Written   Dec.. 28, 2012


About The January 2013 Topic Changes

If you have come to this page for some predictions on the new topics for January 2013, I'm sorry but I don't have any, as I explain below. All I can predict is that, (if the exam follows the usual practice), several of the Part 1 & Part 2/3 topics will change during the month of January. And then these topics, along with the topics from before January 2013 that stay in the test, will continue to be used until the next change time, in May.


Parts 2/3

        Probably about 24.

        This is almost impossible to predict.

The reason is that the test managers now do not make the same predictable changes that they used to make. From 2001 to September 2009, the changes used to be quite predictable about 1/3 of the Part 2/3 topics were changed at each change time (January, May and September), with the result that each topic was used continuously for approximately 12 months and then retired. However, beginning at September 2009, we see that the test managers started using some topics continuously for much less than 12 months. Now they use most topics for just one testing period (4 months) or two testing periods (8 months), which gives an average of 6 months of usage for each topic. And now they seem to randomly choose which topics will be used for 4 months and which will be used for 8 months. Random => Unpredictable.

Since there are usually 48 to 50 Part 2/3 topics in use during each testing period, changing about 24 of them at each test period gives an average usage of 6 months, which agrees with the observation that some topics are used for 4 months and some for 8 months.

A small number of topics still seem to be used for about 12 months and it's possible that an even smaller number of topics are being used for about 1 1/2 years continuously before being retired.

If you look back at how many "new" topics were introduced into the test at each change time since September 2010, you will see that the number is between 22 and 25, with 23 being the the most frequent number, and 22 and 25 being the number of changes once each. In September 2012 I have 19 changes and in May 2012 I have 35 changes. It's possible that these numbers were more or less correct but it's also possible I made some major mistakes here. All this information can be seen by looking at the "Topic Usage History" files at TOPIC_USAGE_HISTORY.htm.

Probably more than half of the new topics that are introduced into the test at each test change time have actually been used before. So they are not completely new. On the other hand, a few of these might be very similar to previously used topics but have been changed in a small way, or even in a big way. It seems that the part 3 questions that go with each Part 2 topic sometimes are the same as when the topic was used previously but at other times new questions are being used.

Mostly from what people write on the internet and mostly from the renren website. For example, here is the page for what people wrote after the December 6/8 2012 tests on renren: The website also has some information but not as much as renren, and the exact dates of the tests are not clear on that website, only the months.

After recording the information for each test, I can usually see what new topics are in use by the beginning or middle of the month following the change time. This will be February for the upcoming change time. And by the middle to the latter part of that month, I can usually see what topics are no longer being used.


Part 1

The situation for the Part 1 topics and questions is similar to that for the Parts 2/3 topics and questions. There are usually 27 or 28 topics in use for Part 1 during each testing period (although I sometimes have more than 28). It seems that between 10 and 14 Part 1 topics change at each change time. But this is a little harder to see than for Parts 2/3 because quite often the same basic topic, such as "Shopping" or "Sport" is repeated but with a new set of questions.

We can assume that, as with the Parts 2/3 topics, there is a random choice of what topics stay in use for 4 months and what topics stay in use for 8 months or 12 months. This makes predicting what particular topics will be retired almost impossible.