SEARCH
News
02 July 2021

Taylor’s Case Study: Solving Real Problems Through Statistics

If you think statistics is just a collection of numbers, you may be surprised how it could answer world problems.

What do COVID-19 efficacy rates, the amount of time you spend on Instagram, and your final-year project have in common? 

They all deal with some form of statistics. 

For many, these statistics can be used to solve everyday problems and bring about a positive change to society.

For Associate Professor Dr. Chang Yun Fah, from Taylor’s School of Accounting & Finance, becoming an applied statistician stemmed from his background in Arts and Social Science as well as his love for Maths. 

He shares his journey and the most important thing for an applied statistician.

Q: Do tell us about your journey on becoming an applied statistician. 

A: In university, I enrolled as an Arts and Social Sciences student. Back then, we were to take 3 different minors during our first 2 years before choosing our major in the 3rd and 4th year of our degree. I chose Maths as one of my minors because I loved the subject and realised that it was one of my strengths — even my teacher, back in vocational school, would discuss and solve different Maths problems with me. I also minored in Economics and International Studies.

When choosing my major, it was a no brainer — I chose Maths. After graduation, I realised statistics was my strength and I liked analysing data so I continued my Master’s in statistics where I researched specifically on solving road accident problems. 

During my PhD, I continued my research on a new modelling method, that many have not applied or worked with in Malaysia, for different areas like image processing, pattern recognition, and image quality analysis. 

As a lecturer and researcher in Applied Statistics, I started exploring more areas and collaborating with different experts in their respective fields. I believe we must make our research applicable for all, which requires us to collaborate with others in various fields, depending on the areas of study. Without cross-disciplinary research, we lack the expertise to go in-depth into the problem so I always try to continuously expand my network.

Q: So you’ve been interested in Maths ever since secondary school days and you’ve always been in the academic field. What’s one memorable moment in your academic journey?

A: It was while I was pursuing my Master’s Degree and tutoring at the same time. Back then, all the postgraduate students who did tutoring were either doing their PhD or Masters degree and we’d have a room dedicated for us. I remember all the postgraduate students scrambling to finish their work — some going past midnight and even staying the night there! 

At times, it was a struggle for me because there was no one else doing their research in the same field I did mine. So I always had to try very hard to solve the problems on my own before meeting my supervisor without having prior discussion with my friends. 

Even though we’d work on different projects, I felt that was an enjoyable time because the environment brought us together. We’re still in touch with each other even after many years.

Thinking of becoming a statistician? Find out how you can build your foundation in the subject here.

Q: What are some of your research that you’ve worked on?

A: My past and ongoing research projects cover financial and economic analysis, medical issues, image processing and pattern recognition, and social sciences. 

I’m also interested in research on Mathematics education and TVET (Technical and Vocational Education and Training). 

As an applied statistician, we need to head into cross-disciplinary research because we need to see how we can apply our methods to different areas and problems in the world. 

Taylor's Case Study: Inspiration Amidst Hardship 2

That’s why I’m always looking for the opportunities to apply my expertise in statistics and data science to solve real world problems.

Q: Speaking of your research, you’ve done one on the housing market. What made you work on that project? 

A: All statisticians would know that it’s really hard to get raw data. So when I went into this research, I'd a friend who was in the real-estate industry for 20-30 years before retiring. After talking to him, I found out that he had 20-30 years of raw data from his work and the best part was that he wanted to share and work with me! 

I was so happy to get this and suggested to my postgraduate student, who at that time was also looking for topics, to take this project up. During my PhD, I also came up with an advanced statistical technique called functional relationship model. 

With our methods, we’d normally find new ways to apply this method to different areas and problems in order to increase the value of the method. When I was looking at the housing market, I noticed that this model would be quite appropriate after some slight adjustments which were done by the student. 

To me, the most important thing was that I had someone willing to provide the raw data

Q: How does this research aim to help the society at large?

A: In the beginning, this project looked at the micro aspects of the housing market — like the location and area of the house. These are usually determinants for the buyer to decide the value of the house and whether a particular house price is reasonable or not within that area. 

From the project’s development, I realised that there was an untapped potential to solve macro-level problems too. 

