FAQ about the Master of Science in Statistics programme in NUS

Category: [Machine learning & Statistics]

Tag: [MSc Stats]


6min read

It’s been a while since I last worked on this blog, but today (Sunday!) I have some time in my hands to write a little bit here.

Prior to this, I have been receiving more questions about the MSc. Stats programme in NUS - the timing makes a lot of sense since it was just before the start of the new Academic Year in NUS (AY20/21).

In this post, I have collated a number of questions that I have received, most rephrased to be more generic, as well as my responses to them. Moving forward, I may also update this post as I receive more questions.

Before I continue, I would like to sincerely thank my readers for reaching out to me and asking these questions. I hope that my answers were helpful for you in your decision-making and I hope you excel in the programme. I also hope you don’t mind me sharing these generic questions to a wider audience.

And for those of you starting the new AY in NUS - hope you have a good year ahead!

On my background

Q1. May I know what your educational/professional background is before the MSc?

A1. My degree was in biostatistics before the Msc.

Q2. What field were you in before joining the programme?

A2. I was and still am in the data science field.

Q3. What made you want to join the programme?

A3. I was a few years into the workforce and wanted to go back and study something. Stats happened to be the area that I am most interested in.

On the programme in general

Q1. Does the programme require knowledge on programming languages for data science (R, Python)? Does it teach any programming languages for data science?

A1. There is no emphasis in programming in the Stats masters. Some modules may require you to code in R, but not many, and not much within those modules.

Q2. Is the programme grueling?

A2. I found it manageable, but the start was a little challenging, getting used to studying again. And certain modules are more theoretical in nature, with some bit of theorem proving.

Q3. Would you recommend the programme?

A3. Yes definitely. I found it useful for my work, and managed to learn what I wanted to.

Q4 .Do you find the statistics content that you have learnt useful for your current job?

A4. I would say yes. I work in the data science industry and it’s always useful to fall back on fundamental statistics principles to think through certain issues and problems.

On admission

Q1. Do you know the competitiveness of the program? Is it hard to get in for NUS grads?

A1. It really depends on your background prior to the MSc. Competitiveness in terms of getting in may not be difficult, but coping in the workload may be so, both for full time (5 msc modules) or part time (2-3 msc modules + a day job). Also, having an honours would help as you can be in Track 1 straightaway (the shorter track with 40MCs).

Q2. Do you know around when you got your results of the admissions?

A2. I got my offer letter via email sometime near late May during the year I applied.

On modules and workload

Q1. Do you have any advice on which module combination to take in each semester? (MOE subsidy for AY2020/21 requires students to complete the required modules within 2 years for part time track)

A1. I would recommend finishing the two core modules of 5201 and 5202 as soon as possible. It’s hard to plan ahead of time for the MSc in terms of modules because modules offered can differ significantly from academic year to academic year. It depends heavily on lecturer availability and some modules only appear once in 4 semesters I believe.

Q2. What is the difference between “Coursework” and non-coursework? Seems like some non-coursework modules are also offered at 7-10pm for part-time students.

A2. For the “Coursework” tag, it is used by the department to tag whether it’s meant more for the MSc Research students as opposed to 100% course work students. Technically the non-coursework modules are supposed to be harder - but not always the case. And also whether it’s possibly taught during working hours vs. 7-10pm.

Q3. For most of the statistics modules that you have taken so far, is it true that they are very theoretical, with majority of the tutorial questions based on proving theories rather than application questions?

A3. It heavily depends on the nature of the module, and more importantly the lecturer. E.g. the same topic on regression can be taught in both heavily theoretical and more applied methods, depending on the lecturer’s style. And some lecturers are “renowned” for being theoretical no matter what topic they are teaching. That said, since these are all stats modules, proving-type of questions are inevitable. I would recommend finding out more about the lecturer’s style.

Q4. For the ten modules that you have taken, was the workload and difficulty level for those modules somewhat similar to ST5201?

A4. Yes, largely the same with minor variations.

Q5. On average, how many hours per week do you think you spent on your lectures and tutorials?

A5. I typically spend 1 Sunday (morning + until before dinner) and ~1 to 1.5 weeknights per week on catching up on lectures and working on tutorials, on top of the 7-10pm lectures.

Q6. Were there a lot of group projects or assignments?

A6. Not for me, usually stats mods in NUS don’t have many group projects. I only had one with ST5227 applied data mining.

Q7. How was the bell curve for the statistics modules? i.e. Were the exam papers rather simple but with steep bell curves or very difficult with shallower bell curves?

A7. Again this heavily depends on the lecturer again. The variation in style, content level, exam format etc in the MSc is larger than that in NUS undergrad courses, because there isn’t really a tight oversight from the Faculty or Registrar, as compared to undergrad courses. I think I have equally sat in both steep and shallow bell curves.

Q8. Do you think reading lecture notes and doing tutorials alone is sufficient to score a B+ for most of the statistics modules?

A8. Doubt so. For the majority of students, B+ is a respectable 2nd Upper standard. Reading lecture notes and doing tuts can warrant a pass (C/C+/B-?) probably. I would say attending lectures, reading the recommended textbooks (just very briefly on some of the harder topics) and consistent revision are important. If you are a part-time student, consistency is even more important because we would only have some portions of our week allocated to the MSc, and on top of work and life. Also, the tutorials themselves can be very short (4-5, at most 6 questions?). It’s not enough to assess yourself to see whether you have a holistic understanding of the chapter.

Q9. Of all the statistics modules that you have taken so far, is there any that you would not recommend to take?

A9. None I think. I enjoyed my modules. The hardest part for me was at the start, taking ST5201 during the first semester. It was relatively theoretical plus the fact that I was a little “rusty” in the theoretical stuff, plus the need to get used to studying again. After that, it got a lot better and more manageable.

Q10. Are you able to share some materials from the MSc?

A10. Sure, write to me at my gmail (see About) and we will see how I can help you.