AY2014-2015 Semester 1 Module Review

I am so relieved to have overloaded this semester and cleared 6 core modules at one go as I just realized my schedule will only be getting increasingly more hectic in the remaining semesters. On one hand, the modules I intend to take in the subsequent semesters are a lot more difficult and time-consuming and on the other hand, having to juggle work on top of studies and meeting the deadlines set by the professor isn't something easy for me.

But anyway, this semester wasn't as tiring as I had imagined it to be. Rather, I thought it served its purpose of preparing me for even tougher semesters ahead. Nonetheless, I am contented with my SAP this time round and hope to be able to maintain such results for the rest of my candidature.

EC3304 Econometrics II

This module was taken by Eric Fesselmeyer. I thought he had poor planning in the sense that he spent almost 3 lectures only to cover the first chapter which was basically just a review on the concepts taught in EC3303. After he came to realize how much time he had left before the midterms, he started to rush through the next two chapters so that he can include both of them in the midterms as well. And when midterms was over, he took his time on the following chapter only to rush through the last two chapters again. In fact, he cancelled the lecture in Week 11 or so and made up for it with a webcast that I guess exceeded the usual 1 hour 35 minutes (I am guessing as I did not watch it). If he had not done so, there was probably no way he could have finished what he wanted to test by Week 13. What's more the syllabus he showed during the start of the semester included one more chapter on "Quasi Experiments" but due to the lack of time...  By the way, there wasn't even a tutorial on the last chapter.

Anyway, I don't think there's a need to recommend whether you should get the textbook as chances are you might already be holding onto it considering that it's the same textbook used for EC3303.

Topics covered before midterms
Review
Introduction to Time Series Regression and Forecasting (Chapter 14 of the textbook)
Dynamic Causal Effects (Chapter 15 of the textbook)

Topics covered after midterms
Cointegration (part of Chapter 16 of the textbook)
Panel Data (Chapter 10 of the textbook)
Regression with Binary Dependent Variable (Chapter 11 of the textbook)
Instrumental Variables Regression (Chapter 12 of the textbook)

I have to say the proofs that Eric does during lectures are a lot clearer than those in the textbook. And yes, proofs are important as there was one 35-marks proving question in the finals which was the differentiator. And well, he said things that he showed using the visualizer which include the proofs will not be in the webcast but apparently, he doesn't have a good idea of how the webcast works so I actually skipped the lectures after the fourth one and I'll watch the webcast just for the part on proofs cos' other than that, the lectures are pretty much useless.

The workload for this module is very light as tutorials are not weekly but fortnightly instead. His tutorials include application questions that require the use of Stata and I personally think that these questions shouldn't be neglected. There are questions in his papers that will show students the output from Stata and answers will have to be inferred from there. But the point is these questions will give you a good idea of how the concepts learnt in class can actually be applied in real life and you tend to better appreciate this module after doing them.

Weightage
Tutorial attendance and participation: 20%
Midterms: 30%
Finals: 50%

I thought the first component was absurd cos' this module only has 5 tutorials and you have to present just once. Most people will be very 'enthusiastic' in presenting the solutions for the first few theoretical questions but when it came to those questions on Stata, there would be a sudden silence in the class. So I would think that it'll be good to prepare those questions at the back so that you'll get your chance to present. My tutor, though very approachable, can't teach for nuts and yes, I seldom say that so the tutor has to be really quite bad. I would have loved to skip the tutorials if I could as it wasn't value-adding in any way. Due to some reason or another, I did not attend one of the tutorials and even though I went for the makeup in the odd week, my tutor actually forgot to mark my attendance so that I had to go to Eric personally but now I've doubts as to whether he marked my attendance for that week as well.

Ironically, the emphasis in midterms was on stuff covered in EC3303 and I honestly do not admire the questions set by Eric or probably that's because I've taken ST3131. Specifically, there are 3 questions on the 'Review' part and 2 questions on 'Time Series'. So if you are good at the stuff in EC3303, you'll probably do well for the midterms though if you suck at the subsequent chapters, chances are you'll still screw up the module eventually. I do know of someone doing very badly (way below average) for the midterms but still got away with a decent grade and also, another person doing very well (way above average) for the midterms but got a not-too-decent grade eventually. Median was, if I did not remember wrongly, 63/100. I got 73 which was exactly the 75th percentile. Anyway, Eric takes a pretty long time to mark the papers; the papers will be ready for collection at the general office probably two weeks after the midterms.

