Meena Doshi is visiting faculty at the MIT School of Bioengineering Sciences & Research in Pune, India. It is the only bioengineering program in the state of Maharashtra.

It has been well over a decade that I have lived in the US and gone to school and worked here. Typically, I have visited my family in India once a year, over the Christmas/New Year holiday season. But I was on a project in Pune, India—the city in which I grew up—recently. After a long hiatus, I went to stay for an extended time and had the opportunity to teach elementary statistics to first-year economics graduate students at a well-known private university. The teaching opportunity had me wondering about how different the students, ways of teaching, and classrooms would be. I thought this was going to be fun. I was so excited!

Before I go on to write about my experience teaching elementary statistics in India and the US, I want to mention that I don’t know it all. I do not have years of experience in teaching like most professors do. I was a teaching assistant in the US for two years and conducted a lecture series in elementary statistics, as well as a workshop in India—that’s about all! I would like to share my impressions, opinions, and thoughts about teaching elementary statistics.

Oxford of the East – Pune

A bit about my undergrad days first. I did my undergraduate degree in Pune, India—also known as the Oxford of the East. Students from every corner of India flock to Pune for their studies due to the high concentration of the best engineering, medical, and architectural schools.

The teaching system in India is very different from western parts of the world. In my undergrad days, the professor would enter the classroom to deliver his lectures/concepts in a didactic manner, while the students would diligently take down notes and be attentive (or at least give that impression). It was almost as if we were not encouraged to question or think for ourselves. The lectures were a one-way process, if I may call it so, and many times—depending on the teacher—quite boring and difficult to sit through (with eyes open). All coursework was theoretical, not experiential. Also, there were no computers in class or at home then. So it was the old-school way of teaching—chalk and black board.

Now when I pause to reflect on those days, I think it was hard on the professors. We were given hypothetical examples, many times described from something that didn’t exist, arising from a situation that was hard for students to imagine, and drawn from something we found difficult to put together. You see how this might not have been the simplest mental somersault for professors or students. And yet, the love of the subject kept us both going.

Columbus, Ohio

While pursuing my first master’s degree at The Ohio State University, I was on a teaching scholarship and taught introductory statistics to undergraduate students for two years. The teaching fellowship in the first quarter gave me lessons on how to teach. I learned to explain confidence intervals making student’s sample 30 M&Ms from a bowl, counting the brown-colored M&Ms, and plugging the numbers into the formula. The applied approach to teaching statistics was so fun and made so much sense.

Honestly, it was in those years that my understanding of the basic concepts of statistics started to become clear as students would ask questions such as the following: What does the p-value really mean, and where is it used? What are confidence intervals? What is a normal distribution? I think it was the best thing that happened to me as a statistician. I loved it.

The classroom was a stimulating environment, where the learning process was made interesting by encouraging questioning and lively discussion. Class interaction was important and contributed to the students’ overall grade. It was also quite a casual environment, in which it was okay for students to have lunch or a snack. These two elements were unlike the stiff, formal, monotonic environment I was familiar with in India.

All classrooms had computers. I taught elementary statistics using Datadesk (point and click) as the software. It was simple and made it easy to visualize and get used to handling data on the computer. The students could play around with various features and get a better understanding—gaining a crucial ease with data.

Back in India After a Decade

Today, India is known to the world as an IT hub. Things have changed dramatically. We live in a world where the term “big data” is splashed across every media; there were huge billboard signs featuring big data, the cloud, and the latest iPhone along major roads. There were OLA and UBER cabs zipping by; almost every other person seemed to have a smart phone and use WhatsApp. I could go on about how my city seemed to have access to the latest technology. I told myself the scenario inside the university must have been revamped to a completely different level. Sure enough, the classrooms were made over with projectors, air conditioners, better lighting, and computer terminals for the teacher that projected onto a whiteboard visible to the whole class. And the students looked more westernized in so many ways. However, the mindset of the students was “same old, same old”! That was sad and hard to comprehend.

