Tuesday 10 December 2013

Question: PISA and cheating

Has anyone been able to find the part of the technical appendix in PISA 2012 which deals with the issue of whether China cheated? Or, for that matter, whether any other country cheated?

Andreas Schleicher wrote in his blog today that this is a nonsense, and that critics like Time magazine “didn't bother to read the PISA 2012 Technical Background Annex, which shows there was no cheating, whatsoever, involved.”

I can’t find that bit in the technical background annex, and the page to which Schleicher directs readers in his blog is a very general one, including parts which the system says I don’t have access to. I have sent an email requesting access, and have looked through lots of tables but can only find lots and lots of country tables and figures.

Can anyone help direct me to the relevant part of the annex?


  1. Well-meaning-amateur10 December 2013 at 22:20

    "... which shows there was no cheating, whatsoever, involved."

    Under what circumstances could such a thing possibly be shown?

    1. Hear, hear! I do get fed up with public statements that are an insult to my intelligence.

    2. In the 2009 PISA test, Azerbaijan got caught cheating by the PISA staff. Of course, we can't know what countries cheated more adeptly ...

    3. Thanks. I had forgotten that. Rindermann dropped those results from his national surveys, but based on your research into the PISA technical papers it seems probable we will have to apply a "Sailer correction" to the results. Would you like to calculate it now?

  2. Zhejiang Province (China) 2012 PISA scores are out in a local education-related Chinese newspaper :

    Math: 623 (Shanghai 613)
    Reading: 570 (Shanghai 570)
    Science: 582 (Shanghai 580)

    According to OECD PISA Programme:

    1. 80% of the scores are taken from the rural areas of Zhejiang in both 2009 and 2012

    2. All elite schools of Zhejiang didn't participate in the PISA test.

    So Zhejiang's PISA scores by and large don't include the far right side of the IQ curve.

    Zhejiang Province: population 54 million. GDP about USD 10,000 per cap. Urbanisation rate: 62%. Zhejiang is one of the homelands of Wu Chinese, a sub-race of Han Chinese.

    ~ SP ~

    1. Thanks very much for these data. I think that Steve Sailer posted, some months ago, that the Chinese had good scores even in other poor provinces, so I am not alleging cheating, mostly because so many of the small, unpublicized and apparently unimportant IQ tests all show higher scores. For once, IQ test to the rescue!

    2. Sure about that?


      "A long time ago I checked through his China references and almost none of them even contained IQ data! It was totally baffling. I have 100s of Chinese studies, many of them in Chinese. I would love to start working on the China post, but it would be a crazy amount of work, and I would probably need a good Chinese collaborator.

      You are right, though. I have not carefully examined them, but many studies from rural China report IQ scores in the 90s, 80s and lower."

      More like "As always, Richard Lynn to the rescue! (because I won't accept anything else.)"

  3. Thanks, James.

    More related data:

    Below is 2012 score scaling of Chinese national Gaokao exam.

    A full score of 100 degree in University Entrance test - Shanghai students, for example, got 67.5 to enter a university, ranking only 17th of 33 provinces within China in the important Science section.

    Note that Gaokao separates into Science and Liberal Arts 2 big categories. Both test maths, but maths tested in Science is far more difficult.

    This ranking is more or less the general ranking in Gaokao year after year. Note that Zhejiang has been one of the traditional heavyweights. Shanghai usually ranks average or slightly above average, even though the core population of both Zhejiang and Shanghai are Wu Chinese. One of the general perceptions is that kids in cities get distracted by videos games, etc and are much less motivated and hardworking than their rural and sub-urban counterparts.

