Distributed Mentor Project - Summer 2007
Week One (5/28/07 - 6/3/07)
I went out to lunch with my research partner and my mentor to discuss the specifics of the
research project we were to be working on. Our assignment for the day was to find an IM Client
that would log a user's idle time sufficiently. My partner and I looked into AIM (AOL Instant
Messenger) and a universal client program, Pidgin. We began logging idle time using both. We
then researched various types of methods into creating a chat bot for the AIM Client.
My partner and I spent a majority of the day filling out paperwork to gets IDs, keys, and
wireless accounts on campus. We had a meeting with our mentor about the interview process for
HCI research, and using the AIM SDK for creating a chat bot. Our assignment for the day was to
go over some suggested reading and create a concept map.
We had another meeting to touch bases on the project, discuss our concept maps and other ideas,
and to finalize some decisions. We were given more suggested reading to look through, and were
assigned the task of coming up with general questions for the interview process.
I completed the certification to perform interviews with human subjects for HCI research. I
began implementing a chat bot using the AIM SDK, and ran into a lot of problems trying to run the
sample bot. There was little discussion on the issue. The ultimate goal was to get the bot online to begin
logging idle time, and printing it out to a file. I spent the rest of the day continuing to read
My partner and I attended two meetings today, and met most of the other people in our lab department.
We were able to learn about other research projects going on, and even share some of our ideas.
I spent the rest of the day starting the pilot study with one of my friends. The monitoring and
the initial interview questions took almost half the day.
I was finally able to find a fix and a tutorial for the AIM sample bot in Java. My partner and I
started up the bot, and made some minor modifications.
Week Two (6/4/07 - 6/10/07)
We had a meeting to touch bases on the project, and to make a timeline of our goals for the project
during the summer. My partner and I refined the interview questions, and started a new set of
questions to be asked on a daily basis. These questions were going to be made into surveys for
easy access to users. We then began the pilot study on our mentor, only to find out that we were
still lacking key follow-up questions.
We revised the interview questions again, and came up with some open-ended and more specific follow-up
questions for the interview. We sent those out for approval, and continued modifying the bot in the
meantime. We were able to get it to successfully log user status changes, and print them out to a file
including a timestamp.
I finished website construction for my DMP website. Later that night we ran the pilot study on our
mentor again, with the revised interview questions and follow-up questions. This session went a lot
better, and we were able to finialize some decisions.
We had our weekly meeting to discuss the project. We started talking about advertising, running the
chat bot on a central server, different forms of feedback, and even creating a client-server based
form of feedback. Later, we refined the bot's logging system so that we eliminated privacy issues.
Week Three (6/11/07 - 6/17/07)
I started looking into various community websites that I could post ads to for the initial study.
I also looked into different types of ads, and which ones might be more effective than others. I
also created a seperate screen name (SN) for myself on AOL Instant Messenger (AIM), so that I
didn't need to use my personal SN for the study.
I began creating a couple of ads to see what they would look like, and I continued research into
online communities. I spent a portion of the day working on my DMP website. Mostly, I worked on
the "Summer Life" page, finding pictures and information.
I finished the first mock up of the ads and sent them to the rest of the team for comments and/or
suggestions. I continued looking into online communities, and I discussed making a script to
have the bot run automatically on a central server.
We had a morning meeting to discuss the project while our mentor was away for a week. I finalized
some decisions on the ads, and where they should be posted. We talked about starting another pilot
that would last a couple of days, before we started recruiting. At this point, we were mainly
waiting on IRB approval to start. Later, I rewrote the ads according to the specifications discussed,
and I made accounts on the online communities I would be posting to.
I mostly worked on my DMP website, while still waiting for IRB approval to start the study.
Week Four (6/18/07 - 6/24/07)
While still waiting for IRB approval, I put most of the initial interview survey on Survey Monkey,
and edited the daily survey. I also read a paper on the differences in energy consumption for when
a computer is put in "sleep" mode verus "hibernation" mode.
I finished the putting the initial interview survey on Survey Monkey, moved the daily survey over
to a paid account, and finished editing both. I also looked into reinbursement policies if you pay
your subjects with online gift cards that were purchased with a credit card.
We're still waiting for IRB approval, but we've started another pilot study that should last a
couple of days.
