Tabitha Belshee, 22, Class of 2020
Tabitha Belshee majored in politics and earned two certificates: Statistics and Machine Learning from the Center for Statistics and Machine Learning (CSML) and American Studies.
Belshee’s independent project for her CSML certificate analyzed and explored the quantitative impacts of President Donald Trump’s Zero-Tolerance Policy, which went into force on April 6, 2018. The policy was meant to increase prosecution of people crossing the border into United States illegally; however, the policy is controversial because many people entering the country could have qualified for refugee status and according to numerous news reports, parents were being separated from their children, including breast-feeding infants.
With this backdrop, Belshee set out to gain “a deeper understanding of who, specifically, this policy harmed.”
“Thousands of kids were probably lost in the system and the system has no idea where they are,” said Belshee. “And so, I looked to see if there were any patterns in the children that were able to be reunited or not. Which ones were most likely to be separated?”
Some data was made available to her by the University of San Diego’s Center for Public Interest Law, Belshee said. This data had information on when the kids were separated, what country they were from, basic demographic information, whether they were reunited or not with their family.
In terms of conclusions, Belshee saw that children who were apprehended before the official policy date had longer detainment time. Belshee also looked at different border offices, places such as El Paso and Laredo, and saw that the policy was implemented inconsistently.
By modeling the data, she created a synthetic dataset that mimics the characteristics of the actual data. Belshee then developed a neural network model that sought to predict the likelihood of a child being reunited with their parents,. She contextualized her results using the profiles of three fictional children who were from Brazil, El Salvador and Guatemala.
“It was really important to me to make my independent work be accessible to everyone, not just people with a background in statistics and machine learning and be able to tie it back to a story that I could explain to my parents, for example,” said Belshee.
Looking back, Belshee said she found her CSMl classes helpful because they inculcated in her a deep appreciation for data.
“In politics, in general, anybody can claim that they're making informed choices, but if we ground them in unbiased data, it’s good insurance that we are making policies that make sense for people, and they don't have any nasty consequences,” she said.
After graduation, Belshee took up the position of college and career advisor/mentor at her former high school, Orange Glen High School in Escondido, California. Belshee is a first-generation college student and she said she wants to help students like herself. In the future, she plans on pursuing law school and going into politics.
Belshee was a residential college advisor for Wilson College, the head political statistics tutor at McGraw Center for Teaching and Learning, an officer at the Quadrangle Club, and she also participated in the Princeton Hidden Minority Council and the Scholars Institute Fellows Program. In her senior year, along with several students, she won the Spirit of Princeton Award, which recognizes undergraduates for positive contributions to campus life.
Belshee practices and teaches yoga, reads many books, and spends a lot of time with her Chihuahua named Babushka.