Zeynep Tuna (she/her/hers), 2021 APS Education & Diversity Intern

Biography

SPS Chapter: Bowdoin College

I am currently a junior at Bowdoin College, pursuing a double major in Physics and Computer Science. I am an international student from Istanbul, Turkey and I have attended the United World College in Mostar, Bosnia and Herzegovina, where I had the opportunity to live and learn with peers from 60 different countries.

Even though I have always wanted to study physics, the close-knit physics community at my college was definitely a deciding factor. At Bowdoin, I co-lead Women in Physics, Girls Who Code, and the International Student Association. Throughout my life, I have been lucky to be a part of many amazing groups where I collaborated with exceptional people on community outreach and mentorship programs - encouraging more diversity and inclusion in STEM is my passion! I have no doubt that my time working with APS this summer will be as exciting and rewarding.

Internship

Host: American Physical Society

Project

Abstract

In light of a global pandemic, the Black Lives Matter movement, and the efforts to stop Asian and Pacific Islander violence, it is more important now than ever to focus on making our physics community more diverse and inclusive. In order to succeed in this mission, we must better understand the rapidly-changing demographics of the field of physics. During my internship at APS, I had the opportunity to analyze the national physics education statistics based on race, ethnicity, and gender, observe improvements made in diversity and inclusion in physics year by year, and realize the further changes and improvements we need to make. This presentation will focus on how diversity and inclusion data for physics degrees is obtained, categorized, and presented so that the institutions and the community can make active decisions to reach racial and gender parity in the field of physics.

Final Presentation

Internship Blog

Week 1: Decisions, decisions, and more decisions!

My first week as an SPS intern began with a very fun 2-hour virtual orientation. I know, 2 hours on Zoom does not sound enjoyable at all - but believe me I could have stayed for another hour to talk more about bucket hats and listen to staff members’ stories. The rest of the day was filled with back-to-back meetings where I got to set up my laptop (shoutout to Chuck!) and meet some very cool people in the Programs department. Even though there was a lot of information to process and many new faces to remember, the APS staff made these meetings incredibly comfortable and easy-going for me!

The next day, we delved right into the Education and Diversity Statistics project. Aside from the fact that many people use the APS statistics and graphs on race, gender, and ethnicity in the physics programs across the country, this project is especially important to me for a reason: I am a female international student studying and planning on pursuing a career in physics. I haven’t yet had the time to sit down and study the numbers to get some insights on how diverse and inclusive physics programs are in postsecondary institutions in the U.S., but I am very excited to do so over the course of my internship.

Upon meeting my supervisor Michael, I was very happy to see how much we had in common, and it was great talking about Maine, Europe, and being a college radio station DJ. Then came the decisions we had to make, and I am happy I was not making these decisions alone. We talked about which major milestones we should set over the course of these 10 weeks and which products should we aim to hand in at the end. This was the easy part as my supervisor and I agreed easily on the priorities for this project.

The first milestone to accomplish is deciding whether to use the Integrated Postsecondary Education Data System (IPEDS) or the National Science Foundation (NSF) database for the project. Although IPEDS has been used for a couple of years now, NSF has a more user-friendly interface which could save me some time in writing the Python code to produce the statistics and graphs. Upon completing a comparison study of the 2019 IPEDS data vs. the 2019 NSF data, I was not happy to find out that the data did not match! NSF still seems to be more future-proof and has more data points; however, it still looks like there is some uncertainty in this decision for me.

It was for the entire spring semester that my professor reiterated the importance of making decisions in data science in my Big Data in Astronomy class, but it wasn’t until now that I realized how right he actually is. Another thing he taught us was that there is no right or wrong decision, rather there is reasons and consequences. Now that I’ve got my reasons, I’m looking forward to seeing the consequences of all these decisions!

Week 2: Data retrieval is harder than it looks...

This week, I was delighted to realize that I had made a mistake in choosing a certain criterion during my data retrieval process and the IPEDS and NSF data actually matched. I don’t think I have ever been this happy to have made an error! After that, I started documenting the reasons for choosing the NSF database and how to retrieve data correctly. Who knew building tables were this hard? You would think it is just a few checks and clicks here and there, but not really. Anyway, there is nothing that beats the feeling of perfectly formatted Excel sheets of data.

In the middle of the week, my supervisor and I ran into an interesting issue. California State University – East Bay is not included on the APS Compare Your Institution webpage even though it is included in the spreadsheets! We quickly learned that it is because the institution changed its name from Hayward to East Bay (thank you Bardia!). Phew, so happy to be able to get back to the department chair of CSU East Bay with good news.