Week Five (6/25/07 - 7/1/07)
We finally receieved word from IRB. We had to sign some more forms and fax them in. Now we're
waiting for final approval from IRB. We started the pilot study on myself so that we could catch
any last bugs in the system. I had to spell check a couple documents.
Continued the pilot study on myself. Received word that we would be able to get reinbursements for
online gift cards purchased with credit cards.
We had our weekly meeting to discuss different types of feedback and analysis once the initial
study is complete. We finally recieved full IRB approval, and could start advertising online.
Finsihed the pilot that night, and made some last minute changes.
Finished posting advertisements, and started processing the initial responses. We started to mark
who was qualified and who was not, but it was obvious that we were going to be overrun by responses.
We sent out hundreds of e-mails that night, and made it through part of the consent form
signing process. A couple of participants didn't seem to understand the process very well, and it
was hard to handle those situations.
Logged into our e-mail account in the morning only to find a horde of e-mails. I deleted the ad
from Craigslist, only to find that individuals had taken it upon themselves to post our ad elsewhere.
It was frustrating to deal with, but we knew we couldn't find all the ads to delete them. We simply
had to keep turning people away. I put a cap on the number of signatures we could have on the
consent form, on Survey Monkey, so that we didn't get too many responses. We recieved clarification
from our mentor that we needed to send out the link to the initial interview survey and have that
filled out before we started the actual study. I went and found instructions on how to show idle
time for Trillian users. We wanted to make sure that everyone was showing up as idle at the
appropriate times. We then started e-mailing initial instructions for the study and the link to
the initial interview survey. I used a random number generator to pick out ten random numbers out
of forty, so that we would know who to give the additional interview to at the end of the study.
I also sent out an e-mail to my team with my suggestions for feedback devices once this intial
study was concluded.
We finally had all forty signed consent forms that were needed to start the initial study. I began
to check which participants had completed the initial interview survey, that had been sent out to
them at the receipt of their signed consent forms. I made a spreadsheet that showed each participant's
e-mail, SN, whether they were getting the additional interview, whether they had completed the initial
interveiw survey, and how many of the daily surveys they had completed. I color coded the four
groups, since we weren't starting everyone on the same day. I let my partner update who had added
the bot to their buddy lists.
I updated the spreadsheet and the email account one more time, before I handed everything over to my
partner. She started sending out e-mails to start the daily monitoring process. Since I was leaving
for vacation, I would be missing the first week of the study. All the initial prep was completed at
that point, though.
* One week vacation to attend a friend's wedding (7/2/07 - 7/8/07)
Week Six (7/9/07 - 7/15/07)
I was able to catch up on what I had missed during my week of vacation. During the time I was gone,
it was decided that we were going to try to add desktop users to the initial study. It was a last
minute decision, and there was some trouble recruiting. I spent the remainder of the day looking
up examples of AIM plugins and how to implement them. The initial study is still in progress.
I looked up credible census information for how many computer users there were in the United States
and then specifially the area of Pittsburgh. This data would be used later on to analyze how much
idle time is accumulated on a larger average. I rewrote the ad for the desktop user study, and
reposted the ad in various places. Our response rate improved and we were able to continue with
the initial study. The study should end at the beginning of next week. During the team meeting
we discussed the progress of the CPU monitor we will be using for phase two of the study.
I started the first group of laptop participants on the IM Diary. This diary will ask the participant
what they were doing and the status of their computer every time they come back from idle. This
diary will last one day.
Week Seven (7/16/07 - 7/22/07)
I drew up some mock-ups for the feedback windows in Adobe Photoshop and sent them to the group for
comments. I e-mailed the first two participants that I would give an additional interview, scheduling
a time for the interviews. I analyzed their data, looking over the daily surveys, general surveys,
the IM diaries, and the logs showing idle status changes. From this analysis I created some specific
follow-up questions for the interview.
I went over the data analysis one more time for the interviews. I also made some adjustments to the
feedback windows in Photoshop.
At the group meeting, we discussed the issues with starting the second phase, including the initial
survey that the participants have to fill out. We're behind schedule to begin the second phase. We
also discussed what will happen to after I and my partner leave.
I went through the analysis I did for the interviews and commented on the conclusions I had come to.