There were 4 MCQ questions in the finals and each was worth 5 marks. I got the first question wrong which I believe could be the contributing factor to my grade on top of the tutorial attendance issue. But really, finals was manageable; just know what you are doing and you should have no problem in scoring well. In fact, the last question on 'Binary Dependent Variable' was a giveaway provided that you made no careless mistakes.

Result: A-
This was one disappointing module as I was hoping for an A. But overall, I did enjoy this module to some extent as the concepts covered were pretty relevant (though simplistic).

EC3332 Money and Banking I

This module was taken by Seet Min Kok who is a new professor. He is very helpful as he will answer your queries very promptly. I did not attend his lectures at all (not even the first one) as his notes are just a summary of what's covered in the textbook. In fact, I did not print any notes too; I just used them as a guide to what's covered in this module. And one thing about this module is that there are too many assumptions behind the concepts and when I asked the lecturer about them, there are times which he had to agree with me or that he couldn't give me a satisfactory answer. 

I recommend getting the textbook as it is a lot more detailed. Well, I was told by the lecturer to get the 10th edition as he will take questions from the textbook for the tutorials and so I stupidly heeded his advice and bought the expensive book only to see that I can count with one hand the number of tutorial questions he referred the students to the textbook. Anyway, I am not going to list the topics covered in this module as it is very long and more importantly, the syllabus is already written very clearly in the IVLE workbin. I actually think that taking this module with EC2102 will give you a bit of an advantage as there is some overlap.

Weightage

Tutorial attendance and participation: 20%

Midterms: 30%

Finals: 50%

As usual, the first component is absurd though not as bad as EC3304 since the tutorials for this module are on a weekly basis. I think your participation marks is contingent on which tutor you get. For my tutor, I participated only like 3 or 4 times and he mentioned he was very generous with participation marks so he usually gives every student full credit.

Midterms was one of the stupidest papers I have taken in NUS. Yes, strong word: 'stupidest'. Why? Because it is a 30-MCQ paper which tests you on memory and basic concepts. So if you did not read the textbook properly, chances are you'll not do well like me as I only got 25/30 and most people got 27. Well, I could have gotten 26 if I did not misread one question but 26 is still bad enough as that means 4% is instantly gone. At that point in time, I thought the bell curve was insanely steep as this module was really a breeze-through such that if I needed a break from my other modules, I would read this module's textbook for leisure.

Finals was again really easy; there were 7 structured questions and I am guessing you can answer them at any length you wish as students are given a booklet. Some people left early so that's quite indicative of how fast you can finish the paper if you wish. There were even a few questions which was directly taken from what's covered in the notes and textbook. I felt like this was a paper of "who can explain better" rather than "I know my stuff better than you". 

Result: A

I was taken aback by the result so apparently, I might have overestimated the bell curve quite a bit as I was expecting at most a B+. But I was one of those people who did not have free time on hand during the finals as I answered every question in a detailed manner; listing down all the assumptions, explaining how the theory came about, etc. I actually recommend doing that because most people found the finals an easy feat so it was not about whether you can do the question and from my grade, I think finals will still be able to make a difference if you take the paper seriously as much as the questions make you not want to.

EC4102 Macroeconomic Analysis III

This is easily one of my favourite EC modules so far simply because of the introduction of mathematical concepts into the content despite these concepts being very easy, perhaps not even reaching that of a level-1000 MA module. I suggest that instead of figuring out the integral and summation signs around, I think time will be better spent on appreciating the concepts behind the equations. And rumour has it that EC4102 is the hardest EC module around but I can say for sure that the difficulty level is really not like what people make it out to be. It might start out hard or at least it did for me but as you get more comfortable with the content, you'll find the subsequent chapters becoming increasingly easier. The only difficult thing is probably the mere number of people taking this module and the fact that only honours students are left cuts down a lot more leeway. Although this is supposed to be a macroeconomics module, there is extensive use of microeconomics in getting the macroeconomic picture. They term this as 'microfoundations'. Basically, you have to analyze the behaviour of consumers and firms in order to derive the performance of the economy as a whole.

