DREU Summer 2011 - Jianling Fang

PSU is huge! During the first two days of this week I basically walked all over the campus to process my campus ID, apply for my wireless account, and do paperworks for my work status. There was some drama going on, because when it comes to international students, everything is SO different and complicated. However, it made me realize the difference between state university and small college. There are only 2400 students in my home university, and PSU has more than 40000 students!

Enough for sad stories. There were a lot of exciting things happening this week. I met my mentor Erika Poole and my teammate Rachael. Both Rachael and I are fascinated by the dog walking project which Professor Erika proposed. So the basic idea is that walking dogs ultimately promote health for dog and dog owners. We are gonna use pedometers to record how many steps they walk everyday and update the results online. So the dog and dog owner as a team can compete with other teams in the community.

For the second week, Rachael and I decided to gather relavent info about dog health. It was tougher than we thought it would be. I went to the PSU library and hoped there should be some professional books or journals related to dog health, and I did find some related research ariticles which talk about similar subject. Not too bad but not enough.

On friday, Rachael and I walked all the way to two of the veterinary medical centers to see if there are any vets available to answer our questions about doggie health. It turned out nobody was available and we left our email addresses to the lady at the reception desk to schedule another appointment with the doctors.

On the weekend, I found some really useful information online about dog health. It lists very specific data regarding how breed and age can affect the perfect weight of a dog. That is EXACTLY what we wanted! =)

This This week I learnt a lot about Adobe Flash Builder, which is a open source framework for building web applications, and somehow similar to Dreamweaver. The interface looks very neat. The only thing that bothers me is that the different versions of flex have different formats. If I put the codes of two versions together, it could mess up the whole page and renders an error. I have to go back and check my code thoroughly to make sure there is no incompatible element.

Although flex is not very flexible, I think it’s more like a professional version of dreamweaver and HTML. It allows the user to better interact with the interface, and is a very good choice for big projects.

After two weeks of work, we finally have our interface.

So when the user logs in to our system. A dog info page will pop up and the user can input the dog’s basic information.

After the user fills out the basic information and clicks "Submit" button, an acitivity tracker will pop up.

The user can switch between the activity view and the ranking view by clicking the bottom left button.

This week we started to integrate the ideas of the papers we read, so we could be prepared for our own paper. I also learnt a lot about IEEE format paper and the citation rules. So far we have got the introduction, literature review, and the system & design sections. I found it kind of hard to condense all of the things you want to say and also to make it looks professional. I’m also aware that it is a very good habit to make notes of the literature sources we have. It’s a pain if you want to cite other people’s idea but do not remember where the source comes from. So we put all of our references on our website under the “Related Works” column.

It’s very handy for us to get the information we need and for citing the source.

To follow Prof. Poole’s decision, we dragged all of our flex code to the recycle bin. From now on, we are going to work with php and html to simplify the interface because we might not have enough time to implement complex features to our interface. What we decided to do has been reduced to: 1. for the baseline data collection week, a dog picture and daily dog facts will display on the kiosk. 2. for the rest of the experimental weeks, we will have the progress bar and the ranking system displaying on the kiosk instead. And if the participants exceed their goals (over 100 percent), a congratulations message will show at the bottom of the page.

Professor Poole helped us implemented the basic HTML code with the cute dog pictures, as well as the php scripts to retrieve data from pedometers. Unfortunately, I’m a noob in php scripts, and it was a pain to learn everything in a short period of time. Gladly, Rachael had some knowledge about php. As she worked on the php part, I was mainly fixing the HTML, css style sheets, and also the paper that we are going to submit.

WeWe now have two appointments for next week, so we need to be prepared. We already have the kiosks and keyboards. All of our pedometers and harnesses have arrived this week. Rachael and I set up, charged, and labeled all of the pedometers. We also set up individual account for each pedometer.

We currently have two kiosk modes for participants. For the first week of the study, the participant will only be shown the dogFactView which displays daily interesting facts about dogs though the kiosk, so we can collect the baseline data and calculate the goals for the second week. From the 2nd week on, we are gonna change to dogFactView to application view which displays distanced walked, miles to go, the ranking system, and the reward system. We fixed the application view to fit to the kiosk screen. However, Rachael found a ranking bug which the system does not show the participant’s own ranking. So we fixed that as well. There might be more problems coming up in participant’s home during the study, but we are confident that we can fix them!

This is the dog facts view:

This is the application view:

We had meetings with two participants for this week. I was definitely not prepared for the overly active bulky dogs! The dogs of one of our participants were really aggressive to strangers. They were jumpy with me and I didn’t even dare to move. It reminded me of one of the biggest obstacles of dog-walking that dogs can be aggressive to strangers. The owners may worry about people getting bitten by their dogs. Fortunately, our participant has a good control of her dogs and they finally calmed down so we could proceed our study.

This was the first time that I conduct experiment at people’s home, and everything went smoothly–except that I got scratched by the dogs. So the participant completed some paper work, and then we explained to her how the kiosk and pedometers work. There aren’t many things to keep in mind: just to put on the pedometers in the morning and take it off at night, and make sure the dog doesn’t chew the pedometer. Then we had a brief interview and asked the participant some questions related dog-walking habits. We audio-recorded the content of the interview for future use.

Overall, the study went well and I was very excited that we finally get started!

This is the second week that we have launched our study. Not surprisingly, problems were found at participant’s home. It has been consistently raining since last week, and the thunderbolt always caused power outages in the town, and then the internet went out, which means we could no longer receive data from the participants, and the computer will not go back to kiosk mode after electricity recovers. Of course, we were not prepared for this. So we decided to write an auto-run apple script to make sure that the computer automatically run the kiosk mode on restart.

This is the last week of our study! We had more participant meetings on Monday, Wednesday, and Thursday. One of our participants has problem with the kiosk, and for the last week, there were only two day records of her physical activities because of power outage. So we replaced her kiosk with the new one which has the auto-run scripts on. If the power goes out again, the participant can just re-plug in the power chord and restart the computer.

We recorded all the data in a separate excel sheet so we could analyze the results. Unfortunately, we found that the result data are not very pleasant because it’s showing a negative effect of our application. I was confused, but then we realized that there were actually 5 rainy days in the past week. So the discouraging results do not necessarily mean that our application is discouraging. I know our application is helpful because some participants were very excited about the progress bar and the rankings we have through interviews. One of our participants always wanted her dog to be the “queen” of the rankings. Our application to some extent motivated the participants to walk their dogs more frequently.

The research study in this summer for me is always like the process from problem found to problem solved, which is the most challenging part of the study. However, I did learn how to deal with unexpected issues and how to solve problems by the help of online resources.