I sent out my analysis, the status change logs, and the interview dialog to the rest of the group for
I scheduled three more interviews for the laptop participants on Monday.
Week Eight (7/23/07 - 7/29/07)
I performed my analysis for the three participants I would be interviewing, and only interviewed
one of them. The other participant failed to sign on for their interview, and I sent them an e-mail
My partner and I went to a presentation to Intel to tell them about our research. The Intel representative
informed us that their research team was performing similar research.
The group had a lengthy meeting to discuss all the issues with launching the second phase and how
we could get the issues solved faster. At this point, we're very behind schedule. I finished my
analysis for the final interviews, and conducted the last interviews. It turned out that the last
participant was scamming us. It was a messy situation to deal with.
I sent out my analysis of the last three interviews to the rest of the group. I started my
analysis of phase one to begin making feedback suggestions for phase two.
I had a progress meeting with my mentor. We discussed jobs and possible graduate school programs
that I could apply to. I finished my initial analysis of phase one and sent it out to the group
for comments before I started making feedback suggestions. I sent out and e-mail to my mentor
so that the last participants interviewed would receive their payment. I began to take over
phase one for the desktop users so that my partner could work on other things.
My new responsibilities for the desktop users is to check the daily surveys every morning and
update the spreadsheet. I'll then e-mail any participant who missed the survey, letting them know
how many surveys they have remaining that they could miss. Every night, I send out reminders to
fill out the survey via AIM and e-mail.
Week Nine (7/30/07 - 8/5/07)
I wrote down ideas for the slide presentation that's going to take place this Friday. The
presentation is to show the rest of the department what our group has been working on this
summer. My partner and I were responsible with making the presentation.
I finished making the feedback suggestions for phase two and sent them out to the group for
comments. I sent out e-mails to the first desktop user to finish phase one of the study. The
e-mail included uninstallation instructions for the CPU sensor that each participant had to
install on their machine.
We had our group meeting to finalize everything for phase two (i.e. who will do what, etc.).
I finalized another participant for the end of phase one, sent out emails to coordinate
interview times, and began to pay participants who had finished. I started my analysis process
for those participants who were going to be interviewed. I also updated the general survey for
phase two of the study.
I started the bot diary for a participant finishing today. I also finished my analysis for the
participants I interviewed later that night.
My partner and I had our presentation to the department this morning. It went fairly well. We
should have done this presentation at the beginning of the summer, since we received some very
good ideas. However, it was too late into the project to implement any of those ideas. Today was
my partner's last day, and she handed over to me all the bot code that she had been working on.
We couldn't seem to get the code running on my machine, so that's something I'll have to debug
later. One of the professor's wrote a data structure for me in Java that will help with the code
I need to write to generate personalized suggestions for a participant. I also finalized another
participant, and paid the last three.
I continued to debug the bot code on my machine with no real luck at getting it to run correctly.
I began to update an online spreadsheet that contained the data from both phase one and phase two.
I started on the code for prioritizing personal suggestions. I also began to back up the status
changes log from the bot onto my machine. I will have to do this every night until phase one is
Week Ten (8/6/07 - 8/12/07)
I started a participant on the bot diary. I updated the online spreadsheet with new information.
I tried to implement a new version of the bot on my machine, but it still does not run properly.
I continued to work on the suggestions code.
Continued to finalize participants from phase one and pay them. I also updated the online
spreadsheet and worked on the suggestions code. I had to end a participant who missed too many
of the daily surveys and no longer qualified.
We had our final group meeting today to touch bases. I finished phase one of the study, paid the
participants who finished, and set up interviews. I updated the online spreadsheet to look like
the hard copy spreadsheet that I used for phase one. I updated the bot code and sent it to my
mentor to review.
I backed up the log from the bot one last time, and separated out the lines that only pertained to
the desktop users for phase one. I completed the analysis and final interviews for phase one. I
went through all the separate participant logs and pulled out the total idle time for each
participant for each day, and put that information into a separate spreadsheet.
I sent out all of my remaining files including interviews, logs, analysis, etc. I sent my mentor
the remaining payments that were unused in phase one. I dropped off my key to the lab, and finished
filling out some paperwork. I updated the online spreadsheet one last time to make sure everything
* Over the next couple of weeks, I will spend time helping out the team when I can, and finish
updating this website.