DREU Summmer Internship

If, somehow, you ended up on this page without any context, let me tell you what's going on. I took part in a program called Distributed Research Experiences for Undergraduates (DREU). It allows undergraduates to work with researchers at institutions across the country and get funding! I was sent to work at the beautiful Roosevelt Island campus of Cornell Tech.

The plan

I will work on a pre-existing study investigating teachers of the visually impaired use of tactile materials. We explored which materials they used, how they procured them, and most importantly looked for insights about how to improve them. I will help finish the analysis of the data and write a paper showing our findings that will hopefully be accepted to a conference.

• Week One: The grad student who ran the study on tactile materials had left for an internship so he could not give me instructions on how I was supposed to finish up the analysis of the tactile materials study and start writing the paper. So, in the meantime, I worked on the literature review for the text-to-speech study. I gathered papers for the related work section, while also getting sense for the style of paper in the human computer interaction field. When the study runner was availible, I got instructions and I was sent to pick out data from a giant spread sheet and input it into the draft of a paper. However I found the data was very disorganized so I asked permision to re-organize a little myself. As I organized more and more and went further and further into the data I found more things to organize and eventually ended up completely changing a catagory. It was hard managing so much data and trying to understand some of the methods of qualative analysis but it was nice to be trusted with so much responsibility.
• Week Two: As I was looking into the data, I found some of it was missing. I knew we had 13 participants but I only saw data from 9. It turned out the new data hadn't been input into the spreadsheet I was using. So I put the new data into the spreadsheet and organized it too. Then it turned out not only was some data missing a lot of the original catagorization of the data had been lost. So all my work was for naught. It was a little frustrating but it justified why the data was so disorganized. So I started to compile the complete data when my mentor said that actually she wasn't satisfied with the original analysis of the data and that we would have to start again. You may think I would be frustrated by that but actually I was kind of excited to see the whole process and try to do it myself. So I began to "Code" the data, a type of qualatative data analysis which involved labeling what a participant said line by line. I was thrown into the deep end but I learned a lot. Seriously, a ton.
• Week Three: After I had coded data from two participants, I had to wait for the study runner to also code the same data and for us to have time to talk about how to sync up our differnt codes. So in the meantime I started the literature review for the tactile materials study. I learned about different tools to manage papers. I found out Mendeley is really bad. It crashed on me over and over again until it just refused to open. Once my partner got back me we went over the codes together and decided which ones we should get rid of and which ones should be used to code the rest of the interviews. Then I began to code all the rest of the interviews with the codes we had agreed upon. Every once in a while I would find something I thought needed a new code and I compiled a list to give my partner.
• Week Four: I spent it mainly coding. It started to get out of control. I had to many codes and I wasn't really sure what they all meant. I had to go back and really understand my codes but again I got it but by the end of the the week it was out of my control again. It's hard. I also participated in two interview studies. It was nice seeing other people's work and talking to people helped shake up my work and make my day more interesting. I started to categorize my codes. I realized that was were I had started before I had realized I had to code. I started to appreciate how much work it took to get to that part of the study.
• Week Five: I finally finished coding and continued to categorize the codes more seriously. Then I had to organize all the data by category. I started by copying and pasting each instance but I asked my mentor if there was a quicker way and she suggested writing a python script to automatically sort them. She helped me get started and I ended up with a nice script. It was very satisfying after all the work I had done running it and being done with categorizing in less than 5 minutes. Then I began to look at all the data in those categories, see if they were categorized well and begin to imagine the final paper. It was some mind bending work and I began to feel like I was going off in the wrong direction many times. I am honored to get to do it, but it feels a little beyond me. I went to a happy hour (don't worry no booze) in on honor of somebody who was leaving. It was very sweet i saw how close-knit this community really was.
• Week Six: The partner for my study got reviews back for the demo paper for his other project and I gave some critiques. I also helped him print out a 3D model to take a better picture to include in the paper. It was nice to learn a little more about how it works. I also talked to the people at the maker lab on campus and got invited to a session teaching how to use the vinyl cutter but unfortunately missed it. I also spent a while trying to take a picture of the model and I figured out a way to use a phone standing and a camera timer to get something I was proud of. I spent the week really refining the categories and understanding, analyzing and finding themes in the data. I started a real outline and then began to draft the paper. Progress is going slowly but steadily.
• Week Seven: I participated in cool study in which I got to be a in a pretend self driving in virtual reality. I also went to an event though WiTNY. I heard some amazing women in tech and I fot to connect with girls in tech my age. It also seemed promising for future internship since I got some contact info. I began to become pretty burnt out on the paper front. It was a battle up hill to get work done but started time blocking and got back into it, its very easy to lose sight of what you're really doing.
• Week Eight: I was really working hard to get a draft before I took a week off. Another person at the lab who had helped me understand how to code and categorize had finished a draft of what she was working on and I gave some feedback. It actually helped me in my writing because I saw an example of what my work could look like and helped me look at what I was doing from a different perspective.
• Week Nine: I got back from my week long break pretty refreshed. On Monday the lab had lunch at a nice place in Manhattan. We got to order a bunch of starters since the tab was covered by the institution, I think. A lot of people weren't there unfortunately but I got to meet two new people. It was really nice. I actually really finished the draft but it was still really pretty incomplete but I had to move on to get serious about the related work section. That's some of the hardest stuff. Very overwhelming to try to read all this stuff and figure out what is and isn't important. It's interesting that I feel most out of my depth here even those it probably is the most similar to work I have done in school. I also participated in a really interesting study my lab mate was doing about group dynamics and voice assistant and had lunch with the study leaders and other participant to give feedback. It seems now that everybody is back from the summer it's a lot more lively and social.
• Week Ten: A new master's student arrived and was assigned to work with me to finish the paper. It was really nice working with another person. I learned some collaberative skills but I also got to socialize and ask her more questions about grad school. We finished a draft of introduction, related work and abstract. I promised to stay in contact and help complete the last edits and the design implication section. The last day everybody who was in the lab that day had lunch together as a way of saying goodbye to me. It was really nice and I felt it was a good close to overall a good summer.

The End Result

I had to leave before the paper was finished but here is a link to the most updated draft I had a little after I left. I plan to stay in conctact and help finish the submission on the paper to CHI 2019.