This module was taken by Aamir Hashmi. He is a very good lecturer face-to-face. By that, I mean if you look for him to ask questions, the amount of patience and clarity he has for you is wonderful but you might be slightly put off by the replies he gives if you're talking about emails. Anyway, there are times when students ask him questions and he'll be like "I don't actually know the answer to your question". I actually like his honesty as I've met some professors who smoke their way through students, thinking students would actually believe them. 

The textbook for this module is 'Advanced Macroeconomics (4th Edition)' by Romer. I strongly recommend getting the textbook for this module despite it costing a bomb. For frugal people like me, I believe you do know that there is some way or another to get the textbook without purchasing it. Perks of doing so are the larger font size and more importantly, the ample space for scribbling. But some people might think that his notes are good enough. In fact, I do think that for some chapters, his notes are even better than the textbook. As for tutorials, even though the questions are taken from the textbook, he will also upload them onto IVLE. By the way, given that this is a seminar-style class, tutorials are not given out systematically. Instead, only when Aamir thinks that he has finished what should be done will he upload the relevant tutorial questions. 

Topics 

Ramsey-Cass-Koopmans Model (Chapter 2 of the textbook)

Overlapping Generations Model (Chapter 2 of the textbook)

Endogenous Growth Model (Chapter 3 of the textbook)

Romer's Model (separate paper will be uploaded)

Real-Business-Cycle Theory (Chapter 5 of the textbook)

New Keynesian Models of Incomplete Nominal Adjustment (Chapter 6 of the textbook)

Dynamic Models of Incomplete Nominal Adjustment (Chapter 7 of the textbook)

Well, I guess the good news is that the finals is toned down a lot after 'live feedback' from students several semesters ago whereby there were some people who cried while walking out of the examination hall. I actually enjoyed the past papers Aamir set as it makes you feel that this is indeed worthy of a level-4000 module. 

Weightage
Tutorial attendance: 10%
Midterms: 30%
Project: 20%
Finals: 40%

The first component is free as long as you attend all the tutorials. Midterms was disappointingly easy; anyone can score well if they don't make careless mistakes (though it's pretty hard to) and have studied for the test. How the grading works is that everyone's marks will be standardized and you'll get a grade accordingly. Highest was full marks and average should be around 23/30. I got 27 together with nine or ten other people. He'll upload everyone's scores in a PDF file within a week after the midterms so that speaks quite a lot about his efficiency.

Without a doubt, the stupidest component of this module has to be the project. Basically, every group gets to draw a paper and the task is to come up with a 1000-word non-technical summary on the paper you've drawn. This is worth 15%. Then there is also a 10-minute presentation which makes up 5% of your grade. Not everyone presents at the same time though as it depends on which paper you get. The presentations are spread out over the last three weeks of school. The bell curve for this component is really steep but the unfair bit is that if you get an easy paper like I did, he will mark you more strictly. One thing to note is when he says non-technical, he really means it. Anyway, he turned this component into a summary because he released a freestyle topic when he first took this module whereby students just have to sieve out articles and apply the concepts learnt in this module. However, it was somehow too difficult and so Aamir decided to remove it and change it to something at the other end of the spectrum.

Finals was non-cumulative but was again too easy; straight from textbook, simple manipulation... You name it, you have it. While the median used to be a failing mark in the past few semesters that he took the module, I think the median for this semester should be pretty close to the full mark. Aamir claims that if you do sufficiently well and he feels that you deserve the grade, he'll give it to you. Nevertheless, I may have overstated the ease of the finals as I know of people who lost over 10 marks in the finals and still got away with a decent grade.

Result: A
I was expecting an A- since I screwed up the project component and even though I believe I should have gotten close to full marks for finals, I'm quite sure everyone else did too. I think it is important to attempt all the derivations again before the finals and ensure that there is no room for careless mistakes during the finals.

MA2311 Techniques in Advanced Calculus

Stats majors get to choose to take either this module or MA2108. I personally think that this module will be more useful for Stats majors as it is important to know multivariate calculus. And this module makes it simple enough such that it does not improve any proving as things like Green's Theorem wasn't introduced plus there is no vector calculus. So only computation of multiple integrals will be tested. Besides, the focus of MA2108 is unnecessarily abstract for Stats majors. But if you have talent for that and you're in need for a CAP puller, then take MA2108 by all means. Nonetheless, because MA2311 is the only opportunity for Stats majors to be formally exposed to multivariate calculus unlike for Maths majors, I still hold on to my stance. Anyway, there could be certain Stats electives that may require double or triple integrals (though I don't see them being offered anymore).

