One with a curious mind, i-CATS University College Assistant Professor Dr Donald Stephen often wonders about the human mind. After earning a degree in Cognitive Science and being exposed to psychology, data analysis, programming, and artificial intelligence, he found his true passion when he realised psychometrics is what brings all these fields together.
As a psychometrician today, he finds it satisfying to dig into data, spot patterns, and figure out what everything means — though most of the time, it leads to more questions than answers. Enjoying the process of understanding the human mind through data and patterns, Donald shares with Sarawak Tribune more about his job.
Q: What is your role as a psychometrician?
A: As a psychometrician, I use data to make sense of the human mind and behaviour, such as perception, motivation, and attitude. I design tools to measure these invisible traits and ensure they are reliable, meaningful, and useful for real-world decisions.
How did you end up becoming a psychometrician?
Becoming a psychometrician isn’t something most people just stumble into — it’s a very specialised field that you grow into, usually through years of postgraduate study and research. I pursued a PhD focusing on psychological measurement, and over the past 15 years, I’ve been immersed in this work. What keeps me going is the fascination with the art and science behind measuring something as abstract as the human mind. It’s a challenge I genuinely enjoy.
What was your initial career goal?
Back in secondary school, my teacher asked us to write our ambition on the class notice board. I wrote “lecturer”. And I’m grateful to say — that’s exactly what I am today. What I didn’t expect was ending up as a lecturer with a niche in psychometrics.
What does your day-to-day job look like?
Most days are a mix of teaching, research, and helping clients make sense of their data (that’s where the psychometric work really comes in). I’m also involved in research activities and admin work at the Research Management Centre. It keeps things dynamic, and I enjoy the balance.
What kind of assessments or tests do you usually work on?
I work on a variety of assessments — from measuring stress, motivation, and attitudes to evaluating job satisfaction or learning experiences. Some are custom-built for companies or communities; others are improvements of existing tools. The goal is always the same: to make sure what we’re measuring truly reflects what people feel or think, and that the results can be trusted for real decisions. If we can understand people better, we can support them better.
What happens behind the scenes when you create a psychometric test?
Behind the scenes, it’s a mix of science, statistics, and quite a few cups of coffee. I start by defining what we want to measure — like anxiety or motivation. Then I draft the items, pilot them, and collect data. After that, I run analyses to check if the test is reliable, valid, and free from bias. It’s a careful, sometimes repetitive process, but it’s worth it.
How do you ensure the reliability and validity of a psychometric test?
I’m very tempted to start a lecture now, and if I do, you might be stuck with me for the rest of the semester. But in short, it’s a bit of detective work. There are many methods and theories to do that, like Classical Test Theory, Item Response Theory, and more recently, Network Psychometrics. You interview, observe, run analyses, refine the tool — and probably need a few more cups of coffee along the way. You can’t rush art.
How can psychometric testing help improve education, hiring, or mental health support?
Psychometric testing is not about what I can do, but more about what the client needs to know. Whether it’s learning gaps, competency, or emotional wellbeing, the tools are built around the questions you’re trying to answer. People often assume psychometrics only refers to those personality quizzes during job applications, but that’s a misconception. It actually covers a broad range of assessments used in education, healthcare, research, and beyond.
As a psychometrician, how differently do you view the human psychological aspect compared to before you studied psychometrics?
My view hasn’t really changed. The human mind is still deeply mysterious. The difference now is that I know how to understand it better — even just a little.
As a psychometrician, being meticulous — especially with numbers — is crucial. Do you see yourself having the same soft skills in life situations?
Giving me data is like giving me a box of Lego. There are countless things I can model and patterns to uncover. Honestly, I do enjoy spotting patterns in everything. But yeah, it can be a bit much when your brain just won’t shut off — like when I’m on Shopee and suddenly I’m comparing specs like it’s a research paper … and still end up buying the wrong size.
What are your thoughts on the use of AI in psychometric testing?
People sometimes forget that the mind is mysterious and incredibly complex. One valuable use of AI in psychometrics is modelling comorbidity or partial independence between psychological traits. For instance, anxiety and depression often occur together, but they are not the same. AI helps us see where they overlap and where they remain distinct, which is important when designing assessments and interpreting results. I also use AI tools to handle large datasets, cluster responses, and even predict outcomes based on behaviour patterns.
Can a test actually predict how good someone will be at a job?
Having measured both competency and job performance for my past clients, I can say tests can offer useful insights. They help identify whether a person has the potential to perform well in a role. But they are not the full story. Real performance also depends on context, leadership, team dynamics — and even timing.
How much can people fake a psychometric test? Can you tell when someone’s faking?
As often as we think. It’s not foolproof, but by analysing response patterns, I can usually tell when something feels off. For example, when someone answers every question in the most extreme way, tries to appear overly perfect, or gives contradictory responses to similar items — these are clues that the answers might not reflect their true self.
I get this question a lot from clients too. Take, for example, a test where someone with high ability gets the easy questions wrong but answers the harder ones correctly. That kind of pattern stands out. And yes, analysis can detect those kinds of anomalies in the data.
But here’s the important part: an anomaly doesn’t automatically mean someone is cheating. It just means their responses don’t follow the usual pattern we’d expect. That could happen for all sorts of reasons — maybe they were distracted, misunderstood a question, second-guessed themselves, or just had an off moment.
The analysis tells us what is unusual, but not why it happened.
