Psychologists – the Secret Data Scientists
As someone in their early twenties who has worked as a Data Scientist for almost two years, it’s pretty easy to assume I jumped right into computer science from high school to learn about the sweet spot between statistics and programming.
The fact that I was lucky enough to secure one of two Data Science spots in REA Group’s graduate program this year, reinforces this assumption.
However, many are surprised to know that I have a Bachelor of Psychology and spent more time at uni studying psychology than I did anything else. Here’s why my psych background stood out during the hiring process, and why many psychologists are around 90% trained as a killer Data Scientist – without even knowing it.
Making the Big Switch
At the start of my final year of my Bachelor of Psychology, I realised that psychology just wasn’t for me. However, I’d fallen in love with statistical analysis through units exploring research methods in psychology. After nervously chatting to several career counsellors, I got the help I needed to make the career change to Data Science.
I did a two-year Master of Data Science, got an internship for a start-up in my first year and secured a permanent position three months later. Shortly after this, I was offered the opportunity of a lifetime – to work alongside the best of the best at REA Group as a Graduate Data Scientist!
While my master’s degree taught me the fundamentals of concepts like machine learning, data wrangling, programming and visualisation, it’s a fraction of what I’ve learnt on the job at REA Group。 This point here is really important and I’ll come back to it later。
Different Cultures, Same Practices
Going into my Master of Data Science, I was insanely intimidated by the term “Machine Learning”。 I was expecting these terrifying computer algorithms that could be used to create Skynet… until I realised that I’d been creating them since my first year of my bachelor’s degree。 At uni, I learnt when testing the impact of a variable on a group (such as the number of choices on a participant’s happiness with their choice), psychologists might perform linear regression to see how well different variables can predict a certain outcome。 If it’s a classification problem, they’ll happily use multinomial regression or tree-based methods。 If they need to cluster groups, K-means clustering will be used。 All of these were taught as statistical methods in psychology。
In Data Science, they have a different name: machine learning algorithms. Psychologists simply use these algorithms to see how well variables are related, and they’ll make several interpretations on a single output; about confounding variables, extraneous variables, statistical significance, and many other factors. They’ll manipulate the data in amazing ways, perform different statistical tests like Chi-square tests and ANOVAs, and they’ll do all of this without breaking a sweat.
They’ll then thoroughly explain the results in their study (regardless of whether the data tells the story they were hoping for)。 This is a vital skill with which many data scientists struggle。 A great algorithm fails if you can’t explain how and why it works。 My knowledge of these concepts helped me immensely when it came to exam time, as I’d already spent a huge chunk of my time working with them。
Learning at University vs Work
For those who are familiar with Data Science, I’m sure you’ve seen a Venn diagram at some point showcasing the different skills you need to call yourself a Data Scientist like the following:
Weirdly enough, while I learnt a decent amount about programming, computer science and statistics in my course, I learnt around 80% of what I know in half the time on the job at REA Group。
Touching on what I emphasised earlier, I’m almost certain that my skills and knowledge of Data Science would be almost identical if I had never done my master’s degree and started this role six months earlier。 You learn so much faster when you’re applying your knowledge to real-life problems and surrounded by experts! However, there’s one significant skill that I bring to this role which I have found to be really difficult and time-consuming to learn。
And that skill is called empathy
Programming and math can come easily to some and difficult to others, but one thing that almost everyone can agree on is that empathy is grossly underrated in this line of work. My job involves using the behavioural data that we have collected from millions of consumers, across millions of listings, to understand our audiences and personalise their experiences.
足彩胜负14场If I want to build a model that predicts which of our users are likely to be a buyer, I have to put myself in the perspective of a buyer and think about how I would behave and why, and then test to see that behaviour is reflected in our data. This is something that comes naturally to a psychologist after years of training, and they can do it better than most without hesitation!
Looking forward to the future!
With only 8 months left in this grad program, it’s pretty amazing knowing that I’ll soon be a fully-fledged Data Scientist at REA Group and teaching up new recruits at this organisation.
Nick de Silva
I'm a Graduate Data Scientist, and joined REA Group in February 2019 as part of their Graduate Program (through the Data Science and Analytics stream). I'm a big fan of deep learning and AI, and specialise Python, SQL and machine learning.View profile for Nick de Silva