Gender Distribution at ITU

Tali Melchior has investigated the gender distribution at the IT University of Copenhagen. Being a female student at the MSc in Software Development she has experienced being a minority, which sparked her motivation for the project. Using data from the Analysis Unit she has created graphs to visualise the gender gap. Below you can read her findings.

In the latest “Regeringsgrundlag” (a declaration signed by all parties in the government) the Danish government took on the topic of gender equality and stated:

“If Denmark is to succeed in a globalized world and continue to create economic growth through knowledge and skills at a high level, it is necessary to train skilled IT specialists and advanced IT users for the future labor market. Today few women choose the field of IT and programming. The government wants all talents come into play, and therefore we will launch an effort to promote girls’ and women’s interest in IT and programming across companies and educational institutions- for example through role models, internships and activities.” (regeringsgrundlag, p. 82.)

Ever since I started studying at ITU in the MSc in Software Development Program, I started becoming more aware of being a minority. This, ironically enough, comes after having been one my entire life (I am both Jewish and red-haired!). But at ITU it is not only my religion or hair-color that make me stand out; it is my gender! In the Software Development programs (BSc and MSc) out of 245 of the students only 41 are female (14%)! All of this made me wonder: Why is it I only have one female lecturer? Why are there so few girls around me in class? What can be done to change this imbalance and allow all talents to “come into play”?

This lead me to the research I have performed in the last semester at the ETHOS Lab. I realized that the first step to finding answers to my questions was to find out more about the past and current gender distribution at ITU, among both students and employees. The Analysis Unit were very helpful in providing me the data I requested, in form of tables and figures. Since numbers can be hard to grasp, I decided to put the data into a visual context by using Tableau. I assumed that when seeing the data in the following graphs we would be able to better understand the significance of the data, and expose patterns and trends that may confirm my assumptions about the gender gap at the university.  

The visualizations are based on 2 datasets, students and employees. The student dataset was received on October 26th, 2016 and contained anonymized data of all students enrolled at ITU since it was established in 1999.  The employees dataset contained anonymized data of all current employees as of November 4th, 2016.

These visualizations focus on gender distribution and contribute to the prime goal of this project- to raise awareness to this matter. This project can help the university management, employees and students gain insight into the current state of gender distribution at ITU. I hope that these insights will lead to operational objectives regarding hiring policies and student intake, which will eventually create a more even gender balance. In the long term, I would like this to be the basis for further research and calls to action that potentially will attract more women to the Computer Science Department at ITU.

Gender Distribution Among Students

When looking at the student data, the first question I asked is: How many women and men graduated from ITU since the university was established in 1999?

Graph 1

In Graph 1 we can see the total number of graduates since it was established in 1999 until today. When looking at percentage rates we find that 24% of the Bachelor’s graduates are female (less than a quarter!), and 36% of the Master’s graduates are female.

As a side note, it is important to remark that the first degree programs offered by the university were master level, hence the higher number of MSc graduates. Bachelor programs were introduced in 2007.

In order to compare total ratios over time with the current distribution and then explored the following question: What is the current gender distribution across programs? In the following graph are the stats for all students currently enrolled:

Graph 2

In Graph 2 the first thing that stands out is that only one program has a majority of women, and that is the MSc in Digital Design and Communication (DDK).

The three programs that have a vast majority of men is the BSc and MSc in Software Development, and the MSc in Games. These programs are considered more technical and some even perceive them as ‘hard core’ study lines (which is definitely a loaded word in our context…). However, many assumptions related to this tendency require more research, such as examining how many women apply for these programs in the first place, which considerations go into choosing a certain program, how these programs are marketed and presented, etc.

Graph 3

Graph 3 shows the gender distribution across programs since the establishment of the university. This longitudinal data is a great tool for examining processes and changes over time. There are a few interesting things to notice in this graph. The first things that stands out are the thin green lines in most programs. Even in SDT where the number of women has grown, they are still a clear minority compared to the number of men.

However, I found the DDK graph most interesting because it seems like in the first few years the gender distribution was rather equal, and over time it became the (only) program with the biggest majority of women. I would like to suppose that there may a process of polarization happening, in which over time certain programs got a reputation of being more appropriate for a certain gender, and then students’ choice of study-lines began to reinforce that reputation. This is an assumption that would be interesting to investigate further, not only in the ITU context, but in other institutions and areas as well.

The final question I will present here regarding the student dataset is: What are the graduation rates vs. the drop-out rates of women in the BSc in Software Development? I checked these stats because it is important to look both and enrollment and at graduation to see if women who begin these study programs complete them and graduate. I also had heard that in 2012 the three women who participated in this program all dropped out. The graph presents the gender distribution showing the number of students of each gender who either graduated or dropped out, over a six-year period:

Graph 4

In the graph we can see that the same thing happened also in 2009, and in the two years in between there was only one female graduate in each year! In 2013 one woman graduated, and four are still enrolled and will hopefully graduate with a delay.