Work aside, I have had a very social week. It was amazing being at the Birthday Bash, seeing everyone together, and introducing myself. I now feel like I am officially working at APS as an intern. I also enjoyed the Game Day a lot and I even won a gift card in the raffle! (Shoutout to Shana!) I have had a couple of Zoom meetings where I got to meet some amazing staff members, and the SPS Happy Hour was a great ending to this week. Scribbl.io is fun, but it is 10x more fun when it’s physics words!

Week 3: Too many for loops mean too big of a Big O

Switching to the NSF database from the IPEDS database was a difficult decision, and naturally, it gave rise to more decisions. Deciding on how to construct the data tables required lots of experimentation and code review. I realized that I had to carefully choose how to build the data tables so that parsing the data would not take long time computationally. While deciding on the format, I tried many different layouts for the table. I have had to download, upload, re-download and re-upload. After many trials, I think I have found the final format I can follow to make the parsing more efficient! Of course, I have had to reflect all of these changes to the documentation.

It also took me a while to follow and understand the prior code to create the Uberspreadsheet. I found myself lost many times within the many lines of code on my dark VSCode IDE. I think I finally started understanding certain functions. Then, I started planning how I could modify this code to fit my needs when parsing NSF data. Turns out this will be harder than parsing IPEDS data due to the way NSF formulated their “Build a table” property. The grand totals for each race and ethnicity are actually not included in the table and including those would mean losing the grand totals for females and males. Yet again, another decision to make arises!

Another problem was the fact that some of the software and the packages used in creating the program for the Uberspreadsheet in 2019 are not available publicly anymore. Thankfully, I managed to choose a different environment manager and Excel package easily. With this, I will be trying to use dictionaries to my advantage to parse this data efficiently – fingers crossed!

Lastly, we have been receiving messages from more and more institutions about their place on the Compare Your Institution webpage. It seems like we have a lot to update, but that means we have a lot to learn as well. As this “Light meetings week” comes to an end, I wish everyone a good time celebrating/observing Juneteenth!

Week 4: It’s taking a bit too long!

This week, I spent most of time on Excel formatting data for the Compare Your Institution webpage on the APS website. In my opinion, this is an important aspect of the Education and Diversity Statistics project as most institutions and most people use this as a resource very often. Even just while I was formatting data, I could see the huge difference between the level of diversity in 2015 physics programs and the level of diversity in physics programs now. Seeing this improvement has made me happy, nonetheless we still have a long way to go.

I spent the rest of my time buried in Visual Studio writing Python code. As I was doing so, I realized that some of the Excel files I am trying to read are very long. Naturally, my code is taking a while to run – frankly, I have not been patient enough to even wait for it to finish. I think either a change in my code or the data format might be necessary.

An exciting event of this week was meeting the Nobel Prize winner Dr. John Mather, who works as a senior scientist in the NASA James Webb Space Telescope project. It was an honor to meet an incredible science figure as an SPS intern and have the chance to listen to his journey and scientific projects firsthand and ask questions. I am looking forward to the SPS virtual picnic as well!

Week 5: Life is full of surprises!

At the end of last week, I was quite worried about my process of modifying and debugging the code that adds the new year’s data into the Überspreadsheet. At the end of this week, I am at a completely different place: the code has been modified, debugged, and complete; the final Überspreadsheet has been generated; and the documentation is ready! It turns out that the solution to my code taking a long time to run was quite easy: use a read-only mode for speed optimization, load one workbook into memory at a time and close it before moving onto the next one (big shoutout to Arthur Smith from the IT department for his help!). The rest of the debugging process went very smoothly: lots of breakpoints and prints. Life is truly full of surprises – or at least, I still cannot be sure when my code will be complete and ready.

As I was talking to my supervisor about the final Überspreadsheet, we realized how far we still have to improve the diversity and inclusion in our physics programs. The total number of “Historically black colleges or universities”, “Black-serving institutions”, and “Hispanic-serving institutions” were upsetting. Also, realizing that the inclusion of the “Two or more races” category into the “Underrepresented minorities” definition made a greater difference in the 2018-2019 numbers than 2015-2016 numbers showed me how much we have improved; seeing this gave me hope for the future!

As this week comes to an end, I am ready to rest and recharge to start working on the graphs we all see and use on the APS website – exciting! I have also been having fun brainstorming ideas about my contribution to the SPS Demo Competition and I cannot wait to film it. Happy 4^th of July to everyone!

Week 6: Who’s playing in the Top Educators league?

This week, I started working on the Top Educators lists. In order to create the Top Educators lists for each demographic, I need to use the Überspreadsheet to create a separate sheet with all the institution’s 3-year averages for degrees awarded in physics. At first, I started calculating all the averages in the Excel sheet by hand. After a while, I realized this was going to take way too long! So, I decided that I could write a program to read in the Überspreadsheet, calculate all the averages and write it into an Excel sheet.

