering project.
The first thing I did at our lab was taking the tutorial: "Human
Participant Protections: Education for Research Teams" by National
Institutes of Health (NIH) since I was going to be involved in a research
involving human participants. Then, I started learning the WYSIWYT
(What You See Is What You Test) testing methodology for end-user programmers.
helps end-users test and find errors in their spreadsheets. It was recently
integrated into a commercial spreadsheet, Excel. I also started familiarizing
myself with the Visual Basic for Applications (VBA) since I would also
be helping our team debug in the integration process. Later in the week,
I debugged a minor bug and prepared a handout about the WYSIWYT for the
upcoming study with the high school teachers and students.
Week 2 ...
During my second week, I helped my mentor find and read papers to answer
the reviewers' comments on her journal paper, Interactive, Visual
Fault Localization Support for End-User Programmers, she was about
to publish. We found an interesting paper about people being polite to
computers when they had to evaluate or criticize the computers face-to-face
(a paper by Clifford Nass on communication). I also helped our team debug
a few more minor bugs.
Every week I had three meetings and a picnic. On every Tuesday, the whole
group had a research meeting, which we discussed about the issues in our
research project in general. On every Thursday, there was a technical
meeting, in which we discussed specifically about technical issues of
our project. Right before lunch on every Wednesday, I had my weekly meeting
with my mentor to talk about any issues or problems that I might have
had. And we had the OSU Computer Science Department picnic every Wednesday
at the lawn just outside our building.
At the research meeting, I was assigned to give a presentation to the
high school teachers about our WYSIWYT prototype and work with them in
Academy Program at OSU in the following week. I prepared more handouts
and tutorials to help them understand our WYSIWYT testing methodology.
I was a little nervous about the presentation since I myself was in the
Tyler, Andrew and I, who were in the technical group, worked on Sunday
to fix some bugs in our WYSIWYT/Excel prototype and got ready for the
presentation next week. I was very glad that I was in a group of people
who were very hardworking and wholehearted about their project.
Irene and I moved into the West International dorm at OSU. I had fun having
Irene as a roommate. She taught me how to knit and we both had good conversations
at night while knitting.
Week 3 ...
During this week, I gave a presentation about the WYSIWYT prototype (Excel
Version) to the high school teachers as planned. Although there was a
technical program at the beginning, we were able to fix it in time and
the presentation went quite well. :) I was kind of relieved after that.
Our computer science department presented altogether three prototypes,
which were UCheck, Gencel and WYSIWYT, and one existing feature of Excel
to those teachers. I worked with one of the teachers who chose to learn
more about WYSIWYT. We worked through a tutorial that I prepared and talked
through how she could incorporate this testing methodology into her lessons
plan for the Saturday Academy students whom she would be teaching the
following week. Our goal of introducing these prototypes to the teachers
and the Saturday Academy students is to bring in the notion of taking
responsibility in using the technologies.
At the research and technical meetings, I talked about the feedback I
got from my presentation and from working with the teacher. The feedback
would help us in developing our prototype or in planning next year Saturday
Academy program. I presented one bug, which was that our system couldn't
handle the spaces in the file names, and the issue of the testedness and
fault-localization indicators or toolbars at the top. The issue was that
since both indicators were implemented as rows of separate buttons that
would look like progress bars, the users were confused about the reports
by those indicators on what percentages of their program had been tested
and the proportions of the error likelihood in their program. I found
out that the latter issue had been around for a while and their attention
got diverged to other issues since there were a considerable numbers of
issues to pay attention to. I could imagine having minor and major issues
in developing a prototype, and I was very glad that I brought the issue
that they forgot back to the table.
For the rest of the week, I worked on arrows implementation in Excel with
the guidance of a phD student in our group. Since I was adding a new feature
to WYSIWYT Excel system while I was still learning how it was implemented,
the progress of my implementation project was slow; however, I learned
a lot along the way.
This weekend is the 4th of July weekend and Irene, Laura and I went to
downtown Corvallis to see the fireworks. It was fun! I had India food
for dinner at the 4th of July fair, and was there until about 11 pm enjoying
Week 4 ...
I was a technical supporter to the teacher who presented the WYSIWYT prototype
to the Saturday Academy class. There were about 7 high school students
in that class. The Saturday Academy class just started that day, and each
class would take 2 hrs long. The students were with varying background
on Excel; therefore, I ended up helping them how to use Excel to create
a simple spreadsheet for the whole class time. We could only present our
WYSIWYT prototype to them only in the last ten minutes of the class time.
This was also because we hoped to have a spreadsheet that they created
themselves and test their spreadsheet with WYSIWYT after that. In this
way, they would completely know how their spreadsheet should behave and
therefore be able to utilize the help our WYSIWYT could offer to test
their spreadsheet.Of course, we ruled out the fact that in the completed
integrated system, where the WYSIWYT system was fully integrated in Excel,
the end-users would test their program while they were still creating
Since the students had to learn how to use Excel and at the same time
use our WYSIWYT prototype, it was just too much information for them to
take in. Not so many questions were raised about it, and they barely had
time to play with it. Therefore, I didn't have any feedback from them
to present at the meetings except the one that we should give them more
time to play with it next year.
For the rest of the week, I worked on the arrow implementation again alone
and with the help of a phD student. I was getting frustrated about it
since it was going very slowly and also there were a lot of problems with
the coding. :(
Week 5 ...
This was my fifth week. In this week, I helped my roommate, Irene, with
her Think Aloud Study on Gencel. I got to know the two phD students who
were working on the Gencel prototype and another undergraduate student
in that group. We had two or three small group meetings for that Think
Aloud Study. I also took the tests and answered the questionnaires as
a participant so that Irene would know how long the tasks in her assignment
would take. It was a good experience for me since I got to know the other
prototype Gencel and how to conduct a think aloud study.
