To be honest, I was very nervous. A few panicked thoughts started to flood my mind as I prepared myself to enter a conference of 500 or so young female hackers and technology enthusiasts.

Would I be seen as an outsider?
Would people be disappointed to find out their mentor is not a woman?
Would people even listen to me during discussions?
Would any of my words hold credibility?
Would I be judged as being ignorant or unaware?
Would all of the men’s bathrooms be temporarily converted to women’s bathrooms?

Earlier this month I had the privilege of attending WECode (Women Engineers Code) — the largest student-run conference for women in computer science conference — held at Harvard University. I was invited through my company as a mentor, along with my Senior Director and the University Relations Tech Recruiter — both of whom are female.

I was confused about being invited — I was both male and a recent college grad — and a bit nervous about attending. But I ended up gaining an invaluable perspective on inclusivity in the tech industry that I would not have experienced at a ‘normal’ conference. And by “normal”, I mean the usual men-dominated gatherings of technologists.

Learning to Listen

One important thing, if not the most important thing, I learned in this conference was the value of listening.

When I entered different workshops and talks, I usually gravitated towards the back. I was also hesitant to speak up to ask questions. However, during this time of self-conscious sitting, I learned a whole lot about women technologist’s experiences, opinions, and plans for the future. One notable speaker was even setting her stage as being a woman engineer running for Congress.

I heard stories of conquered obstacles, society-changing successes, and powerfully inspirational women. Surprisingly, by the end of the conference, I had seemed to develop a voice as well. I was ready to ask questions, and ready to offer my own contributions.

Being an Ally

Prior to the conference, I didn’t understand what it meant to be an ally. And I’m still not sure I know exactly what that means.

What I can say, at least for now, is that the best I can offer is to be an informed colleague — a friend who listens to the experiences of my colleagues and supports them.

Taking Part in the Solution

One thing that became clear during our discussions is that it takes everyone to solve an issue.

Women can’t do it alone. That would only result in an isolated solution. Men definitely can’t do it alone. It will take communication and hard work from both sides to figure out the inclusivity issues that plague the tech industry.

What Now

There’s still much to learn about the obstacles women in tech face. But now at least I’m equipped with some degree of insight into what’s happening in the workplace.

Will I go to a women’s tech conference again? If I’m invited again, yes. Will the questions of insecurity I held before still float in my mind? Probably. But doing my part to make tech a more inclusive field is much more important than my own insecurities.

The legends of mathematics that almost never were

Math on a chalk board in Afghanistan

Mathematical genius is misunderstood. The term “genius” evokes feelings of envy. It paints a picture of a pre-determined minority of people who have truly mastered a discipline. A chosen few.

Yet there is nothing pre-determined about it. Mathematical genius resides within every one of us. Most people just don’t know it yet. That’s because genius is fragile. If you don’t embrace genius and tend to it with care, it will slip away, leaving behind just a subdued vision of the mathematicians we could have become.

If you take even a cursory look through the annals of mathematics, you’ll see the fragile and often fickle nature of genius. Many of the legends of the field — whose talents and achievements are beyond dispute — could so easily have been forgotten by history. Their stories should act as a warning sign to educators today.

From last in class to mathematical folklore

Jacques Hadamard was a French mathematician with a knack for number theory. In proving the Prime Number Theorem (loosely, showing that primes become less common through the number line), Hadamard established his place in mathematical folklore.

How surprising then that, prior to fifth grade, he was placed last in his class for arithmetic. Hardly the mark of a child prodigy. You wonder by what alchemy he was reduced to the sobering status as ‘last in class’ — most likely a test, of the same narrow kind that pervades school mathematics today.

Jacques Hadamard — no longer last in his class

Tests can serve a profoundly useful purpose in providing feedback and diagnosing students’ learning needs. But as blunt labels for students’ abilities, they can be devastating. Hadamard prevailed, but how many other primary-aged students give up for life in the face of these cruel, snapshot judgements?

Ditching the garden for Linear Algebra

The stated hopes and beliefs of our parents can hold the key to unlocking our genius — or stifling it. Hermann Grassmann, one of the founding fathers of Linear Algebra, did not excel in his formative school years. Grassman’s own father — also one of the school teachers — advised his son to give up mathematics altogether. In his father’s eyes, Grassmann’s calling lay as a gardener. With all respect paid to my botanically-inclined friends, Grassmann did well to honor his intellect.

Hermann Grassmann — gardening's loss was the gain of mathematics

Women make the best mathematicians

Sophie Germain faced resistance of another kind. This precocious teen discovered mathematics in the stack of books that filled her father’s home library. Germain’s commitment was nothing short of incredible. She even taught herself Latin and Greek to access the full range of texts.

Germain faced just one problem — this was eighteenth century Europe, where it was uncouth for women to participate in mathematics.