While teaching, I incorporated the “American” model, which allowed students to talk in class, ask questions, and think for themselves. I played interesting video clips (e.g., Hans Rosling explaining the concept of correlations), shared my experiences, and encouraged questions. I tried to set my voice for the class as “statistics is not boring and definitely nothing to be afraid of.” However, in a class of about 40 students, only about three or four looked like they were interested and asked maybe one question. This made me ponder, why do the students not question? Is it that most students are simply not interested in the subject, or is it the unchanged culture of not thinking for themselves and only accepting what is being told to them? How will this change? When? And who will initiate this change?

Few undergraduate schools in Pune include a statistical programming language in their curriculum. Even if they do, it is mostly R (an open source language) at an extremely basic level. The majority of schools do not have computers. Lack of funds is one major issue. Besides, there is a thought process among educators. I spoke to some of the faculty there and they said, in statistics, there are some educators who believe the mathematical underpinnings of statistics are sufficient to provide students with an intuitive understanding of the discipline. Because of this, they do not find it imperative for statistics education to be accompanied by computation. In this paradigm, students learn basic concepts about sampling, distributions, and variability and work through formulas by hand. At most, they use the computer to assist with their arithmetic calculations (statistical computation).

This argument has two roots. One is some statisticians do not believe computers are necessary to do statistics, and the other is that even if doing real statistics requires a computer, students do not need to use one when they are first starting out. However, this is an unfortunate line of reasoning because it prevents students from seeing real applications of statistics (i.e., the interesting stuff) until they have progressed to a real job.

Global Issues

From my perspective, statistics is perceived as a branch of math and has a bit of an image problem the world over. It is considered boring, difficult to understand, and scary. What remains jarringly universal is that few students are truly interested in the subject. Why?

First, this could be because the subject is not always taught in an appealing manner to draw students in, but instead presented in a complicated manner that can scare students away. I think the reason for this is two-fold: the attitude of the teacher and the method of teaching.

Teacher’s attitude: This sets the interest level to boost student performance and development in class and other areas, therefore elevating students’ overall confidence level. Other professionals such as architects, designers, and marketing managers are required to develop soft skills to become successful; so should teachers of statistics. I am aware the centers for teaching and learning do provide training in these areas, yet I want to mention the following two areas that need special attention:

1. Design: the organization/planning of lectures and slides with inclusion of images, videos, etc.
2. Storytelling: drawing examples from current news; understanding student backgrounds and encouraging class participation; and setting your “voice” and employing it with consistency, intelligence, and sometimes humor

Method of teaching: Reshuffling the method of teaching from time to time is both difficult and necessary. Why? Difficult because teachers value their autonomy, worry about their ever-increasing workload and time constraints, and shy away from risk and change. Necessary primarily because it will help get more students interested in the subject.

Most introductory statistics courses follow an order unchanged: begin with probability theory, basics of data and its types, and sampling strategies; then move to measures of central tendency and dispersion, correlation, distributions and their parameters, basics of inference, and inference of numerical and categorical data; and finally end with regression. If presented mathematically, that’s probably the right order because it depends on other concepts already introduced. But if we drop the math, we can adopt a different order—one that follows the gradual building of students’ intuition and deeper understanding. The math can be introduced after the intuition is in place.

Say, for example, we start with a problem; represent it physically; do experiments; represent the results as trees, two-way tables, or Venn diagrams to get conditional proportions; and then do probability. Teachers need to adapt their craft to the rapid-paced global traffic of information, changing students, and students’ needs.

Second, the lack of interest in statistics could be partly because few students are aware of this critical and growing field and all it has to offer. It’s probably one of the least well-known among the fastest-growing STEM (science, technology, engineering, and mathematics) careers. Statistical jobs are among the top 10 fastest-growing occupations and expected to continue to be through 2024, according to the Bureau of Labor Statistics. And fields from health and science to journalism, marketing, human rights, agriculture, actuarial science, and psychology (the list is not complete) increasingly require an understanding of statistics. Major pharma companies such as Pfizer, Quintiles, and Johnson and Johnson (to name a few), as well as CROs, are outsourcing their clinical trials to places like India and Europe primarily because the cost of conducting a trial is a lot cheaper there. More awareness needs to be created about the growing demand for statisticians and this lucrative career in undergraduate and graduate schools. Perhaps a fact sheet from the McKinsey Global Institute needs to be circulated throughout social media and on campus. The opportunities for a statistician are immense the world over. This is one field in which one can develop a career that straddles countries.

