Tuesday 3 December 2013

Warning on PISA: Intelligence may be involved

 

Late in the day, I have developed an early warning system to detect when an intelligence related story is entering the public domain: politicians start accusing each other of having fouled up the public education system. The exchange of insults indicates that a set of exam results is coming out, and politicians are playing the blame game. This has now become relatively sophisticated, as these things go, because it includes reference to lagged indicators, and estimates of representativeness and error terms. For example, if current UK 15 year olds are scholastically weak by international standards, is that the fault of the current administration (responsible for the last two years) or the previous administration (responsible for the last 13 years)? Or can we trace the fundamental errors back to the 1870 Education Act? Is our political and journalistic class becoming even more sophisticated? There’s a thesis in that for somebody.

Anyway, last week there was a ritual exchange of insults about the United Kingdom education system which, it was intimated, would shortly be shown by an international study to be “stagnating”. Today the story came out, as trailed, with the PISA results showing that the UK is 26th in the world. League tables seem to have more impact than anything else in British culture, a reference to sport, apparently. Makes it hard to see that a whole bunch of countries are “stagnating” together, within the usual error terms. This is the snapshot most people will see on the news. A more detailed look might follow later.

Here is the BBC graphic:

image

 

Attention centres on Shanghai, Singapore and Hong Kong who are way ahead, and South Korea, Vietnam and Japan which also do well, but the discussion is all about the methods of education. At no stage in the discussions is it mentioned that, from the point of view of intelligence, these results are entirely predictable.

For example, here are a list of relevant international intelligence test results, based on the comparison group of white Europeans having a “Greenwich Mean” IQ of 100, sd 15. The columns refer to country, age range of people tested, numbers tested, test used, IQ result corrected for Flynn effect as estimated by Richard Lynn (one of the discoverers of the effect), and the name of the author and date of publication. (I am working on getting these into one easily accessible list).

China

6/16

660

WISC-R

107

Dan et al., 1990

China

5/15

5,108

SPM

101

Lynn, 1991

China

14/15

297

Various

103

Li et al., 1996

China

6/12

269

SPM

104

Geary et al., 1997

China

4

60

Arithmetic

109

Ginsberg et al., 1997

China

6/13

463

DAM

103

Cox et al., 1998

China

6/8

160

SPM

107

Goa et al., 1998

China

17

218

SPM

103

Geary et al., 1999

China

19

218

SPM

113

Geary et al., 1999

China

6/8

300

BTBC-R

107

Zhou & Boehm, 2001

China: median

     

105.5

 

Hong Kong

9/11

1,007

CCT

105

Godman, 1964

Hong Kong

16

5,209

AH4

106

Vernon, 1982

Hong Kong

10

1,000

SPM

109

Chan & Vernon, 1988

Hong Kong

6/13

13,822

SPM

103

Lynn, Pagliari & Chan, 1988

Hong Kong

6/15

4,500

SPM

110

Lynn, Pagliari & Chan, 1988

Hong Kong

10

197

SPM

108

Lynn, Pagliari & Chan, 1988

Hong Kong

9

376

CCF

104

Lynn, Hampson & Lee, 1988

Hong Kong

9

479

SPM

122

Chan et al., 1991

Hong Kong

15

341

APM

120

Lynn & Chan, 2003

Hong Kong: median

     

106

 

Singapore

13

337

SPM

103

Lynn, 1977b

Singapore

15

459

APM

114

Lim, 1994

Singapore: mean

     

108.5

 

Korea: South

2/12

440

KABC

113

Moon, 1988

Korea: South

9

107

SPM

109

Lynn & Song, 1994

Korea: South

4

56

Arith

103

Ginsburg et al., 1997

Korea: South

6-16

2,231

WISC-3

100

Georgas et al., 2003

Korea: South: median

     

106

 