Germain’s parents went as far as banning her from studying. They even removed her light and clothes to dissuade her from reading books.

She persevered, reading the likes of Newton and Euler into the late hours of the night, wrapped under a quilt, with stolen candles her only flicker of light. After finding her slumped over her books one morning following an all-nighter, Germain’s parents finally relented. Good job too — her contributions to number theory proved pivotal.

Sophie Germain — persistence personified

Germain worked under the pseudonym M. LeBlanc to disguise her identity. Her talents put her in correspondence with the Prince of Mathematicians, Carl Friedrich Gauss, who later expressed admiration and astonishment when discovering Germain’s true gender.

The most sobering aspect of Sophie Germain’s story is that it could just as easily play out today. The gender gap in mathematics remains unresolved, with girls falling behind their male peers in early schooling; a trend which carries through to the underrepresentation of women in STEM degrees and careers.

The gender gap is cultural, not biological. When teachers have lower expectations of girls, it is hardly surprising that boys show more favorable attitudes towards mathematics.

If Sophie Germain were alive today, would she be any more welcome in mathematical circles than in eighteenth century Europe?

Maryam Mirzakhani gives us reason to be hopeful. In 2014 the Stanford professor became the first female recipient of the Fields Medal; the highest accolade in all of mathematics (our equivalent of the Nobel Prize or, if you prefer, the Oscars).

You do not need to understand Mirzakhani’s work (and unless you are an expert on Riemann surfaces, be grateful for that) to appreciate its significance.

Maryam Mirzakhani — first female recipient of the Field’s Medal

Mirzakhani has torn down gender stereotypes. As an Iranian, she may yet have other threats to face from the Trump administration. It appears that, even in 2017, the country into which we are born profoundly shapes our life chances.

Not all geniuses are born into equal opportunity.

Plucked from obscurity

Srinivasa Ramanujan epitomized the term “genius.” For mathematicians, his is a true rags-to-riches story. Born and raised in rural India in the late nineteenth century, Ramanujan had scarce few opportunities to develop his raw mathematical skills. He worked with the crusty relics available to him. On one occasion, he encountered a primitive, out-dated textbook and from this simple manuscript derived deep mathematical theorems — many of them unknown to the best mathematicians of his time.

Srinivasa Ramanujan in optimistic mood

Ramanujan sent his manuscripts over to England, where they were rejected by everyone but GH Hardy, the famous maths professor at Trinity College, Cambridge. Even Hardy was ready to dismiss Ramanujan’s work, until he realized the remarkable depths of what he was reading. Hardy invited Ramanujan over to Cambridge. The two men formed the most unlikely of alliances. Despite his life being cut short by illness, Ramanujan’s legacy as a mathematical great is secure.

It took a chance encounter with a textbook and a leap of faith from a Cambridge professor for Ramanujan’s talents to be realized. How many more Ramanujan’s are languishing in obscurity?

There are seventy million unschooled children in the world. That’s seventy millions beacons extinguished without a hope. Not that modern schooling is a safe haven to opportunity. Socioeconomic achievement remains closely related to educational outcomes. Access to education counts for little when it is not backed up by quality.

Genius at our fingertips

The world needs mathematicians more than ever before. Yet historical barriers — from parental expectations and gender stereotypes to xenophobia and poverty — persist to this day. A handful of mathematicians may prevail through dogged persistence, sheer luck or divine intervention. But for their triumphs, millions more may perish. Genius can never be taken for granted.

The values of society exert untold influence on children’s mathematical achievement. It is our beliefs and choices that make mathematical genius so fragile.

We have a global talent pool of mathematical geniuses at our fingertips. We can not afford to let them slip through.

I am a research mathematician turned educator working at the nexus of mathematics, education and innovation.

Come say hello on Twitter or LinkedIn.

If you liked this article you might want to check out my following pieces:

These four visuals dispel the myths we’ve been fed about maths education
_The case of Nancy and Liam_mystudentvoices.comEdTech’s culture problem
_Education is innately cultural. The way we learn is shaped by our environment, our experiences and our belief systems._medium.comMathematics without history is soulless
_π through the ages_hackernoon.com

Free Code Camp asked 15,000 people who they are, and how they’re learning to code. I separated them by gender, continent, and whether they’re ethnic minorities, then explored their differences.

_Image by [me](http://davidventuri.com/" rel="noopener" target="blank" title=").

More than 15,000 people responded to Free Code Camp’s 2016 New Coder Survey, granting researchers (like me!) an unprecedented glimpse into how people are learning to code. The entire dataset was released on Kaggle.

Here is how new coders of different genders, ethnicities, and continents vary by wage, age, and hours dedicated to learning.