With advances in technology, the demand for statisticians increases. Therefore, we need good teachers, better methods of instruction, and an improved image to attract more students into this field.

What Are Educators Doing in the Statistics Community?

The International Conference on Teaching Statistics (ICOTS) started in 1982 by the International Statistical Institute (ISI) Education Committee. The conference is held every four years. After ICOTS-3, the International Association for Statistics Education (IASE) was created as a section of ISI and has since organized the conference. The last ICOTS was held in Kyoto, Japan, in 2018.

These conferences get the most experienced statistical educators around the world under one roof so they can exchange thoughts, explore new ideas, and be inspired. They are valuable events and offered under the assumption they can positively affect how faculty teach. However, I think these interventions don’t have much of an effect when it comes to sustained behavior change, unless it is actually implemented in the form of a program/curriculum.

For example, the simulation-based inference CATALYST course created by Andy Zieffler and colleagues at the University of Minnesota reshuffles the traditional order of teaching statistics. First, it introduces simulation from a model (marginal then conditional), then permutation tests, and then bootstrapping. This equates to distributions, then regression, then hypothesis tests, and then confidence intervals, but sets aside formal definitions from the early part. The guiding principal is to teach students how to cook, rather than follow recipes. The cooking method uses randomization and repeated sampling methods to make statistical inferences. Even though there were many challenges, the creators think they developed a course that engages students and stimulates them to think, build and test models, and understand the core ideas of statistical inference. I am sure this makes perfect sense to teachers who have had years to think about it.

Another example is the Numeracy Infusion Course for Higher Education (NICHE) project, created by Esther Isabelle Wilder and colleagues at the City University of New York. The motivation behind NICHE was educators without a statistics background who have to teach statistics and feel uncomfortable about it. This is a big and real problem. The project provides a safe environment for these teachers to boost their statistics skills and share ideas. Pre- and post-testing among the participants showed great improvement in their comprehension. The project has become so popular, the creators now have to turn people away each summer.

The Guidelines for Assessment and Instruction in Statistics Education (GAISE) were developed in 2005 by the American Statistical Association. These guidelines provide a framework for statistics education toward the end of enabling students to achieve statistical literacy, both in their personal lives and careers.

Three important Indian institutes are the Indian Statistical Institute (ISI, founded in 1931 by famous statistician P. C. Mahalanobis) in Calcutta (Kolkata); the Advanced Institute of Mathematics, Statistics, and Computer Science (AIMSCS, founded in 2007 by well-known statistician C. R Rao) in Hyderabad; and the International Indian Statistical Association (IISA, founded in 1991) in Washington, DC.

In spite of the large and growing talent pool of both experienced and young statisticians from India and conferences, seminars, and workshops conducted by these three institutes to rethink new ways to teach statistics, the status quo among teachers and the methods of teaching is disheartening. Somewhere, someone has to initiate change by creating actual programs like CATALYST.

Starting Early

Tightening the focus on statistics in the early grades would give teachers room to help their students think more critically about the subject, rather than relegating it to the last two weeks of the year—if the teacher has time. Sprinkling in basic statistics among mathematics or science would help teachers build a better foundation from a younger age, hence making their students globally competitive in the years to come. Gradually, this change is being introduced in elementary grades in western parts of the world. Hopefully, the east will catch up soon.

Classroom Under a Tree

In the news recently was a poem by Rabindranath Tagore, “Where the Mind Is Without Fear,” cited by Martin Sheen. Tagore was a reformist educator who wrote the national anthem for India and became the first Asian to win a Nobel Prize in literature. It is truly amazing that the poem, which was originally written in Bengali by Tagore in 1910, is relevant today across the globe. Listening to Sheen’s speech, I found myself reflecting on Tagore and his Shantiniketan—a unique learning campus near Kolkata known as the “classroom under a tree.” The idea is that true learning does not begin or end inside an enclosed classroom. A connection with the natural, outside, “real” world is continually fostered in an environment such as Tagore’s classroom. With much more added to broaden and deepen the knowledge base, learners leave such a classroom only to keep growing. Maybe such a classroom needs to be created for statistics students.