Japan

5/15

1,070

WISC

102

Lynn, 1977a

Japan

35

316

WAIS

102

Lynn, 1977a

Japan

5/10

760

MFFT

107

Salkind et al., 1978

Japan

10

212

Kyoto

106

Lynn & Dziobon, 1980

Japan

8/11

97

WRAT

108

Tarnopol & Tarnopol, 1980

Japan

9

223

CEFT

112

Bagley et al., 1983

Japan

4/9

347

CMMS

107

Misawa et al., 1984

Japan

6/11

480

Various

105

Stevenson et al., 1985

Japan

6/16

1,100

WISC-R

103

Lynn & Hampson, 1986

Japan

4/6

600

WPPSI

105

Lynn & Hampson, 1987

Japan

14

2,100

Kyoto

104

Lynn et al., 1987a

Japan

13/15

178

DAT

104

Lynn et al., 1987b

Japan

2/8

548

McCarthy

103

Ishikuma et al., 1988

Japan

6/12

142

KABC

101

Kaufman et al., 1989

Japan

16

175

AMM

113

Mann et al., 1990

Japan

9

444

SPM

110

Shigehisa & Lynn, 1991

Japan

5/7

454

CCAT

109

Takeuchi & Scott, 1992

Japan

6/12

451

MAT

105

Tamaoka et al., 1993

Japan

14/15

239

Various

103

Li et al., 1996

Japan

6/17

93

Gen Info

100

Chen et al., 1996

Japan

19

72

GMRT

102

Flaherty, 1997

Japan

7/11

60

DAM

102

Cox et al., 2001

Japan

17

1,119

Gen Info

105

Evans et al. 2002

Japan: median

     

105

 

 

So, even if we just consider an IQ test to be a “school far” test (testing things which anyone should know who has been brought up in a particular society), and a scholastic test to be a “school near” test (testing things that are taught at most schools in a particular society) then it is clear that we can test ability early in the life cycle, and use that to predict  scholastic outcomes at 15. If you accept that people vary in intelligence, and that some societies are composed of people who are brighter than people in other societies, then intelligence is part of the mix to be considered when doing international comparisons of scholastic attainments.

And now a few things about PISA: there is no real enforcement of the requirement that the schools tested should be representative of the countries concerned; some countries have strength in depth (appears to be true of China, where even poor provinces do well) while others have much more variability (India), and  it is possible to game the test by teaching to it, rather than to a general curriculum. National exams probably have better coverage of the school population and of the curriculum, but they too can be gamed in various ways, by training children in particular test formats. As a rule of thumb, don’t let the providers of education have the final say about the outcomes of education. Get someone else to judge: universities, employers, patent offices, research labs.

However, whatever the quibbles, the results show that the Far East is about three school years ahead of the West. Sure, the West can ask for a re-count with recent immigrants excluded (PISA does actually provide some of those figures, and even second generation immigrants are usually not at the same standard of the locals), but if you combine a population of IQ 106 with an open economy you get high scholastic attainments, higher than societies with IQs of around 100 and moderately open economies.

Within our own society we probably have 7 tribes of intelligence. The same seems to be true of the world, though we are talking mean differences, not different bands of ability on a single national bell curve. It appears to be the case that the Top Tribe of Global Intelligence reside in China, Japan, Korea and Singapore and thereabouts. (For the moment, we leave to one side small minority groups like European Jews, who are as far ahead of the Chines as the Chinese are ahead of the Europeans). If we take 2011 estimates of a world population of 7.1 billion persons, then with China, Japan, South Korea we have roughly 1.6 billion persons. If they are three school years ahead of the West (three years of full time instruction that will not be easily made up as people take on full time work, so are likely to be permanent) then we (Westerners) have a competitive challenge on our hands.

If we assume that Europe has 740 million inhabitants and the USA 314 then that gives us an estimated Western European population of 1.05 billion. Assuming Easterners have a real IQ of 106 they are 0.65 of a standard deviation above the West. By my crude calculations that implies that only 35% of Westerners are above the Chinese mean, whilst 50% of the Chinese are above their mean, of course.

So, 369 million bright Westerners will be slugging it out against 800 million equally bright Easterners (all of them above GMIQ 106). Western wages are currently about 6 times as high as Chinese wages.

Have I ever explained to you that IQ is only a single number, that there is much more to people than IQ, and that Westerners are very creative, and have multiple intelligences, so many intelligences that, so far, they have never been measured?

21 comments:

  1. Back in December 2010, Lord Voldemort broke down the USA PISA scores by race and posted a PISA score bar chart with U.S. whites, blacks, Latinos, and Asians shown as separate bars. As I recall, U.S. whites and Asians came in about fourth or fifth from the top. Voldemort's blog post about this can be found at the following address, but unfortunately the chart itself is now behind a paywall, which you will hit if you click through:

    http://isteve.blogspot.com/2010/12/pisa-school-test-scores-by-ethnicity.html

    ReplyDelete
  2. On the subject of not letting the providers of education have the final say about the outcomes of education:

    "Two Philadelphia School Principals Lose Credentials In Wake of Test Cheating Scandal

    ... Two Philadelphia principals have surrendered their administrative credentials amid the state investigation into test cheating. These are the first heads to roll in an investigation of 53 Philadelphia schools that has been going on for more than a year.