The demographic distributions for the 15,620 new coders who responded to the survey are as follows:

Gender

• 21% (2,840) are women
• 78% (10,766) are men

Ethnic Minority vs. Majority

• 24% (3,253) are ethnic minorities in their country
• 76% (10,392) are not

Citizenship by Continent

Does expected next salary vary by demographic?

This next salary is their first one after advertising their new coding skills. Expected next salary is one of two main questions in the Free Code Camp survey whose answers depend on the quality of coding resources.

Post-coding skills acquisition, North Americans expect the highest salaries and Europeans expect the lowest. Expectations in Asia are all over the board.

The 25th percentile North American expects the same as the 75th percentile European: $50k. The median North American expects $60k per year.

I wonder if some Europeans forgot to convert from pounds, euros, or any of the other European currencies to US dollars.

By the way, here’s how to read this chart (and the other box plots in this article): the “x” is the mean. The horizontal line is the median (a.k.a. the 50th percentile). The bottom of the box is the 25th percentile, and the top of the box is the 75th percentile. Whisker length is 1.5 times the height of the box. The circles are outliers. All y-axes are on a logarithmic scale to better visualize the outlier-heavy data.

Females actually expect higher next salaries than males.

The median female expects $9k more than the median male. The 25th percentile female expects $14k (!) more than her male equivalent. Female new coders appear optimistic about the changing diversity landscape in the workplace.

Like gender, minorities expect higher next salaries than the majority demographic.

The gap in medians is $10k. The gap in first quartiles is $15k. Minorities also appear optimistic about the changing diversity landscape in the workplace.

Expected salary is about the same, whether you dedicate less than 10 or up to 40 hours per week.

Those who dedicate 40+ hours per week have a mean expected salary that is $3k to 5k higher than the others, but this could be caused by random chance. Only 694 of the 15,000+ respondents spend this much time learning.

So expected next salary varies wildly by continent. There appears to be a reverse wage gap trend going on with gender and ethnic minority status.

Less than 5% of new coders are dedicating 40+ hours to learning each week.

Most of these respondents are in their early twenties and have a bachelor’s degree, which suggests that some are forgoing traditional forms of higher education (like master’s and professional degrees) and using those 40+ hour weeks to learn code.

By the way, this is the situation I myself am in with my personalized data science master’s degree.

As the awareness of the quality and affordability of online education rises, I expect more people to join the higher brackets.

Do hours dedicated to learning per week vary by demographic?

The hours dedicated to learning question is another one whose answer depends on the quality of coding resources.

Time spent learning to code is pretty much identical across continents. A median of 10 hours per week is standard.

Same thing across genders.

(Note that for trans people, the difference was not statistically significant.)

And same thing for ethnic majorities vs. minorities.

So unlike expected next salary, hours dedicated to learning doesn’t vary much by demographic. The bulk of each spends between 5 to 20 hours learning weekly.

Let’s dig deeper into the wage gaps.

For both males vs. females and ethnic majorities vs. minorities, two grouped scatter plots follow:

1. Current salary (before advertising their new coding skills) vs. age
2. Expected next salary (after advertising their new coding skills) vs. age

Each has a best-fit line labeled with its correlation, as well as dashed lines representing the median for each axis variable. I removed new coders that are 65+ years old since they are statistical outliers.

Wages Across Ages

Female new coders have higher median current salaries than males (by $2k), but the higher female salaries are concentrated in millennials.

Note the current salary vs. age correlations of 0.267 and 0.192.

Male new coders do have an above average proportion of very high ($150k+) current salaries, which corresponds to a slightly higher mean (not plotted).

For gender, the above relationship between age and wage isn’t maintained when new coders project their next salary.

The correlations are much lower, with both near 0.1. Again, we see the huge gap in medians: $59k for females and $50k for males. These are both higher than each gender’s current salary median.

_Photo via [KuniakiIGARASHI](https://www.flickr.com/photos/igaiga/8524166440/in/photolist-dZfzGU-dBG6fK-mavikK-dZ9LET-dBG7yK-dZ9PFK-dBMupE-dZfzLo-dZ9RUR-mavkDx-dAQVYu-dBG8M4-dBG5Qt-dZ9SCi-dfRvyL-maw2Za-dBG5Wc-dBG8Ec-dBMx5J-dBMxcY-dBMwXy-fwy6Vc-rtdj7S-re4gSB-rtdmmG-rvpEFE-rtdfs7-rvuWeM-rvp4kt-qyvqtw-dZ9Tcn-dBG6uX-mawZrJ-dC4z8C-kRyZsZ-dZ9Npa-dBY8CZ-dfRPuf-dZfvtq-dBG5zK-dfRMHK-dBG79V-dBMvY7-mavZxM-dZfwpE-dZfviC-dZfx9d-maw2ap-dZfwnf-dZ9PSR#undefined" rel="noopener" target="blank" title=")

Minorities have a higher median current salary than majorities by $4k. Unlike gender, the spread of current wages over age is about even.

