Saturday, June 2, 2018

ST4199 review (NUS)

DISCLAIMER: firstly, i am not a professional writer, so if there are mistakes in my language, move on. secondly, the review is mostly based on opinions, so i would advise you to google for more reviews to get a diverse take on the modules you are interested in. thirdly, module details change every semester, so do not expect every assessment will stick. last but not the least, i will not talk about grades. hope this is useful for you, cheers.

this review will be slightly different since the topic of the project varies from student to student. in statistics course, ST4199 offers two different paths as well. one would be the individual FYP and the other is the group/seminar style. note that, even the group seminar/style, in the end the report writing, presentation and other assessments are done as though it is an individual project. meaning, in a group/seminar setting, you can discuss on literature review and what not but ultimately your submissions and what not are done individually. yes, your topics are similar but your "products" vary.


i have a rough idea how group/seminar path works but not in detail so am not confident to share on that as well. i took the individual FYP path and as such, will be talking about that.



assessments

for individual FYP, the assessments include report (2 drafts) and oral presentation, with Q&A. the report has two submissions, both submission is the entire thesis. the first would be around semester 2 just after recess week while the second would be after the oral presentation and Q&A, right before reading week. oral presentation is a 40-minutes session though if you take less than that is fine too.. just try not to exceed that. straight after the presentation would be the Q&A.



my project

my FYP was titled "Survival Analysis of Identified Diabetic Patients". the main few analyses included Kaplan-Meier estimate, Cox proportionate hazard model and computation of restricted mean number of years of survival. I had two awesome supervisors, Prof. Teo Yik Ying and Dr. Stefan Ma.



rough progress timeline

though the department required those doing individual FYP to submit the project details and supervisors' name by 3rd week of semester 1, i began finding supervisors since June and managed to confirm all the details by late June to early July. retrieval of my datasets took quite some time so between August to early September, i did tons of literature review and learning of the few statistical methods that were required and that i have yet to understand fully. 

to be honest, even for the remaining of September, i did mostly literature review. my data came quite late and i only started to do simple cleaning and merging of data so that it can be used for analysis with minimal issues in later stages. cleaning and merging thankfully took about 2 days before i head on to summary statistics and churning out simple results.


October came and yes, still some literature review and reading of research papers. i would say my project had three main sections of analysis/results.. with the last two sections having three sub-sections. let's call these sections and sub-sections A, B1, B2, B3, C1, C2, C3. so in October, i produced the results for A which mainly consist of summary statistics and other simple results. i began coding a little for B1 and C1.


in November, i completed coding and producing the results for B1 and C1. then i went on to B2 and C2. these were a little tricky so it took couple of meetings and tries to iron things out but i remembered by reading week i was near done with them.


in December, after relaxing a while after the final examinations of the first semester, i went on to B3 and C3. due to unforeseen circumstances happening elsewhere outside school work and in my personal life, the break wasn't as productive as i hoped. so this spilled to the first half of January. by middle of January, i had all results out and started to figure the interpretations and what not.


February involved a lot of edits after presenting the stuff to both my supervisors. started writing the backbone of my thesis in early February as well. spent the second half of the month mainly writing up my thesis and letting my supervisors to run through before editing again and again.


March was the submission of the first draft. then also started preparing slides for my oral presentation. by middle of March, my slides and speech were up and ready. practised with my peers tons of time. well, what came after this would be the oral presentation with Q&A and the final submission.



individual VS group/seminar

yeah i know some might be thinking which paths might be better. while certain ranges of CAP will put you in a spot where you cannot choose, there are some who can choose between the two paths. or another way is probably you are now aiming to score within certain ranges of CAP so as to grant you a particular path. nonetheless, i will try my best to weigh out the pros and cons of the two paths to give a rough idea to you all.


time management. notice that all the deadlines for individual FYP happen in the second semester so technically no one would be pushing you or setting a pace for you specifically to follow. good time management and proper planning is needed. then you also cannot set the timeline same as other individual FYP since the nature of most individual FYPs varies. some might face a slower data collection or data extraction process while others have to take a slower pace in other aspects. on the contrary, the assessments for group/seminar FYP are spread out.. so you roughly get the sense of urgency now and then. professors taking students on this path are more likely to check on you and push you now and then too.. well, apart from a few lazy professors.


meetings. on this aspect, group/seminar FYP students generally will meet the professors more often as they will set up say bi-weekly sessions and what not. i think such platforms serve as a good place to clarify doubts efficiently. individual FYP wise, no meeting also can but of course that is for extreme case in which i have never know someone who has done so. however, for individual FYP student, the professor will usually agree to meet if he/she is free.. bluntly speaking, you are not the priority. so time between each meeting could be up to a month or so. i'm sure you can hear some horror stories elsewhere. thank goodness i have awesome supervisors.


freedom of choice. individual FYP tend to have greater freedom of choice.. be it topic wise or supervisor wise. i get to choose who to find and ask if he/she could be my supervisor. of course start early since any professor cannot take a lot of students in an academic year. topic wise, i had a little say as well.. highlighting my interest in biostatistic and/or public health. for group/seminar FYP, you get to rank the professors you want but more often than not, you cannot get the first choice. topic wise, those professors have set the topics even before your selection of professors begin. 


running into trouble. this is where group/seminar FYP is the obvious better option. while reports might vary.. background topic can be similar.. so anything unsure you have a few others to discuss with you. for individual FYP, for obvious reasons you are on your own if you run into a mini hiccup or what not. yeah your supervisor might hint you but they wouldn't usually suggest a solution.



my opinion

i think ST4199, regardless of the path you take, is a really good exposure to research work. if you think you are definitely not going into the statistic or analyst or research line.. year 4, and thus FYP, might not be your cup of tea. if you decide to got for FYP, know what kind of project you probably want.. like roughly the "industry" and what not. be it individual or group/seminar FYP, then go ahead and choose professors based on their interest of research and topics offered. 

ohh which reminds me of a common misconception. difficulty. both paths are difficult.. come one FYP sure got tons of challenges. that is why it is important to choose properly.. i am sure it is less painful if you are doing something you prefer.

actually, i think time management is important in both paths.. just one might be more reliant on good time management. this module is clear very different from any other modules. last minute work or cramming will never work out fine. nonetheless, embrace this opportunity and whatever challenges it may pose.. for the finish line is sweet.

click here to see the modules i have read and their respective reviews. do feel free to contact me via Telegram @alvinngjh for any questions or what not.

3 comments:

  1. as for individual project, how do you find the prof and the topics for the project?

    ReplyDelete
    Replies
    1. hmm first know which areas of statistics you want to explore in then email them based on the research interests of the staff members listed on the faculty website.. don't be afraid to contact a few at one go or even meet up in person to discuss.. your main supervisor has to be someone from that list and depending on the project, you might or might not need a co-supervisor

      for me, i knew i was interested in biostatistics/public health/epidemiology.. there isn't many faculty members who have those primary interests.. topic wise, he/she might suggest a few to let you choose or just introduce just one.. this really depends on your supervisor

      Delete
  2. hey there, have you heard what grades (group seminar) students usually get for their FYP? is it very very hard to get an A- at the very least?

    ReplyDelete