ECON20003: Quantitative Methods 2

3 assignments weighted 5% each, a mid-semester online test weighted 5%, tutorial participation worth 10%, final exam worth 70%.

So instead of studying for my next exams I’ve decided I’ll review QM2. Some of you may recognise my username from my favourable review of QM1 and Jonathon Thong (maybe if I was in this years cohort for QM1 it might’ve been different…) so it’s not like its unexpected when I say that QM2 was good.

Firstly the old argument of introductory econometrics (IE) versus QM2. We were shown a slide at the start of the semester saying how IE is primarily regression based and involves a more depth of knowledge whereas QM2 focuses on regression for only part of the course and involves a more breadth of knowledge. As the semester progressed I was pretty happy I did QM2, honestly the regression in IE looked scary (from peers showing me), whereas QM2 really didn’t provide anything ridiculous. I can’t really provide much more information on the choice between IE and QM2 but honestly I personally am glad I chose QM2. I have been told by friends that IE isn’t hard if you put the effort in, whereas I categorise QM2 as one of the easiest subjects I’ve done and definitely the easiest in my semester (other subjects were corp law, cost management, and ob)

Ok so I’ll start with the lecturer. Reza! The main man. Honestly my opinion of him is pretty good, he’s pretty chill. Coming into the first lecture everyone was a bit disappointed as Reza speaks in a very quiet tone with not much expression. There isn’t much to say about him but I probably did go to 20 out of 23 lectures and he did get the content across effectively (in most circumstances). Don’t be too put down that you have some boring lecturer – the best thing I did was listen to lectures at 2x speed, which was pretty effective actually. The lectures themselves don’t pile on the content crazy quick, so they’re pretty good. Reza will generally just read out the lectures in most cases, however sometimes he didn’t – but you could easily get by without going to any lectures and just reading the slides (I do commend him for very good slides).

Next, the tutors. I had James as a tutor and honestly if you can, choose him. He also was our online tutor for the semester so it didn’t provide too much of a boon to have him as your tutor but he really gets the content across well. He definitely knows his stuff and will help you if you get stuck on an assignment if you ask. I can’t speak for any other tutors but James was and currently is the best tutor I’ve had at Melbourne uni. The tutorials themselves are quite easy; a set of questions to be answered before class and a set of questions that are answered during class. The homework set (to be answered before class) is collected during class, and THIS IS THE ONLY TUTORIAL MARK. Unlike other subjects where you are required to contribute, QM2 only requires you to hand in the pre-set work. Answers to both sets of questions are posted at the end of the semester (I think the first half answers are put up at the end of each week) and these questions really help for exam revision.

The content itself is probably a little bit dry (its QM2 like what do you expect) but I enjoyed it. The first 3 (or 4?) weeks are simply constructing hypothesis tests. That’s it. Don’t slack off in these weeks though, I found it pretty helpful to remind myself of the knowledge I gained in QM1 to make sure I had a strong foundation going into future weeks – but honestly, not much is required of you in these weeks. Come exam time all you need to do is memorise the types of tests (i.e what type of data you need, is it normally distributed) and you’ll be set. After these weeks, regression starts with Ordinary Least Squares. About 60% of the marks on the exam are dedicated to regression, with 20% dedicated to the hypothesis tests and 20% on time series. The main problem I found with regression probably just had to interpret different types of variables. It isn’t really required at all in QM2 to do any outside research or anything, just constantly reminding yourself to do it the same way the lectures do it is all that’s required. Reza also gave us a support lecture slides which helped us develop that skill pretty well. I don’t have too much to say about regression honestly. It’s pretty cool I guess, and I was glad we didn’t have to learn any proofs. In SWOTVAC I found most of my time on regression dedicated to just miniscule things (like ensuring you get 100% in a question rather than 98%) because I find it’s the small things that add up, like using percentage point instead of unit in your interpretation. Lastly, time series is introduced in the last 2 weeks of the course which proves to probably, in my opinion anyway, be the hardest part of the course. Time series is pretty cool itself, and the lectures were well structured until the last lecture of the subject. In this lecture Reza introduced random walks and non-stationary time-series which really require more than a lecture to understand. I found my self in a continuous exchange of emails with my tutor, and constant youtubing to develop some sort of comprehension of non-stationary time series. This didn’t really help and sort of just left me with more questions than answers. Anyhow though, few marks on the exam are dedicated to this and the rest of time series is a walk in the part. Overall the content is relatively straightforward provided you maintain concentration during the semester.

