Sarah Morton’s Journey into Data Science and a New Career
The journey through earning a PhD in sociology is different for every student. Observing the educational path followed by one of our newest doctoral graduates, Sarah Morton (PhD, 2019), reminds us about the changes occurring in the knowledge needs and interdisciplinary skills that are becoming increasingly important for applying sociological learnings in the practical world of work.
We asked Dr. Morton to describe her path through graduate school and how she managed to link her sociological interests with statistical perspectives critical to working in a field now described simply as “data science.”
By Sarah Morton (PhD, 2019)
I did not know what data science was until a few years ago, but it has changed my life and career forever. During my third year in the Sociology PhD program, I started a master’s in statistics, which fueled my interest in data science. At that time, I was also thinking about my sociology dissertation, and decided to study gender and STEM (science, technology, engineering, and mathematics) education using data science and social network analysis. This offered the benefit of being an interesting project that would help me fulfill the dissertation requirements and build employable skills along the way.
To measure organizational culture, I used texts from each program’s “about us” or equivalent webpages. I scraped information from the web via a text miner I built using Python programming. I used machine learning to classify programs as having masculine, feminine, or gender-neutral culture based on a set of criteria taken from the gender and STEM literature. I then used social network analysis to visualize the texts and built regression models to test my hypotheses. I gathered other data sources using the Integrated Postsecondary Education Data System (IPEDS) and the IPEDS Completions Survey available through the National Science Foundation to complete these analyses.
To complete my dissertation, The Gendered Substructure of STEM: A Quantitative Analysis of Organizational Culture, Organizing Processes, and the Proportion of Female Graduates in Six Disciplines, I had to put to the test all of the data management, programming, and statistics skills I learned. My goal was to see if there were organizational cultural differences across six STEM disciplines (biology, chemistry, computer science, mathematics, physics, and psychology). I wanted to find out if organizational culture and the interdisciplinary structures of STEM programs were associated with the proportion of female bachelor’s graduates.
Several interesting findings emerged, some of which I did not expect. As expected, I found that biology and psychology—two female dominated disciplines at the undergraduate level—were more likely than computer science to be classified as having feminine culture relative to masculine culture. However, I was not able to link the organizational culture of a program with the proportion of female bachelor’s graduates for the most part, and the significant findings I did have were the opposite of what I hypothesized. For example, in a few of the disciplines I found that the higher the score was on feminine aspects of culture (e.g., science that helps society), the lower the proportion of female graduates was. On par with my hypotheses, I found that certain interdisciplinary department structures in computer science, chemistry, and psychology (e.g., a department of mathematics and computer science) were associated with higher proportions of female graduates than single-disciplinary departments (e.g., a department of computer science only).
My findings bring diversity issues in academic STEM to light. Since many of my significant findings regarding culture and the proportion of female graduates were the opposite of what I hypothesized, it is possible that universities are “window-dressing” their publicly available texts to sound more female friendly than they actually are. On a more positive note, my findings support encouraging departments to have more interdisciplinary components, inasmuch as departments with those structures tended to have higher proportions of female graduates than single disciplinary structures.
My dissertation helped jump start my own career as a woman in STEM. In late November 2018, I applied to work as a data scientist at Engie Insight, a sustainability and energy management company in Spokane, WA. As part of the onsite interview, I gave a presentation on the data science project I had worked on, using a portion of my dissertation research.
Less than a week before I defended my dissertation, I accepted Engie Insights’ offer of employment as a data scientist. I owe a lot of this opportunity to my research experiences in sociology. While it is a seemingly unconventional path, there is definitely room for sociology in data science, especially considering how sociology has very interesting data and scientific problems to solve. As I start my career, I am excited to see where both the energy industry and data science take me. And, I hope to help others break into data science, especially those from underrepresented groups and unconventional backgrounds, to increase diversity in this exciting and growing field.