From Economics to Data: My Journey to a Data Career

Although my university background was not data-focused, I decided to pursue a career in data after graduating—here is the story of how I got where I am today.

What sparked my interest in data

My interest in data began during my university years while pursuing a degree in financial economics.

Throughout my three-year program, I was exposed to econometrics modules, which I found very interesting and engaging. Econometrics, for those unfamiliar, is a branch of economics that utilises mathematical and statistical methods to develop theories or test existing hypotheses in economics or finance.

Think of it as a toolkit that economists use to dissect data and provide answers to questions about how various factors impact economic outcomes. After learning the theories and methods in lectures, our practical sessions involved us mostly running different regression models to answer specific questions.

One particularly vivid memory from my first year stands out: a data analysis lecturer demonstrated how using the very same dataset, one could produce different charts that convey almost contradictory interpretations at first glance.

This fascinated me, highlighting how the media can exploit the limited data literacy of the general public to mislead them.

Sometimes, it requires a bit of time to thoroughly analyse the visualisations presented, as they may not always offer immediate clarity upon initial inspection. It taught me a lesson to always look carefully and a little deeper, asking questions, even if the answer seems obvious.

An example is shown below. Think about how these graphs initially tell us different things. This is using the same data! If you are interested, you can see more real-life examples of misleading graphs here.

Image Source: Down the Funnel

Projects at university

During my degree journey, I had various econometrics projects. The structure was the same for all; we were given a dataset and had to answer some questions. What varied was the type of dataset, the questions, and the regression models required to answer them.

For the final year project, we were given a large longitudinal dataset (a dataset containing observations for many variables across many points in time). The dataset included variables such as population density, life expectancy, case numbers, vaccinations, tests, hospital beds, and stringency indices.

The main task was to investigate the determinants of COVID-19 cases across North and South American countries. I had to construct a regression model that yielded statistically significant results, while justifying each of the chosen variables of my model using literature. After experimenting with different combinations of variables and regression models (including linear, log-linear, linear probability, probit, etc.), I could answer the project’s questions.

These projects helped me realise a few things.

Firstly, just how important data quality is. For example, the dataset could barely have any missing observations, but if the collection method is questionable, how much can it be trusted?

Secondly, and more importantly, what is done with the data is critical. Even with the cleanest, highest-quality dataset collected meticulously, misusing or using it in an unfit model can lead to inaccurate analyses and conclusions. This can be very costly for organisations and governments. On large scales, this can result in legislation, regulations, or policies that lack the necessary data-backed foundation and consequently fail to achieve their intended objectives.

Lastly, when answering a question using data, there will always be external factors we may not necessarily have the data for or even be aware of. In the real world, a model will never be 100% accurate, but it can certainly be great at answering some questions and helping organisations make data-driven decisions and move forward.

After graduation

Following my graduation, I started looking for roles in the field of data. I chose this field because data is everywhere, and the skills needed are not necessarily industry-specific. Further, more than ever, organisations are trying to become more data-driven.

Then I came across Rockborne. Their training programme for graduates is what drew me to them. This was particularly appealing as I did not come from a data-focused background. I recognised the immense value of acquiring the skills and confidence necessary to advance my professional career. I am happy to say that it has been 8 months since I joined Rockborne, and I have learned a lot, both technical and soft skills.

Interested in joining our diverse team? Find out more about the Rockborne graduate programme here.