    Barbara McCreery, the former principal at Communications Tech High School, and Lola Marie O’Rourke, who used to be the principal at Locke Elementary School, have both given up their credentials “in lieu of disciplinary action” from the Pennsylvania Department of Education....

    The move follows an investigation into allegations that students were coached and answers were changed in state PSSA (Pennsylvania System of School Assessment) test booklets....."

    http://philadelphia.cbslocal.com/2013/04/03/two-philadelphia-school-principals-lose-credentials-in-wake-of-test-cheating-scandal/

    ReplyDelete
  3. Last time I glanced at Lynn's Chinese IQ data sources, all studies I was able to access were convenience samples of either Shanghai and Beijing populations.

    B.B.

    ReplyDelete
  4. PISA apparently went further afield to poorer provinces, still finding good results. However, it would be usual for city samples to have higher scores, and usual for researchers to go where it is most convenient, so it is still an open question.

    ReplyDelete
    Replies
    1. Shanghai is actually a consistant average player in China's all-important domestic Gaokao exams on Maths, Sciences and Literature. There're about 10 poorer and much poorer Chinese provinces outscore Shanghai (some by a big margin) in Gaokao every year.

      See some detail of Shanghai on my posts here:

      http://staffanspersonalityblog.wordpress.com/2013/11/28/the-sour-grapes-of-pisa/comment-page-1/

      ~ by SP ~

      Delete
    2. Thanks for directing me to your blog. I think we are both on the same page about the major factors in these results. Are we overlapping too much? I may have to turn to other things, currently on the back burner. best wishes
      James

      Delete
    3. Oh James, it's not my blog, but a Swedish guy called Staffen's. I am only one of his commenters.

      ~ SP ~

      Delete
  5. A very good piece as usual, James!

    I disagree on 3 small things:


    1. “while others have much more variability (India)…”

    It appears to be an urban myth, largely propelled by “high“ caste Indians themselves and their Anglo-American counterparts. Some recent studies show that unlike what previously thought that there’s “north-south divide”, Indians are actually quite homogenous genetically at the sub-continental level. It fits into intuition as well, have you ever confused an Indian with a Korean?

    2. “and it is possible to game the test by teaching to it, rather than to a general curriculum.”

    Both Steve Hsu and Steve Sailer had entries showing that though it is possible to game a test such as IQ test, or PISA, or SAT, the results are statistically insignificant.


    3. “…small minority groups like European Jews, who are as far ahead of the Chinese as the Chinese are ahead of the Europeans…”

    It is a hypothesis, not a fact yet. Perhaps also an urban myth. Because

    A. It is just me or...I have never seen an IQ study, or a standard test such as SAT, PISA, where the statistically valid results of European(Ashkenazi) Jews have been directly compared to any of East Asian population groups being the Chinese, the Japanese or the Koreans, apple to apple, let alone the gap you mentioned.

    B. None of IQ –related specialists (e.g. Rushton, Lynn, Jansen, Murray, Cochran…), as far as I know, has presented any evidence that the average brain size of European(Ashkenazi) Jews is equal or larger than that of the East Asians.

    Most, if not all, “evidences” presented to "prove" European Jews' avg IQ and the "gap" betwwen them and the East Asians, are usually how many European(Ashkenazi) Jews won Nobel Prizes, or go to elite US universities, or are wall street billionaire financiers, or just who is whom name count on celebrities , percentage-wise vís-a-vís the East Asians…

    these “evidences” could be very subjective and could be indeed gamed sometimes (e.g. even Obama won Nobel Prizes, went into Harvard & Columbia, and likely to be a billionaire or multi millionaire in the future as well.)

    If you have any evidence of above missing links, I appreciate it if you could give a source. Thank you, James.


    ~ SP ~

    ReplyDelete
  6. ISBN 978-1-59368-036-7 "The Chosen People: A study of Jewish Intelligence and Achievement" Richard Lynn, Washington Summit 2011 is the best general source of data. Kindle edition here http://www.amazon.co.uk/s/ref=nb_sb_noss?url=search-alias%3Daps&field-keywords=the+chosen+people+Richard+Lynn&rh=i%3Aaps%2Ck%3Athe+chosen+people+Richard+Lynn
    For another approach, see La Griffe du Lion http://www.lagriffedulion.f2s.com/ashkenaz.htm This is very good on defining what would constitute a real world test of intellect, and going through the statistics of exceptional achievement in competitions. By the way, can you change your anonymous to anonymous with a number attached? It allows me to distinguish you from anonymous. Yours, Thompson2013

    ReplyDelete
    Replies
    1. Thank you for the reply, James. I only can reply as anon since my another account somehow doesn’t work at the moment, sorry. For your convenience I’ll sign my name as “SP” at the end of my posts. My background: I have a Phd in Finance and am in banking industry. I generally go through the article of La Griffe du Lion you mentioned. To be brutally direct I see it as a bordering piece of intellectual retardation ( excuse my wording) if the author is intellectually honest.