At the macro-level, you’re no longer talking about a particular house value but the market itself and would relate to the GDP, employment rate, inflation rates, and other areas that influence the market itself. 

Taylor's Case Study: Inspiration Amidst Hardship 2

We realised that the data needed for this are time-series data which doesn’t fit into our current model. As of now, my research team is currently working to modify and extend the functional relationship model to fit the time-series data. 

This bigger area of data would help the policy makers and developers to see the future trend before making their decisions.

Did you know that there are two kinds of statistics? Theoretical statistics and applied statistics. Find out how both of them function here.

Q: As an applied statistician, what’s important for your research to have a greater impact on the related fields?

A: There are two aspects that are very important. 

Firstly, data accessibility and accuracy. For applied statisticians, data is very crucial because without it, you can’t do anything. On the other hand, when we do have this data, we must make sure it’s reliable and we do this through data cleansing. If your data isn’t ready, you can’t move on to the analysis stage. 

Secondly, statistical skills. Statistical software eases the difficulty faced by many people to generate statistical results — even if the researcher has no statistical background. However, the skills of choosing the appropriate statistical technique and ability to interpret the outputs are essential. I’ve come across articles where inappropriate statistical methods were used and wrong inferences were produced by the researchers. 

This is why statistics researchers would always need to collaborate with a subject expert to produce reliable outputs.

Q: What are some of the challenges you faced as a researcher in applied statistics and how can they be fixed?

A: Unfortunately, as I mentioned earlier, in Malaysia, it’s very difficult to get raw data and the data that’s usually easily accessible and available are second-hand data. If we were to get it from companies, it’d be very expensive. 

So, many researchers resort to collecting data from surveys but there’s also sampling design problems and a low response rate. This will ultimately affect the reliability of the research findings.

I do hope that the government can improve its current open data system to not only provide selected and processed secondary data, but also censored raw data to registered researchers. 

On top of that, also for the public to be more open and proactive in responding to questionnaires.

Aside from that, though many disciplines carry out their own statistics, most of the time it’s surface information where you’d need more technique, skills, and experience to get the in-depth information. 

One of the main solutions for this is to head towards a cross-disciplinary approach.

Taylor's Case Study: Inspiration Amidst Hardship 2

Q: What are some of the biggest challenges that we could face in the next 5 years and how can we prepare for them now?

A: When we look at mental health, these problems are masked and not obvious. The most important thing to do is identify these issues early on. The short-term impact is that people are worried, fearful, and stressed. But we hope that it stops there and not lead into depression because that can lead to suicide and losing loved ones.

Q: If you could give one piece of advice to your younger self, what would it be?

A: If you want to do it, just do it. Don’t wait. Research is uncertain so you have to start equipping yourself with the skills and then kick start it instead of waiting for the best solution to come about. If you can, do it as early as possible, especially when you’re still young before you have bigger commitments. Otherwise, there could potentially be a lot of struggle.

Q: Do you have any advice for researchers or anyone interested in taking up research?

A: Many have a misconception that researchers work alone and rarely mix with others. Nowadays, this is far from the truth. As a researcher, you must build your network connection. I realise that there are many of us who are passive when it comes to this but in reality we need others because we can’t always come up with new ideas on our own. 

You’d need to explore and discuss with people to get new input, to develop your skills and find the opportunity to collaborate so that your project can develop and shorten your process without sacrificing the quality of your research. You’ll never know one day your network would come to you for help on a particular problem. Like my friend in the real-estate industry!

Also, if you’re interested in the academic field, consider going into research for a deeper career development. Many opt for coursework-based and head out to grow a career in the industry. The misconception here is that research is very difficult and super uncertain resulting in research being overlooked. Though it has some truth, there are a lot of opportunities research brings about. 

So if you really want to do it, just go for it!

Interested in continuing your Postgraduate studies in research? Find out how Taylor’s can be the place to kickstart your research journey!

RELATED STAFF MEMBER
Ailyn Low
Editor
TAGS
STATISTICS
CROSS-DISCIPLINARY RESEARCH
MATHEMATICS

OTHER RELATED ARTICLES