Big data, big improvements

As Manager of the UTS Peer Assisted Study Success (U:PASS) program, Georgina Barratt-See wanted to know if it was really making a difference to student success. With 10 years of data to analyse she didn’t know where to begin. That’s where postgraduate student Len Tay and the Master of Data Science and Innovation’s subject iLab came in – upskilling staff in data literacy while empowering students to work with real clients.

Georgina Barratt-See

We have 110 high-achieving and compassionate leaders running study sessions with students, mainly in first-year subjects. We know peer-based work is much more effective as the students see the peer leaders as just another student.

When I heard the Master of Data Science and Innovation had a subject giving master’s students projects to look at real industry data, I instantly wanted to know more. I've been doing my job for a very long time now – almost 10 years – and we compile a lot of individual data sets each session and perform a lot of evaluation. We look at how often students come to classes, the types of students that attend, their opinion of U:PASS, that kind of thing. But, there was no linkage of that data.

I don’t have the ability or time to look at our data in-depth and I always thought it would make a good research project. This was a great chance to look at the big picture results, like how the different subjects are performing and which ones we could work on. I put together a proposal and next thing I knew I had an email from Len telling me she had selected my project.

We instantly connected. Working with a student can be hard because you spend so much time teaching them what needs to be done. But Len just ‘got it’, instantly. I immediately saw how intelligent she is.

If you meet Len she's just so calm and thoughtful. I'm kind of extroverted and passionate, so this pairing between us just worked.

I gave Len so much data and thought, 'She's not going to cope with this’. But she was able to unpick all the questions we were asking from a data perspective. As the project progressed, she clarified, probed and asked good questions.

We’re already seeing the benefits of this data project. I was approached by an associate dean asking for data from their particular faculty. I pulled out some of Len's data and explained what we'd discovered over the semesters and how we could see the different patterns across the subjects. It was great. 

Working with Len was really rewarding and easy. I hope to stay in touch with Len and I like to think I’ve helped her in her new data scientist career.

Len Tay

I’ve been working in computer science for close to 20 years and have seen a lot of things we’re doing in the field already being outsourced. That’s why I decided to go back to university and study the Master of Data Science and Innovation. It’s a degree for students who’d like to become experts in data science.

I was attracted to the degree because it pointed to the future – I wanted to move into something that I can see being ‘alive’ in years to come. Plus, I really like writing algorithms. I like doing programming. I like analysing data.

I met Georgina through iLab 1 – it’s a subject that requires students to team-up with real clients to analyse their data. Out of a list of possible projects, both at UTS and in industry, I chose Georgina’s U:PASS project because I wanted to do something in student analysis. I’m a student myself, and my two kids study at UTS. I was curious to see what sort of experience students are having on campus. And the U:PASS data could tell me that.

I knew from the start that we’d work really well together. Georgina is a very open and helpful person and was flexible with how we worked. That’s a big thing. I’ve met other clients who are very demanding but Georgina wasn’t. She was very respectful of my time.

Georgina’s more of a morning person and I prefer to work at night, so we’d just email back and forth. I’d say I spent a total of 50 or 60 hours mining the data over about three months.

Georgina was also very clear on her objectives. Some clients just give you the data and ask you to find correlations and patterns, but Georgina is smart and knew what she wanted to ask of it: which are the best performing subjects? Which are the worst performing subjects? If a student stops attending U:PASS will they still continue to improve or not? Essentially, what’s the mark, if any, that U:PASS leaves on the student’s learning.

Before the project, I wasn’t sure if I could do data science confidently on my own. But working through it gave me enough confidence to say, ‘Yep, I know what I’m doing’.

I think Georgina was surprised with what data can do. Or better, what we can do with data. When we started she was wary of handing over hundreds of thousands of records; like it was too much for me. But, I smiled and said, ‘Don’t worry, the programming language can handle it’. Plus, data analysis gives me a kick! To be able to answer the questions people have, that gives me so much satisfaction.

So, what did the data tell us?

• 16 per cent increase in the grades of students who regularly attend U:PASS
• U:PASS students half as likely to drop out of a subject compared with non-U:PASS students

Unlock your data

Are you a UTS academic, faculty or unit keen see how data analytics can help the work you do? Get in touch:

• Masters of Data Science: Just like Georgina, pitch your data needs for an analysis project headed by MDSI students. Email Amanda McGregor or Rhiannon Tuntevski at FTDiPartnerships@uts.edu.au
• Connected Intelligence Centre (CIC): Gain insights from your data by working with CIC’s Data Scientist Michael Pracy. Email michael.pracy@uts.edu.au