QM2 marking is pretty good. There are 3 assignments during the semester each weighted 5%, combined with a 5% mid-semester online test, and 10% for tutorial participation (handing in your work), with 70% for the exam. The assignments are pretty extensive for only 5% but I found them really useful for my understanding. It’s pretty easy to get full marks for the assignments if you put the effort in, and go to consults if you find problems! The tutors will help. The mid-semester test I believe was 15 online questions in 30minutes, which again, wasn’t too hard as it was only conducted in week 4 or 5, where the only real content we had learnt was hypothesis tests. Again, the tutorial participation should really be a free 10%! The tutors don’t care if you get the questions wrong, only if you show an effort, so just do them. It was quite common to go into the exam with 25/30% of the marks available or higher. Oh and the assignments are to be done in pairs or alone.

Lastly, the exam. Honestly it was just a general rehash of the assignments. If you were familiar with the assignments then you would recognise the exam. The only problem I faced was that the random walk/non-stationary time series crap I yapped on about earlier, constituted 7 marks of the 70 available on the exam. So it was quite rough considering it was on one lecture and probably the hardest concept overall in the subject. Other than that the other 63 marks should have been easily attainable (I think…haven’t got my mark back yet).

Overall, QM2 was a lot better than QM1 hah.

Yes

Reza Hajargasht

Yeah, we were given a 25 page file completely decked out with past exam questions.

4.9 Out of 5

That SSK textbook from QM1. Not needed at all.

Two one-hour lectures and one one-hour tutorial per week

2016 Semester 1

TBA

5% Mid-Semester Test (Online); 15% Assignments (2 best of 3); 10% Tutorial Participation; 70% 2-hr Exam

The content of QM2 spans a good range of topics, from parametric and non-parametric tests to linear regressions, logit models, forecasting, and ARIMA. This subject gives you a very good overview of different statistical techniques and emphasises when to use it and how to use it (at least on E-Views). This subject focuses less on calculations and more on analysis and interpretation of results - and I actually like it that way. Some people dislike the fact that sometimes 'there is no one right answer' despite it being a quantitative subject, but being an accounting & finance major has probably increased my comfortability with vagueness.

What I really really dislike about this subject is that I felt like I was in a computer class during tutorials. I was just being taught to click some buttons, type in some numbers, produce some charts, etc. The most 'analysis' we do during tutorials were just 'interpretation' of the results. In the end, we don't get to sit in front of the computer during exams like we did in all tutorials. This contributed to the students' feeling of unpreparedness. Also, unfortunately, QM2 lectures fail to give us adequate knowledge to prepare us for our assessments. There's always the feeling that you don't know enough, or worse, that you don't know anything at all. During SWOTVAC and consults, you can pretty much feel/see/hear that everyone's pretty tense and confused and bursting with questions that should've been answered had Joe (and other tutors) been clear in teaching the content. I cannot remember how many emails I've sent to my tutor, or how many questions I've submitted on the Online Tutor, during my revision. Probably the reason I got H1 is because I did some intense self-study sessions to help me get prepared, because unfortunately, the lectures and the tutorials did not.

To get full marks on participation, attend tutorials and always do Part A (pre-tutorial questions). Trust me, tutorial marks will SAVE you (friend got 48; had he been present to even a few more tutorials he could've probably passed). The mid-semester test should be a piece of cake - just have your notes handy and you'll be fine. For the assignments, do the first 2. Some people go off thinking they can always do the last one, but BOOM the last one's pretty hard to ace. So do the first 2, so you don't feel pressured to do the last one (still do it, though). It's pretty easy to get a perfect mark for the assignments as long as the answers do not deviate that much - they're pretty lenient in marking assignments. It's nice to compare your E-Views output with that of your friends just to be sure.

Read all lecture notes (also, listen to the lecture recordings) and study them by heart. Memorisation won't work (well, I guess it will, to an extent). Attend consults, and abuse the OLT!The bottom line is just be sure to know the definition of terms, the difference of one technique to the other, when to use a certain technique, which technique to use in various situations, and interpret E-Views results.
Overall, this subject is quite chill. Just don't leave everything to the last minute. It's up to you if you ditch lectures. But attend your tutorials and do your pre-tutes. Submit Assignments 1 and 2. Ace the mid-semester test - that 5% is still gonna be helpful. H1 is hard to achieve, but it's possible..

Yes, with screen capture

Joe Hirschberg

There is a mock final exam with solutions. Also, tutorial questions are past exam questions.

2.5 out of 5

Business Statistics (QM1 book - not relevant)

2x 1-hr Lectures; 1x 1-hr Tutorial

Semester 1 2014

H1