This phenomenon is in no way exclusive at ITU. In Carnegie Mellon’s undergraduate Computer Science Program female enrollment had hovered below 10% for a number of years, and the fraction of women leaving the program was approximately twice that for men. However, during the mid-late 1990s, a research study was designed specifically to diagnose and find remedies for the gender gap. As a result of that process the figures were changed drastically: the percentage of women entering the School of Computer Science rose from 7% in 1995 to 42% in 2000. I recommend looking into the project that lay behind this transformation. 

Gender Distribution of ITU Employees

As of November 2017, ITU has 564 employees, of which 337 are male (60%), and 227 are female (40%). In the following set of graphs I focused only on academic staff and their distribution among sections. As a side note, the section based system as presented here ended in 2016, and as of 2017 the university is divided into three departments.

The following graph shows the distribution among staff members across sections. The acronyms of the sections stand for: Digital Society and Communication (DISCO), Information Management (IM), People and Computational Things (PACT), Technology in Practice and Games (T&G), Software and Systems (S&S), and Technical Computer Science (TCS).

Graph 5

Here we can see that in 5 out of 6 sections there is a majority of men. S&S and TCS, which were recently merged into the new Computer Science Department, both have a vast majority of men.

My experience of having one female lecturer out of thirteen in total also reflects the gender distribution among faculty employees at ITU. Graph 6 shows the gender distribution in the four ‘VIP’  academic positions:

Graph 6

The highest academic rank in Denmark is the full professor. ITU has six full professors as of fall 2016 and of these, all are male. In the other three positions the ratio of men to women is approximately 2:1.

The last graph I will present shows number of current PhD students in total, and their distribution across sections. This is an important aspect to consider because it reflects how the next generation of researchers is being trained, and here too we can see similar tendencies:

Graph 7

On the left we can see that as of November 2016, only 1 in 4 PhD students are female (approx.). On the right we can identify a similar distribution across sections to that in Graph 5.

Conclusions and reflections: Is this a satisfactory picture?

I believe this is not a satisfactory picture, and that it could and should be changed. There are endless articles and blog-posts on why fewer women choose to study Computer Science and what can be done to address this. Several “Women in Tech” groups have been formed all over the world, including here in Copenhagen. In Danish press the topic has also been brought up time and again. Mads Tofte, the Vice-Chancellor of ITU, was interviewed in an article published by Politiken and said that ITU has begun taking steps to address this issue. I am aware of an internal research project currently being conducted for ITU by Valeria Borsotti. Valeria is focusing on the student experience at the Bachelor Software Development program, and on the current gender gap in the Software Development education. The university has also nominated a woman as the head of the new Data Science program and shows only women in the main video advertisement that has been circulated in various media platforms. (I find it quite curious that they avoided showing any men in this ad, maybe creating a slightly too transparent attempt).

It is encouraging to know that our university is somewhat engaged in addressing this issue. Some further suggestions that could be relevant are highlighting more female role-models, assisting and supporting women as a minority, employing more women to VIP positions, accepting more female students, attracting women by redesigning the brochures and the university publications, displaying more success stories of females on the info screens, and more.

What this project illustrates most is structural inequality, which as such is quite difficult to change. The end-goal of the whole discussion about women in tech should be to change this structure. Because it will improve our tech industry. Because it will promote gender equality in our society. Because IT is and will for a long time be one of the main channels to material wealth and influence – and why shouldn’t women get their fair slice of that? 

This project is just a starting point for raising awareness, and I hope that others will continue looking into this topic. Many more aspects can and should be researched. Numbers tell one story but students’ experiences are equally important in understanding the dynamics of gender at ITU. Therefore it would be interesting for example to interview students and faculty regarding their perception of ‘hard-core’ versus ‘soft-core’ programs and courses. That would maybe show how stereotypes play a role in the choices people make, and in the way they perceive study programs, courses and jobs.

I acknowledge that this is indeed a “wicked” problem, almost impossible to solve. Nevertheless, I would like to ask from my position, as someone who has been cast as a minority: Will management build on the research conducted in this area in order to tackle the wicked problem of gender inequality in higher education institutions?

I would like to thank Rachel Douglas-Jones and Valeria Borsotti for their insight, comments, and references to the right people and resources for this project. I also want to thank the ETHOS Lab members for this inspiring experience. Being a junior researcher has exposed me to many interesting questions and useful tools. Moreover, meeting with faculty and students from other programs opened up a whole new part of the university to me, and I feel grateful and lucky to be a part of this lab.

-Tali Melchior, January 2017