    For Science(as full score=100)

    1.Guangdong Province 76.5 (a big surprise that the Cantonese did exceptional well last year. Guangdong is China's , and world's, manufacturing powerhouse, with relatively high average GDP per cap. A traditional above-average performer in Gaokao. High score are perhaps also due to new mega-cities there formed by migrations from other parts of China in the recent dacades)

    2.Zhejiang Province 76.2

    3.Sichuan Province 74.9 (Western pandaland, 100m+ strong population with relative low GDP per cap. it is one of the traditional Han Chinese enclaves stretching back to BC)

    4.Shandong Province 73.9 (homeland of Confucius, a Han Chinese stronghold)

    5.Beijing 73.3

    6.Hebei Province 71.7 ( Northern Han)
    7.Liaoning Province 71.7 ( Northern Han)

    8.Jilin Province 71.3 (Northern Han, with a sizeable ethnic Korean population)

    9.Jiangsu Province 70.4 (traditional heavyweight in Gaokao the same class as Zhejiang. Another Wu Chinese homeland)

    10.Hubei Province 70.3 (a central China Han Chinese stronghold)
    11. Heilongjiang Province 70.3 (Nothern Han)
    12.Tianjin 69.5 (Nothern Han)

    13.Chongqing 69.3 (Formerly within Sichuan province. China's, perhaps world's, biggest city with population of 34 million and ongoing...)

    14.Jiangxi Province 68.9 (full of mountainous areas. Han Chinese+ minorities, One of the originates of Communist Revolution in the earlier 20th century)

    15.Guangxi region68.0 (largely minorities + Han)

    16.Hainan Province 67.6 ( China's largest island, a weak performer traditionally)

    17.Shanghai 67.5

    18.Henan Province 67.3 (areas contain capitals of old imperial China. Han Chinese + some Jewish and Arabian descendents. The lowest on GDP per cap in Central China)

    19.Fujian Province 66.8 (a rough proxy for Taiwan and Singapore, since Fujianese are the core population base of all three. Shanghai beats Taiwan and Singapore in PISA. Surprise?)

    20.Hunan Province 66.0 ( Mao's homeland, Southern Han Chinese)

    21.Yunnan 66.0 (Minority region next to Vietnam and Laos)

    22.Shanxi Province 65.7 (Northern Han)

    23.Anhui Province 65.3 (traditionally a weak performer. Han Chinese)

    24.Gansu Province 65.2 (sparsely populated minority region)

    25.Shanxi Province 64.7

    26.Inner Mongolia 64.3
    27.Tibet Autonomous Region 62.7
    28.Ningxia (Muslim Region)60.7
    29.Guizhou Province(minority region)59.9
    30.Xinjiang (Muslim region)59.1
    31.Qinghai (sparsely populated. largely Tibetans)51.1

    For Liberal arts(as full score=100)

    1.Guangdong 79.2 (a very big surprise as well)
    2,Zhejiang 76.4
    3.Shandong 76.0
    4.Sichuan 75.6
    5.Hebei 74.8

    6.Shanghai 74.7

    7.Hunan 74.3
    8.Chongqing 74.1
    9.Hainan 74.1
    10.Liaoning 73.9
    11.Beijing 73.2
    12.Guangxi 72.1
    13.Anhui 72.0
    14.Shanxi 72.0
    15.Tianjin 71.1
    16.Jiangxi 70.9
    17.Hubei 70.8
    18.Guizhou 69.6
    19.Yunnan 69.3
    20.Henan 69.2
    21.Fujian 68.4
    22.Jiangsu 68.3
    23.Jilin 68.0
    24.Shanxi 67.6
    25.Heilongjiang 67.2
    26.Gansu 67.1 (minority region)
    27.Ningxia 64.5 (Muslim region)
    28.Tibet 64.0 (minority region)
    29.Inner Mongolia 63.2 (minority region)
    30.Xinjiang 61.3 (Muslim region)
    31.Qinghai 58.0 (minority region)

    ~ SP ~

  4. Thank you very much for this very helpful and highly informative set of findings. Is there any official source which we can give as a reference. Do you have the data in Excel format? We could then offer it up for researchers who wanted to look at it further.