Now that I’m ten weeks deep in a fifteen-week Data Science program, there’s still a subject that weighs heavily on my mind. From the very start of the program, all language and educational examples have almost always been binary. This makes sense, given that computers themselves are binary machines. But Data Science specifically deals with real-life human data and problems. So, it should be able to adapt to the evolving identities of the people it’s about. Additionally, this isn’t a critique of the program that I’m in; this is a widespread problem across industries.
As a disclaimer, I’m writing this as an aspiring Data Scientist. So, some aspects may be beyond the scope of my current knowledge. Also, I am writing from the perspective of an ally; I do not identify as non-binary.
Additionally, I would like to say a huge "Thank You" to my first coach in the program, Joél Collins. They inspired me to not only stick with the rigorous bootcamp, but also investigate deeper into every issue I encounter. They also generously provided personal experiences for this blog post.
Initially, I wanted to investigate the general concept of "non-binary" in Data Science. But a conversation with them unlocked a whole range of insights that I believe are imperative to share. Therefore, this post will primarily discuss the experiences of non-binary people, as a whole.
Understanding What Non-Binary Means
"Most people don’t understand what it means to be non-binary because they haven’t been exposed to it."
For anyone who may not know, the term "non-binary" describes individuals who experience a gender identity that is neither exclusively male or female, or even identify between both genders. Non-binary people may identify as genderqueer, gender fluid, agender (without a gender), or something else entirely.
The antiquated normalization of the gender binary is problematic because a person’s biological sex does not determine their gender identity. One way to combat this is with the normalization of gender-neutral pronouns, like they/them. Although it’s difficult for cisgender people to get used to this language, it’s critical that we use inclusive pronouns for those who exist outside of the normative binary.
Non-Binary Experience in Tech
It’s no secret that tech was once one of the most discriminatory fields, not only for women and people of color; but also for non-binary people. There are countless records of alleged workplace discrimination. In recent years, there have been increasing efforts to promote non-binary Inclusivity. However, Joél described that non-binary people currently operate in this limbo where they are respected, but people just don’t know how to be respectful. This is because a majority of people have never been exposed to the concept of "non-binary."
"The inclusion of non-binary identity is typically only in company circles, but not discussed in the public sphere."

Because of the lack of educational materials available on non-binary identities, people end up asking the first non-binary person they meet for more information. For Joél specifically, most people who inquire about their gender identity or pronouns usually accidentally approach it in a rude way. Even if it comes from a well-meaning place. However, they described that they face an internal struggle because despite this uncomfortable feeling, "at least people are asking questions."
Revolutionary Workplace Protection
For Joél, the most impactful moment in their career was on Monday, June 15, 2020. But for all of their coworkers, it was just another day at work. On that day, the Supreme Court ruled that the 1964 Civil Rights Act protects gay, lesbian and transgender employees from discrimination based on sex. This monumental ruling lifted a huge weight off their chest.
"This means that by federal law, I can no longer be fired solely based on my gender identity."
Although their current workplace consistently makes efforts to promote inclusivity, Joél explained that "you never really know what could happen." This decision was life changing for them, but no one else really noticed. And this representative of the overall experience of being a non-binary person. They expressed that because not that many people live through this experience, it can be very isolating. And despite this new protection, there is still an obvious underrepresentation of non-binary people in tech; and all other industries.
Binary Data Science
The next step is to examine why Data Science is such a binary field, and the answer should be obvious. Pretty much everything we do in this field is binary. From the way that computers store data, to how we write our code. In the Binary system, 0s and 1s are represented by "Off" and "On." This system is best suited to the optical and magnetic storage components to a computer.

From the very beginning of my data science education, I noticed that every example is binary. For example, while learning Pandas, the simplest concepts use "Male" vs. "Female." However, it’s important to acknowledge that things exist outside of the binary, especially when teaching newcomers. Ultimately, the binary is harmful for people who live outside of it.
Machine Learning and the Human Experience
So far on my data science journey, I’ve only learned binary classification models. Such as K Nearest Neighbors, Logistic Regression, Decision Trees and Random Forest. Although some of these can be used in multiclass situations, I’m curious to see what other multilabel classification models are out there. And whether they are capable of "predicting" or "understanding" non-binary characteristics. It’s worth noting that this subject is obviously outside of my scope of knowledge right now, but it’s something that I want to keep in mind as I learn more about the field.
The problem is that multiclass classification makes the assumption that each sample is assigned to one and only one label. For example, a fruit can be either an apple or a pear, but not both at the same time. Although these multiclass models deal with more than two classifications, the aspect of binary categorization still remains. And we can see how that could be problematic towards classifying human data because: what if someone belongs to more than one category?
I hypothesize that machine learning will never be able to truly capture the spectrum that is the human experience. For example, I don’t think a machine learning model will ever be able to fully predict a non-binary person’s gender identity, no matter how specific the predictors are. In this hypothetical scenario, we would build the model by fitting it to features of what we deem are traditional identifiers for each gender. This comes with obvious bias towards the antiquated gender binary.
Because of the class imbalance that comes with any underrepresented demographic, the model would appear to have a high accuracy score. But if we were able to somehow remedy that imbalance, the model would still be terrible at predicting every form of gender.
This dilemma is troubling because Data Science specifically handles human problems. So, if it can’t adapt to nuances like gender identity, then there is a clear limitation on progress.
Towards the Future
As I mentioned, Data Science obviously deals with real-world data and real-world problems. Therefore, it needs to be able to robustly adapt to evolving norms. Before this can even happen, overall society needs to drastically increase inclusivity for the non-binary community. According to Joél, the first thing that needs to happen is that people need to be familiarized with the concept. Destigmatizing or demystifying the concept of "non-binary" can promote monumental progress on its own.
Near the end of our conversation, we came to the conclusion that there has always been a cycle of fighting for inclusivity among marginalized communities. A few decades ago, it was the gay and queer community fighting for equality and rights. Looking at the progress that has been made today gives hope to the non-binary community that they can someday achieve the same equality in the near future.
A small way that readers of this article can contribute to this movement is to announce your pronouns. Even if you’re cisgender, it’s imperative to create a space for the discussion of pronouns. And if someone hasn’t announced their own pronouns, the default should always be they/them. Even if this consideration seems small or arbitrary for you, it could be major for someone in the non-binary community.
As a closing statement, I want to reiterate that I’ve written this blog from the perspective of an ally with the generous help of my previous coach, Joél Collins. If you identify as non-binary and feel that I’ve misrepresented anything, please reach out! And if you don’t identify as non-binary, I hope that you learned a little more about the community.