Undergraduate Research Program

Joia Zhang


Mentor: Dr. Sat Gupta (Department of Mathematics and Statistics, University of North Carolina at Greensboro), Dr. Sadia Khalil (Department of Statistics, Lahore College for Women University)

Contact: joiaz@uw.edu

Current research project: Mitigating Respondent Lack of Trust in Quantitative RRT Models


Joia is a junior majoring in statistics at the University of Washington. Her current research focuses on improving respondent privacy and addressing respondent lack of trust in surveys containing sensitive questions, specifically in quantitative randomized response techniques (RRT). It’s the age of information, and it’s not only about what data is gathered, but how data is gathered. Joia’s research on quantitative RRT models centers around developing a survey technique that gathers data with both model efficiency and respondent privacy. Her other research includes applying topological data analysis techniques to investigate the relationship between political parties and economic factors in US states. She enjoys sharing her research at conferences and symposiums as well as advocating for undergraduate research for all undergraduates in all fields. On campus, you can find her as the vice president of the Statistics and Probability Association (SPA). Joia is from Ithaca, New York and loves to read and watch film adaptations of books in her spare time.

Translate your work so that we can all understand its importance
Surveys collect data that form the basis of decisions made at the individual, company, and federal level. Out of all mediums of surveying, face-to-face surveys yield the highest response rates. However, face-to-face surveys are susceptible to a phenomenon called social desirability bias (SDB). SDB is people’s tendency to give surveyors socially favorable answers, rather than true ones. As such, consequences of SDB include low response rate or worse, untruthful responding. A way to circumvent SDB is randomized response techniques (RRT). RRT models allow respondents to provide scrambled responses that prevent the surveyor from ever knowing any true answers, while the ability to estimate the sensitive trait is maintained. An issue that arises in RRT models, however, is: what if respondents don’t trust the RRT model? My research addresses this question of how to mitigate respondent lack of trust in an improved model that produces better data with higher respondent privacy.


When, how, and why did you get involved in undergraduate research?
When I came to UW, I had never done research, but I was fascinated by the idea of it and the creation of new knowledge. Initially, I was unsure when or how to start. I’ve learned that it’s never too early to start pursuing undergraduate research opportunities and that the search can be challenging but it builds your communication and self advocacy skills, and is very rewarding. As a sophomore, I applied to the Statistics and Probability Association Directed Reading Program (SPA-DRP) which is a quarterly program that pairs an undergraduate with a graduate student to delve into a topic and apply what is learned in a research project (more information and how to apply here https://spa-drp.github.io/). I had an enriching experience doing a project on topological data analysis with my mentor Jerry Wei, statistics Ph.D. student. I got the opportunity to present my findings at the Undergraduate Research Symposium organized by the Undergraduate Research Program. This was my first research experience and my first symposium. It inspired me to keep pursuing research and the following summer, I partook in my first full-time undergraduate research experience in an NSF funded REU program at the University of North Carolina at Greensboro (UNCG) on Complex Data Analysis using Statistical and Machine Learning Tools. This immersive summer program invited guest speakers, hosted workshops, and held panels on topics ranging from graduate school to coding languages to diversity in academia. I was able to research randomized response techniques with my mentors Dr. Sat Gupta (UNCG) and Dr. Sadia Khalil (Lahore College for Women University). I have enjoyed presenting my research in conferences and the manuscript is in progress. My REU experience has been rich, enriching, and inspiring. REUs accept freshmen, sophomores, juniors, and there are also programs for seniors in their summer after graduation. My REU application process began with going to the NSF website that lists programs by academic field (https://www.nsf.gov/crssprgm/reu/reu_search.jsp). I used a spreadsheet to keep track of the programs I applied to, so that I did not miss a step: contact references, confirmation of submitted reference letters, statement of purpose draft and final draft, additional application materials, and finally submission of the application.


What advice would you give a student who is considering getting involved in undergraduate research?
Join an RSO in your academic field of interest! For me, that was UW’s Statistics and Probability Association (SPA). You get to meet an amazing group of people who you share common interests with. You get immersed into the academic community, meeting fellow undergraduates, graduate students, and even faculty members. You also get immersed in the information loop. You join email lists, hear of important updates in the department, and you learn of opportunities—potentially undergraduate research opportunities. Joining an RSO in your field of interest is a great way to meet new people, bond over your shared interests, and gain footing in your discipline.