Please tell us briefly about your current research and teaching activities. What do you find most exciting about them?
I work on developing and implementing statistical methods to analyze data from medical and social-science applications. My current research incorporates various themes.
A part of my work focuses on an area called statistical social network analysis. Networks are often used to represent data in which the units of observation, for example users in a Facebook network, act in a dependent way. Research methods that have traditionally been developed for independent units cannot be used to analyze such data. I work on analyzing such social network data using modern statistical methods to identify communities in the network, understand what characteristics are associated with a higher chance of forming a connection, and so on.
I am also involved in research projects to predict medical outcomes in a big data setting using machine learning methods. When the number of predictors is large (potentially larger than the number of people we have data for) traditional methods often fail. New statistical and machine learning methods are developed to deal with such failures and give more accurate prediction result.
All of my work is essentially motivated by application specific scientific questions. Opportunities to learn about various topics by collaborating with many scientists, and to contribute in a meaningful way to new scientific breakthroughs, make me excited about my work. I also like that my field is quite interactive.
What are the recent advancements and breakthroughs coming up in the field of BioStatistics?
Fields such as statistics, machine-learning and bio-statistics are quite interrelated and a breakthrough in one is probably also a breakthrough in other. I will thus answer this question in a general way. There are many breakthroughs in recent time. Advances in predictive models, advances in theory for inference, and drastic improvement in computing power, all have enabled us to analyze massive amount of data relatively fast and with better prediction properties.
Tell us briefly about your career journey. What made you attracted and what keeps you interested in Statistics?
It was in Mount Holyoke College where I found my passion for numbers, particularly for research in statistics. I had caring and inspiring professors who never doubted our abilities. I was also exposed to many research experiences as an undergraduate, and that more or less paved my path to being an academic. I entered Mount Holyoke as an economics major and did a summer internship in rural India to study an impact of a micro-finance on rural women. During the internship, I instead got interested in collection of data and tools used to analyze that data, which basically is the foundation of statistical analysis. After that summer I started taking more Statistics related classes and decided to write my thesis in Statistics, to analyze brain signals in epileptic patients. That was a turning point. After Mount Holyoke, I went to Carnegie Mellon University to pursue a PhD in Statistics and then to Harvard for a post-graduate research fellowship.
What keeps me interested in Statistics? I like that I never stop learning in my field. I have to work hard to keep up with new methods, algorithms, theories, and applications, which I find intellectually stimulating. I often work on problems motivated by real world examples, and I find the direct impact of my work quite satisfying.
What is the difference between a Statistician and a Data Scientist?
Um, this is a highly debatable topic. Larry Wasserman, my professor in graduate school, used to joke that if it is done in Porter Hall (building where Statistics Department is) it is called Statistics, and if it is done in Gates Hall (Computer Science building) it is called Machine Learning. I suppose you could say something similar about Statistics and Data Science! 🙂
On a serious note, in my understanding, the definition of a data scientist is wider than that of a statistician. Statistician is a rather traditional profession with a focus on making inference about different aspects of data, understanding variation, and establishing guarantees about a bigger population from a smaller sample. (But then this idea of a statistician is also expanding with many new developments in computers and increasing data complexity.) While a statistician can certainly be considered a data scientist, I think there is more to the profession of data scientists. Computer scientist, data managers, algorithmist, machine learner, coders, etc are also data scientists. So data science as a field is broader than what is traditionally considered statistics, and someone who considers herself as a data scientist is not necessarily an expert in every aspect of it.
Why do you think study and understanding of statistics are important for everyone? Do you think statistics is emphasized enough in our education system?
Validating our intuition with data is very important. Hypothesis testing in statistics involves a lot of that. Predicting future outcomes using our current knowledge is helpful in making decisions. Scientific discovery also hinges on having meaningful evidence which is established using data and statistics. So, understanding statistics is important.
Having statistical knowledge is also helpful in day-to-day life. Be it understanding weather reports or changes in the stock market, knowledge of statistics is needed. When media reports on important scientific findings, they are usually exaggerated to make it sensational. A general knowledge in statistical science will help a lay person read the information conveyed through news critically.
I don’t think statistics is emphasized enough in secondary level education in Nepal. Unfortunately, I did not do my bachelor and doctorate education in Nepal, so I am not aware how much of statistics is taught at that level. Even in America, introductory level statistics classes are not the most fun ones. There is a lot of focus on learning formulas and the main motivation is usually lost in the process. A lot of people have told me that statistics was their least favorites class. And I think rightfully so.
Women are very underrepresented in the field of Mathematics. But, they are comparatively better represented in Statistics. Why do you think it is so? What are your experiences and observations so far? What would be your advice for other women who want to get into mathematics and statistics?
To be honest, I am not sure about the differences in the representation of women in mathematics and statistics. I was one of the very few women in my cohort during my PhD. I have been on many meetings where I am the only woman. But I have also seen it getting better. My current department has a good gender ratio and it feels great to be a part of such a diverse department.
I do think, women and people from racial minorities are underrepresented in STEM in general because traditionally people from a very specific race and gender have been gatekeepers of these fields. Unless that changes, it is hard to increase diversity.