This module was taken by Ji Feng. He is excellent, period. Firstly, he is one heck of a good teacher; his lectures are wonderfully conducted as he will take time to write down the steps for his examples on the board. Secondly, he is really helpful; he can get back to you within 15 minutes of your email 99.9% of the time. Thirdly, I had him as my tutor as well and he is very clear in all his explanations. Fourthly, the syllabus for the entire semester was very well-organized in the sense that he could keep a steady pace throughout the semester and still managed to keep Week 13 free. Lastly, his lecture notes have sufficient examples and on top of that, there is often additional information provided during lectures. Lectures aren't webcast by the way.

Topics

Sequences

Series

Vectors

Functions of Several Variables

Multiple Integrals

In terms of quantity, this module is really slack but in terms of quality, maybe it's because I'm not inclined towards the subject but I actually spent a considerable amount of time on this module though I began to neglect it after Week 10. Nonetheless, because this module precludes those who took MA1505 or MA1104, it cuts down the competition by more than half already. However, I must say that the papers set for this module would be an eye-opener if one simply took the tutorials as a gauge to the level of difficulty of this module.

Anyway, there is no textbook for this module though I would recommend 'Thomas' Calculus' or 'Calculus' by James Stewart. I think either one of them should be the same as the textbook used for MA1102R. In terms of content, I think it's pretty redundant to refer to the textbook as the notes are good enough. But I think there are two things in the textbook which are pretty useful: first is the 3d spatial visualizations provided and second is the practice questions. One thing to note about the practice questions in the textbook is that the questions for the first two chapters are very elementary and I don't think they can prepare you well enough for the midterms or finals but for the last chapter, I feel that if there are times when you find yourself not understanding what the lecturer is saying, then doing the practice questions will help a lot.

Weightage

Tutorial attendance and participation: 10%

Midterms: 30%

Finals: 60%

The first component is again, free; just present once and attend all tutorials. The coverage of midterms only includes the first two chapters but because I suck at series and sequences, I screwed up the midterms. The best part was I didn't think I'll screw it up that badly. So the median was around 13/30 and I got only 17. Everyone was cursing the paper after the test in Week 8 but truthfully, it was very doable even though I believed the standard of the paper has increased since the last semester this module was offered as the median was around 20/30 back then. As different as the questions may look compared to those in the tutorials, the concepts required were essentially the same (of course). As typical of a mathematics paper, it just depends on whether you can use the appropriate concepts to solve the problems. Anyway, Ji Feng takes a pretty long time to mark; I think our midterm scripts were returned to us only in the tutorial of Week 10.

Finals was cumulative (obviously since there are only 5 chapters) and I thought the focus was on spherical coordinates as bulk of the marks came from that sub-topic. The thing about vectors and multiple integrals is that most people take the formulas for granted and simply neglect the importance of visualization. Being really bad at visualizing in three-dimensional space, I struggled through the last three chapters but I guess it did help make up for all my lost marks on the two finals questions on series and sequences. As long as you can visualize how vector lines or shapes intersect in 3d space, half the battle is already won.

Result: A-

I was expecting at most a B+. I can't say I got an A- because I could do the questions but it was more of the bell curve that saved me. But I think the most important thing is to understand all the concepts; blindly practising questions will not help at all.

ST1131 Introduction to Statistics

Can I just skip this module? The reason is because I never attended even a single lecture or tutorial and I can count with two hands the number of hours I spent on this module in the entire semester which explains my final grade. Ok, there are a few things I can say: given that the content is insanely easy, the bell curve will also be insanely steep especially with ISE majors around. 

Tutorial questions are taken straight from the textbook. By the way, I don't recommend getting the textbook as the notes are just rephrased from there and will be sufficient for revision. Tutorials require the use of Minitab, a software that I've never bothered to use. I mean why use Minitab when you have R right? In fact, I don't find tutorials useful and I gave up after the fourth one.