      Some simple rebuttal on the central pieces of the argument of La Griffe:


      1. “ Ashkenazim have earned 27% of the Nobel Prizes awarded to Americans, 25% of ACM Turing Awards, and 26% of the Fields Medals. They account for more than half the world chess champions. Ashkenazic Jews, 2 percent of the US population, make up 30% of elite-college faculty, 30% of Supreme Court law clerks, and 27% of Ivy Leaguers.”

      Monks in the old Tibet had such more impressive honours percentage-wise on all fields against 99% of the serfs there that could make the Tibetan monks have average IQ of 300, following La Griffe’s logic. Yes, it sounds a bit extreme but logically as sound as this piece.

      More to the matter at hand, the US has about 300m European White with avg IQ of about 100. Even wildly assume that Ashkenazi Jews have avg IQ of 120, the sheer number of White European Americans with IQ > 145 will be at an order of magnitude bigger than that of Ashkenzais in the US and worldwide, a simple number’s game. The above very impressive percentages of Ashkenazis prove nothing more than A/ a stunningly high percentage of very high IQ White European Americans have been NOT at positions to compete with Ashkenazis in all these varies prizes for variety of reasons. B/ (likely) a stunningly high percentage of Ashkenazis have been in the competitions for these prizes for varieties of reasons. And C/ Ashkenazis are very likely to have a high average IQ vís-a-vís the Whites (strictly speaking, we only know C for sure through a variety of other data other than these prizes, many of which are politically driven with varies degrees of “Affirmative Action” in play) . As for how high, we don’t know at this point. As for the comparison to East Asians, we don’t know at this point either, according only to these percentages.



      2. La Griffe continues with examples from Putnam to USSR Chess Championship tournament to derive that magic number of 0.73 (sd). Basic maths aside, there are countless assumptions involved on who are likely to participate these competitions, their likely avg age, gender, (Chess) , recommendation received in pre-selection (Putnam), etc. etc Quote “…was computed assuming 13 million Ashkenazim and 1 billion non-Jewish whites. Because women are vanishingly inconspicuous at this level of achievement, and the medals are awarded to mathematicians no older than 40 years, the base populations had to be reduced. I applied a factor of 0.2 to these populations to reduce them to more appropriate values.” Even though La Griffe emphasises that some changes in the assumptions will not affect the end result, I am deeply sceptical from statistical point of view, as we all know that a slight change in data or assumptions on how to select data can almost guarantee different results in most case. So called “garbage (data) in, garbage out”. One doesn’t needs to be an expert on Griffe or Gausian distribution to know that.

      Part I

      ~SP ~

      Delete
    2. (Part II)

      In fact, La Griffe derived his “112 average IQ” for Ashkenazis using USSR Chess Championship. In essence La Griffe is using an indirect method via varies indirect (quite subjective: assuming all high IQers that is good at, and like, math… must be at Putnam… and know how to play chess …and were in USSR Chess Champs… ) data sets with many arguable assumptions to trial& error out the figure. With trial & error, and with flexible use of data and assumptions, I personally can churn out infinite of “SP Functions” to illustrate anything you want. But sorry, that is absurd, at least not conclusive. Regardless of the correctness and the accuracy of his function, to start off La Griffe is already very troubled in his methodology(as illustrated by above two points) that can not answer, and goes directly against 2 far more intuitively convincing linear logic on the level of Ashkenazi Jew’s IQ:

      i. If 14 million Jews can be sub-divided into 3 sub-races with Ashkennazi Jews being one of them, there is no reason whatsoever to not to divide 70 million Germans into 15 sub-races, or divide 80+million Koreans into 18 sub-races, or divide 120 million Japanese into 24 sub-races, or divide 300m White European Americans into 60+ sub-races, or divide 1.3 billion Chinese into 250 sub-races…for the sake of a remotely decent apple-to-apple comparison, reasonally assuming that the Jews, as a race or not, is NOT categorily at the level of major races such as Mongoloid, of which the Chinese+Koreans+Japanese consisting almost 95%+ . Obviously, no IQ expert of the mankind thinks that this is worth of mentioning. So no such an apple-to-apple comparison has ever been done.

      ii. Perhaps mostly importantly, the average brain size. No data on average brain size of Jews and Ashkenazi Jews are available. Does P. Rushton see the data irrelevant? Why? I see it critical. Brain size is the central piece to link varies disciplines and make them coherent together. If Ashkenazi Jews have larger brain size than the East Asians, then the entire elegant architecture of Rushton breaks down into pieces. If Ashkenazi Jews have smaller brain size than the East Asians, Rushton stands and Lynn can go take a second and third look at his related Ashkenazi data source.