As I started writing the code, I ran into some problems – as expected. This week was mostly spent on Visual Studio Code debugging and printing to figure out how to make the code work. On another note, as I was studying the COE recommendations for Top Educator lists and the instructions on how to create the tables, I realized certain things. Firstly, when creating the Top Educators list for underrepresented minorities, we need to sum up the values for the last 3 years as we do not have enough/big enough numbers to take averages. Underrepresented minorities include African Americans, Hispanic/Latino Americans, American Indian/Alaska Natives and Native Hawaiians/Pacific Islanders and students who self-identify as multiracial members of two or more of the demographic categories who are U.S. Citizens or Permanent Residents.

Moreover, some of our Top Educators lists will be empty, which means they just will not exist. This is the case for bachelor’s degrees in physics awarded to URM women, master’s degrees in physics awarded to women of color and URM women. The pattern we see here is that women and minority groups receive significantly less degrees than other demographics. This is not a surprising fact; however, it is a concerning one that we should improve our educational system on. Hopefully, I will have more information on these trends next week to share with you all!

Week 7: Sometimes numbers speak louder than words

This week, I continued working on the Top Educators Lists. I completed writing the program that generates the spreadsheet with each institution’s 3-year averages based on the demographic/degree categories decided (yay!). As I started creating the Top Educators tables and considering the suggested rules by the APS Committee on Education, it struck me how the decisions we make on how to design these tables and, more generally, how to present the diversity and inclusion data for physics degrees that we work with affects the progress the institutions and the community can make. It was great to study an extensive document on the suggested rules for the How Does Your Institution Compare and the Top Educators webpage because it shed light on issues and delicacies that I was not immediately aware of.

As I was creating the tables, it was important for me to see and, now share with you all, that the tables with the emptiest cells belonged to physics degrees awarded to underrepresented minorities and women. Most of us were and are aware of this disparity, however, seeing it in numbers made it even more apparent. As the latest year of data that I have access to is 2019, my observations do not reflect the status of diversity in physics right now. I am definitely looking forward to seeing how things have changed in the years 2020 and 2021!

Week 8

This week, I finalized the Top Educators Lists and the documentation for the process. As I was starting to work on graphs, something was brought to my attention. Our How Does Your Institution Compare data for 2018 did not include second degrees in physics although the Uberspreadsheet did. This made me realize that it would be beneficial to think about synchronizing the How Does Your Institution Compare data with the Uberspreadsheet.

Moreover, we realized that the data we obtained from NSF also did not include second degrees, an intentional choice made by the NSF Data Center – tradeoffs in software design and data infrastructure can lead to problems like this. Having switched to NSF and worked on all the projects based on that specific database, this made me quite upset. The rest of my time at APS will be spent on updating the documentation to note this issue, brainstorming about what the future holds for the decision of IPEDS vs. NSF, and evaluating what could be published on the APS website this summer. No matter what, I still learned a lot during my internship, and I am thankful to Galen Pickett from California State University - Long Beach for bringing this issue to our attention. It could have gone unnoticed, and I am glad we get to fix it!

Week 9: Finalizing my work at APS

This week, I spent all my time on revising the work I have been doing throughout this internship so that the data that gets published on the APS website is accurate. I started by making a note explaining the issue of NSF data not including second majors in their numbers to the NSF documentation I had created. Then, I downloaded all the necessary 2019 data files from IPEDS and revised the How Does Your Institution Compare spreadsheets for 2018 and 2019 with this data. After that, I revised my code to read in IPEDS data files and add the new year’s data to the Überspreadsheet. From the updated Überspreadsheet, I generated the Top Educators spreadsheet with 3-year averages. Thankfully, I did not have to revise my code for this as the Überspreadsheet format remains the same. Finally, I revised the Top Educator Lists so that the ordering of the institutions is accurate after the corrected data.

I am happy to have fixed the issue and that my work will get published on the APS website after all. It is very important to our physics community that institutions can use the Compare Your Institution and Top Educators web pages to know where they stand and what they need to do to improve diversity and inclusion in their physics education programs. I am now focusing on my presentation and finalizing my work at APS. I am excited to present and talk about this 10-week journey, where I learned a lot!

Week 10: The Bittersweet End

This week, I spent my time finalizing my work at APS and getting the products of my work published on the APS website. Throughout my internship, I made a lot of tough decisions, experienced the implications and consequences of those decision, and learned a lot. It was awesome to see the results of my efforts go up on the APS website – it felt really rewarding! Of course, I am also happy that the education and diversity data is presented in a consistent and effective manner and the results are available for the public to see and use.

I was also working on my presentation for the 2021 SPS Intern Symposium. It was amazing to listen to all SPS interns’ work and journeys throughout these 10 weeks and share my story with the public. I am grateful to have had this opportunity to contribute to APS’ work in diversity in physics education. It was great to work with all my supervisors and get to know the rest of the interns. I definitely had a wonderful experience throughout my SPS internship!