My advice to young women: recognize your strength, and work hard to be better in it. If someone says you are not good enough but does not offer to help you get better, ignore them. Strong mathematical and computational foundation is essential in Statistics and Mathematics. Try to expose yourself to these ideas from early on. There are a lot of resources available online these days, so try to keep up with new methods and ideas. Finally, don’t be shy, ask for help when you need it.
Tell us about a time you made an exciting breakthrough — or any other highlight in your academic and career journey.
I am still waiting for a breakthrough. But my career has been a culmination of many small victories which I am equally happy about. 🙂
After tenth grade, I got in St. Xavier’s College for high school through a waiting list. Even though Mount Holyoke College (MHC) was my first choice for college in the US, I didn’t get in the first time I applied. I first came to Wesleyan College in Georgia and then transferred my second year by applying again. Both of these events, going to St. Xavier’s College and then to MHC, made an impact in my life and career. I barely made it through, but they had a huge role in making me who I am today. My undergraduate thesis was awarded highest honors and I graduated with Summa Cum Laude. I got into Carnegie Mellon University (CMU) for my PhD degree and then to Harvard for a post-doctoral fellowship. The many opportunities I had to study and receive training from among the best universities in the world on multiple scholarships are definitely the highlights of my career. They were all monumental in bringing me here to NYU.
Tell us about a time you had serious doubts about your own ability in your job. How did you overcome that?
My doubts come and go in cycles. Research is a very slow process that builds on incremental progress. I usually get very excited in the beginning when I start forming new questions. But by the time I start literature review and realize how little I know about the topic, I get anxious and doubtful about whether I will be able to get anywhere with the research. I choose to be persistent and keep going. When the project is complete and I get to write about it, I get ecstatic. So, for me, self-doubt has been a part of the research process. I do not let it stop or deter me from my goals, however.
One memorable example of self-doubt was during my graduate school. I had a major roadblock in my research, to the point that I got depressed, and I wanted to drop out and leave. But I was very lucky to have an amazing mentor who advised that I should stay for a semester and give it a chance; if it did not work out, I could leave. Five years later, here I am with a degree in Statistics!
I am conscious of my writing skills, time management skills, and efficiency. But I also consistently try to improve myself. What I have learned is that it is okay to be aware of our strengths and weaknesses. But we should not obsess on it and let it stop us from reaching our goals. If we don’t have certain skills, it is okay to reach out for resources to learn the skills. Success comes with persistence, patience, and hard work.
What qualities would you look for in a prospective graduate in your field?
In terms of technical skills, I would see whether the student has any training in mathematics and logic. At least a few courses in calculus and linear algebra is needed to be a good researcher in statistical methods. If the student is interested in more applied research, mathematical background is helpful but not always necessary. Experiences in computer science and writing technical papers are both very important and beneficial. Technical writing and public speaking are never really emphasized but they are important skills to have, especially because we work with people from different backgrounds so often. In terms of interpersonal skills, patience, persistence and curiosity are must. Not everyone comes in as a genius, but they should be willing and curious to learn and work hard.
Can you recommend 3 resources for people looking to get into your field ?
-Elements of statistical learning ( by Trevor Hastie, Robert Tibshirani, Jerome Friedman) (available for free online)
-Class notes by Cosma Shalizi and larry Wasserman (both are available online)
Tell us about the role of mentorship in your professional life.
It has been tremendous. I owe a lot of my academic career to my mentors. My parents are my first mentors. My mother, especially, has been my biggest mentor in life. She instilled values of hard work, persistence, and ethics in me from very young age.
My teachers from middle and high school in Nepal, professors at MHC, CMU and Harvard, all had influence in my learning. My PhD advisor Brian Junker at CMU gets a lot of credit for keeping me interested in academia and for that I am forever grateful. I also learned a lot about research and navigating academia successfully as a woman from my post-doc advisor Sharon-Lise Normand, a collaborator and mentor Sherri Rose from Harvard University, and a collaborator Tracy Sweet from University of Maryland.
What is the best career advice you have ever received?
Be confident, outspoken and demand respect, because you are smarter than you think you are.
The career advice you wished you received in your twenties.
Don’t take it too seriously, rest up and exercise regularly. (This is the advice I wish I had received in my late teens and early twenties.)
Please share your experiences on opportunities and responsibilities of being at the forefront of cutting edge research as a Nepali national.
My experience so far has been great. I feel extremely lucky to have received all these opportunities. I was always a good student but never the best. I did not always make it through on my first try. But somehow things have worked out. And I am optimistic that they will in the future as well.
Because I have received these amazing opportunities to study in world-class universities out of the generosity of many donors, I have a responsibility to give back to others who have limited resources. I want to encourage and help young people who are aspiring for a similar journey. I am also passionate about increasing gender and racial diversity in science, technology, engineering and mathematics field. I believe that it is crucial for everyone to make it their priority to encourage such diversity.
Final words of advice for the youth who want to pursue a career similar to yours.
Have perseverance, be curious, explore opportunities and resources, reach out for help, and of course work hard. There is no shortcut to success in life!