Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

ECO00019I Econometric Theory

General Comments

This was the (at least I hope it was xte) pandemic ET1 exam completed over 24 hours in open book conditions.  In order to try to maintain a rea- sonable degree of differentiation between students I resorted to three crude tactics:

1. asking for explanation of intermediate ste ps in derivations that might not receive such attention ordinarily but demonstrate a good level of knowledge;

2. investigating how a result fits into econometric theory by asking how and where it is used;

3. setting the odd ambush.

By and large the examination was done well.  The marks ranged from 9 to 83, with an average mark of 66.  Of the total number of scripts only one received marks below 50, which was undoubtedly an uploading error,whiled 15 received marks above 70, of which two very strong scripts scored above 80. As ever higher marks were awarded to answers that were largely correct, demonstrated strong understanding of the arguments used in the proofs and the implications of the results. Originality was also rewarded.

Specific Comments

The summary statistics for marks  (out of 50) for each question as as follows.

Q       mean      attempts

1        32           38

2        33           27

3        35           9

4        37           5

Question 1 was very popular, with most candidates choosing to attempt- ing this and question 2. Parts a), b) and c), which were similar to exercises seen in the problem sets, were largely done well. Credit was still given to can- didates did not use the moment generating function but worked out the mean and variance in the old fashioned way. Plenty of scripts managed to find the mle and a few produced the comment that mle was the sample mean.  Part d), was very different and relied on candidated having a good understanding of what consistency meant, for which credit was given, without rushing to use Chebyshev’s inequality.

Question 2 was also a popular choice, concentrating on asymptotic results. Most candidate did well in most parts of the question, which required a good knowledge of the material and techniques from weeks 3 to 5.  As this was an open book exam, extra weight was put on accurate answers to questions about when these techniques were used.

Questions 3 and 4 were the unfortunate wallflowers of the paper, rarely involved but delighted to be asked.   Good answers to question 3 showed understanding of the matrix expressions in parts b) and c), which could then be related to the example of the hypothesis test for the sample mean in part d). This involved quite a bit of material along the way with good marks going for clear and accurate explanations of the steps. No one, it seems, attempts the conditional normal question out of necessity and so answers to question 4 were generally very good indeed, give or take the odd numerical slip towards the end and some mistakes in counting the number of parameters in an n × n symmetric matrix.