Weightage

Midterms: 30%

Finals: 70%

Midterms, a 20-MCQ paper, was really stupid with meaningless qualitative questions which were not flattering of Statistics at all. One thing for sure is that this module doesn't give a good indication of what Stats majors will be in for in the following semesters. It just picks a bit from ST2131, a bit from ST2132 and a bit from ST3131 and the bits come together to form one big module. It'll be completely understandable if halfway through, you find yourself questioning where this or that came about because the depth of this module is really... 

The thing about the finals this semester was that the questions seemed a bit open-ended instead of the usual physical science style whereby it has to be this and this only for the answer. There was even an unnecessary question testing on understanding of output from Minitab. 

Result: B+

This result was totally within my expectations as I know how I fared for the finals; feeling sleepy halfway through and doing at 1/10 of my usual pace just says it all. I think even the sloppiness of my review says it all.

ST2137 Computer-Aided Data Analysis

Basically, students will learn 3 softwares: R, SAS and SPSS (you may take the order as a rank of the 'intelligence' of the software beginning with the most intelligent to the least). Most people complain about why we even have to learn SPSS, a software that involves zero coding but clicking instead, when there's R and SAS around. Moreover, the fact that SPSS was totally neglected in the last few topics goes to show something about the software.

This module, for a change, was taken by Lim Chinghway this semester. As good as he is, I don't like the questions he set for the midterms and finals simply because they are not indicative of a supposed programming module. Instead, bulk of the marks came from familiarity with the syntax of the three software through T/F questions. I also skipped the lectures as the thing about programming is that it has to come from your own intuition and understanding. But maybe that is an overstatement given the level of difficulty of this module.

Topics

Introduction to SAS

Introduction to R

Introduction to SPSS

Describing Numerical Data

Robust Statistics for Location and Scale Parameter

Analyzing Categorical Data

One-Sample Tests and Two-Sample Tests

One-Way Analysis of Variance

Regression Analysis
Simulation: An Introduction
Simulation Studies in Statistics
Bootstrap Estimation
Numerical Methods in R

Weightage

Tutorial attendance: 10%

Midterms: 20%

Project: 20%

Finals: 50%

Bonus (IVLE forum participation): 5%

The first component is free. Midterms had 10 T/F questions which served no meaning at all. I mean when you are programming, is there a need to care about nitty gritty details with regards to the syntax? You'll just type the lines and when you realize there's an error, you'll change it and find out later that your code is able to run. I was actually at the 25th percentile with only a score of 35/50 and the median was 41-42 while the 75th percentile was 45. And this just shows how steep the bell curve for this module is. The reason why I screwed up so badly is because first, I lost quite some marks in the T/F questions which was worth 2 marks each and second, I misinterpreted one giveaway SPSS question at the back. Anyway, midterm scripts will be returned during the tutorial two weeks after the midterms but the scores will be uploaded onto IVLE beforehand.

Although the norm is to take this module when you're in Year 2 Semester 1 of your candidature, I strongly recommend to take it only after completing ST2132 and ST3131. It will not only give you an edge over others in terms of the speed at which you understand the content taught here but also, it will allow you to appreciate the content a lot better. Moreover, ST3131 is able to expose you to different methods of regression analysis that are not covered in this module which will allow you to attempt unconventional ways of doing your project like what I have done with my group. Basically, students are not limited in the scope of their project; they can do anything under the sky but you have to prove that you are able to make use of the concepts learnt in class. Students are also free to use any of the three software they wish. Bonus marks will be given if you can come up with something that has not been covered before in class. By the way, credit is given for creativity. The only thing to complain about this project is the limit of 8 pages with reasonable font size but of course, there are ways to get around that using some tricks.

Finals had like 20 T/F questions?! It was absurd yes, 40 marks on that. Well, other than that, finals actually focused quite a lot on theory as the lecturer decided that theory is important to be able to carry out computational analysis which I do agree to a certain extent. But the finals was so easy such that I felt it was impossible to pull up my midterm results. And time was no issue; many people left early.

Result: B+

I did very well for the project and finals which I guessed pulled me back up to a B+ as I was expecting a B- or below. So maybe the bell curve wasn't as bad as I thought. I think if anyone happens to take this module under Lim Chinghway again, know the codes like the back of your hand and I don't actually think that deep understanding of codes is required or at least it won't be tested in either the midterms or finals.