Note the current salary vs. age correlations of 0.253 and 0.243.

Ethnic majorities do have an above average proportion of $150k+ current salaries, but it doesn’t correspond to a higher mean (not plotted) this time.

For ethnic minority status, like gender, the relationship between age and wage isn’t maintained when new coders project their next salary.

Both of the correlations are much lower, near 0.1, again. The $10k gap in median expected next salary is striking: $60k for ethnic minorities and $50k for ethnic majorities. As with gender, these are both higher than their current salary median.

_Photos via [Black Girls Code](https://www.flickr.com/photos/perkinsy/9537573288/in/photolist-fwNyfL-dBMuTu-mavjWa-dZfxz5-mawXAj-max12m-dBMuVd-dfRvWf-kRyigi-dBG5UV-dZfwJW-dBMwDd-dBY7QM-dBG874-dZ9Ric-dZfu75-maviov-dBMwjw-dZfv8Y-dBMwfm-dC4xFW-dfRN4K-dZ9MoD-dZfzm1-dBMwSU-dC4z2u-dZfynG-dBMxBw-dBMun7-dZftHJ-dBMv89-dBMwHQ-dAKsMH-dZfvB9-dZ9RBD-dZfvES-dBMvzj-dZfz1J-mawYim-dZ9QqX-maw34D-dZfvwA-kRyhGc-dZfy53-dfRMcx-dBG7ua-dBG7Va-dZ9TgD-dBY93c-mawZv1#undefined" rel="noopener" target="_blank" title="">Yvonne Perkins and <a href="http://precollegeblog.flatironschool.com/post/121194821475/black-girls-code-and-flatiron-school-partner-for" rel="noopener" target="blank" title=")

So, like expected next salaries, female new coders have higher median current salaries than male new coders. So do ethnic minorities vs. ethnic majorities.

Older females don’t do as well as older males, however, which is the only hint of a wage gap that I could find in the entire dataset.

For both gender and ethnic minority status, it appears that older new coders are willing to take a pay cut when transitioning to a job where they advertise their new coding skills, while younger individuals intend to capitalize on coding demand with a hefty salary early in their career.

What did we learn?

Hours dedicated to learning to code is pretty much constant across gender, ethnic minority status, and continent. Most people spend between 5–20 hours each week.

Expected next salary (post-c0ding skills acquisition) varies wildly by continent. The lowest median is $30k (Europe) and the highest is $60k (North America).

Older new coders appear willing to take a pay cut in their new job where they advertise their new coding skills, while younger new coders look to start their careers with substantial salaries with their in-demand skill.

The traditional gender and ethnic minority wage gaps aren’t prevalent in the 2016 New Coder Survey. In fact, they are reversed. Perhaps new coders aren’t reflective of the working population in general, where data suggests that both wage gaps still exist in 2016.

_Photo via [SHRCC](https://www.flickr.com/photos/shrcc/4535210853/in/photolist-7UL9bz-jXi8bu-drKeWS-EH6mXe-cCoTSN-8xcfPC-6aoYyA-nRkbdo-dVH1tJ-8p5VgK-cEZjtj-soV5rS-spfY1q-efBP6h-axL5Wm-aKypuc-7KmDRt-8mnyNH-4uCYto-RapM6-fNunxg-9cNj9r-cqeakC-oXqXpL-2uPztV-5hWNCu-9TV2wn-9TSDDG-iUMeZi-ouFdL9-brensu-4ETTCT-5QnPiV-5hPTqV-bCVbLW-ecDZDm-dVvEo9-8WqgrX-fiA5cJ-6H6ebv-axL639-a5v941-8zGdQ7-9cauL6-nM7B44-9cTmCG-3KBVk4-nNCjMj-f6UVnA-5FSSKq#undefined" rel="noopener" target="blank" title=")

What’s next?

Do you have a hunch as to why the ethnic minority status and gender wage gaps might not apply to new coders? Please share anything relevant (or contradictory!) in the responses.

You can find a more detailed version of this analysis on Kaggle, where you’ll find statistical tests supporting the inferences in this article.

Be sure to check out my other pieces exploring Free Code Camp’s 2016 New Coder Survey:

Developing Data Scientists and Engineers
_Free Code Camp asked 15,000 people who they are, and how they’re learning to code. I isolated those focused on data…_medium.comThe 6 most desirable coding jobs (and the types of people drawn to each)
_Free Code Camp asked 15,000 people who they are, and how they’re learning to code. I separated them by their job…_medium.freecodecamp.com

If you have questions or concerns about this series or the R code that generated it, don’t hesitate to let me know.

David Venturi (@venturidb) | Twitter
_The latest Tweets from David Venturi (@venturidb). Creating my own data science master's degree. @queensu chem eng/econ…_twitter.com