I went to the meetings and picnic as usual. I was very glad that Andrew
was also assigned to reinforce me in implementing the data flow arrows
in our Excel arrow. I was very impressed by Andrew, who was very good
at programming, in that case VBA. I learned a lot from him and we did
make a progress in our arrow implementation .I also prepared a good spreadsheet
to test our WYSIWYT/Excel prototype.
This weekend I tried to learn about GRE exams and tried answering some
questions. I was very glad that I was chosen as a DMP participant, and
got matched up with Professor Burnett. She was very encouraging and I
felt very positive about going to the graduate school. However, I hate
Week 6 ...
During this week, I mainly worked on a response letter to a journal paper,
"Empirical Evaluations of WYSIWYT Testing for Spreadsheet Languages".
That paper contains the results of three different study conducted at
three different years. At first glance at the paper and the questions
from the reviewers, I was intimidated because there were a lot of questions
about the statistical evaluations. I took a statistic class for economists
during the fall semester of my sophomore year at Smith, and I felt like
I forgot most of it. :( However, Dr. Burnett was very encouraging and
she also told me that I could discuss about it with her. So, during this
week, I spend most of my time reading that paper and trying to understand
the statistical evaluations.
I also helped Laura, who is doing a thesis on Gender HCI, in her study
this week. I got to know the people who involved in that group. There
are two other people, Michelle a high school intern and Shraddha a master
student at OSU, besides Laura. We first gave the participants a tutorial
on how to use our spreadsheets in Forms/3 language and asked them to do
testing tasks with our WYSIWYT prototype. There were about 10 participants
who were not computer science students in each section. I mainly help
them in the tutorial part of the section. It was a good experience for
me since I got to know the procedures of conducting a study and collecting
Professor Gregg Rothermel, who is one of the main researchers in EUSES
consortium and who has moved to Nebraska University, joined us in our
meetings this week. Each of the us in our group introduced ourselves to
him and explained a little bit about the projects we were doing or involving
to him. At the meetings this week, I was assigned to demo our WYSIWYT
prototype, both Excel version and Forms/3 version, to Professor Susan
Wiedenbeck from Drexel University. Professor Wiedenbeck is doing a research
related to HCI. I prepared some example spreadsheets to show off our testing
methodology. I was a little nervous about it since I might not be able
to explain the technical details of our project that she might ask. I
did not know much about the back-end, in lisp, side of our prototype.
This weekend, I tried to finish my progress report to DMP. I read the
Asimov's book again. :) I was very addicted to it. I enjoyed his plots
a lot. I also tried to do some GRE questions.
Week 7 ...
This week was a very hectic week for me. I worked on the answers to the
questions by the journal paper reviewers, and I also had meetings with
Dr. Burnett on that. I learned the process of submitting the journal paper.
I also figured out how systematic Dr. Burnett and her team were in preparing
for journal papers and yet there could be bumps along the way before a
paper got published.
I also worked with Professor Wiedenbeck for the whole morning on Wednesday.
She did not ask me any technical questions in detail. It was fine at the
first part of the section while I was explaining her about the features
in Forms/3. However, to my very amazement, the excel version of our prototype
had a bug. We all missed that bug while we were trying to test it on Tuesday
with my example spreadsheets. We always try to test our latest version
of our prototype in advance before we did our demos, and yet there was
one that we didn't notice. We were still adding more features to our prototype,
and it was very likely that new bugs were introduced to it. Professor
Wiedenbeck was very understanding about it. I had to report to Dr Burnett
about my demo section to Professor Wiedenbeck afterwards and therefore
I reported her about our prototype having a bug. I was very sad that it
happened and I felt like we should had tested more thoroughly before the
demoing. I also reported it to our technical team. :(
I was also trying to finish up taking good screen-shots of a few spreadsheets
being tested by our prototype. The screen-shots were for the book that
another professor at the Computer Science Department was about to publish.
I also had to write a paragraph explaining about our testing methodology.
To write a paragraph about it seemed to be more difficult than I thought
because I had to explain the essence of the whole system in a very precise
matter. However, I managed to do it somehow and I was glad that I was
done with it. :)
One of the frustrations in this week was preparing for an example that
Dr. Burnett was planning to demo at the meeting at Microsoft. I felt very
proud that I had a chance to do that. However, I was not very confident
about my ability to come up with a good example that would be useful to
present all the features of our prototype. The most difficult part of
that process was to come up with a seeded bug in the spreadsheet.
Week 8 ...
This week wasn't as hectic as last week but still very frustrating. I
was still working on the journal paper. I read two papers by Panko on
spreadsheet errors so that we could reference the data from them. According
to his paper, the spreadsheets that end-users created could be very error
prone, and at least, 1% of all cells contained errors.
I also worked on arrow implementation with Andrew. As the end of ten weeks
was approaching, I was frustrated about being able to get things done
Week 9 ...
This was my ninth week at OSU. The journal paper revision process was
almost done. I had to meet again with Dr. Burnett to brush up on it. All
the questions by the three reviewers had been addressed. There were also
some modifications in the paper. As requested by one the reviewer, we
needed a screen shot of a spreadsheet used in one of the case studies
in the paper. Since the study was conducted 3 or 4 years ago, we had a
hard time finding it.
I also applied to DMP reunion during this week. I was hoping that I would
get a travel reimbursement.
Week 10 ...
My last week! During this week, I was finishing up on the journal
paper I had been working on. I finally got a satisfied screen shot of
the pizza spreadsheet that was used in one of the studies in the paper.
I was really pleased that finally this paper was at a stage where we could
resubmit again, and I really hoped that it would get published this time.