      Bottomline: Rushton and Lynn, one of them is very wrong on this!

      To conclude, we don’t need La Griffe’s “2nd or 3rd derivative”- style to function-guess a relative IQ. With only 1 assumption wrong, which is highly likely, the whole thing becomes rubbish. Far more direct, simple and accurate approaches are available right in front of us: The largest concentration of Ashkenazi Jews is in Israel.

      -- Why not show some simple large-sample size IQ test results of Ashkenazi Jews in Israel ( which is far more convincing than listing some percentages of Putnam or Nobel prizes etc – China and Korea have zero Nobel Prize Science so far. The average IQ of the Chinese and Koreans must be 60 then? A “Nobel Prize” logic that is bordering retardation, is it not?)

      -- Why not just simply show the scores of Israeli Ashkenazis on TIMASS, PISA 2009 and PISA 2012 – all proven highly correlated to IQ – so that we can have a simple and direct comparative proof? Either OCED or Israel, or both, have the data, as Ashkenazi constitute about half of Israel’s population. They must have the data. Even The US have such data , though I don’t expect PC-obsessed US to disclose it.

      So James, that’s why methink that the long-accepted mainstream “conclusion” that Ashkenzai Jew’s avg IQ is somehow higher than East Asian’s, and with a large gap in between as you wrote, is far from settled yet according to the dismal level of data and logic presented by the IQ-related experts thus far.

      ~SP~

      Delete
    3. Bottomline: Rushton and Lynn, one of them is very wrong on this!

      (Personally, I think Rushton 's elegant 3-way theory is far more convincing than Lynn's related data source.)

      ~SP~

      Delete
    4. The hhbd movement continues perpetuating the idea of ​​human universality ( leftoid behavior ) , that all people have similar patterns of behavior and intelligence . It has been proven that it can not be , since it is within families can now find wide range of behavior, level of intelligence and cognitive style then it is not possible that large discrepancies are not found in separate populations that remained thousands of years from each other.
      It seems that the correlation between brain size and intelligence is not so great as to be in perfect symbiosis . We know for example that women tend to have smaller brains than men and rightfully so for example are better at multitasking . When it reaches a certain level of intelligence (above 100) , I believe , that the correlation between brain size and intelligence , or at least with intelligence and normative functional neuro- genetic , begins to lose its connection.
      Precisely because they have smaller brains that women (compared to men) seem to have greater connectivity between the two sides of the brain, which results beyond the aforementioned advantage, also in increased capacity for communication.
      Autistics for example, usually have large brains, however, their large heads tend to be a hindrance in relation to their social skills. Of course I am aware of the natural neuro-morphology of this population as one of the components responsible for his ill adapted to modern human societies.
      Sometimes I have the impression that Asians seem more like well-adapted aspies, with great intelligence, introverted personalities, many neotenic traits and a certain self of unconsciousness (ego-social-conscience) that can result in impulsive perseverance and ambition, traits that are critically important to Nobel winners. Asians were like a phenotype that resembles certain features found in the aspie population.
      Of course there is the idea that many of nobel prize winners, were autistic, but as I suggested, Asians resemble in some aspects, positive and negative, with aspies, however, the high-functioning aspies of white race that certainly should show different traits compared to their peers in Asia.
      With certainty that there are many environmental-subjective explanations to explain the Ashkenazi success in Western societies, however, believe that there are other cognitive factors besides IQ scores that can help us understand the rapid rise of this group such as cognitive style . Ashkenazim are predominantly neuro-different from most of us and this difference comes as a surprise advantage, resulting in the rapid social escalation of the group in the U.S., example.
      I never understand what the actual amounts of verbal IQ and performance IQ (nonverbal?), But I've read that verbal IQ is less g loaded and therefore more influenced by environmental factors such as literacy and acculturation. What I notice, the average IQ of races and nationalities are computed by their IQ scores in the average performance while Ashkenazim are punctuated especially through verbal IQ. As a result we have 105 iq for Asians, 110 for non-verbal IQ and 98 for verbal IQ,
      100 for whites, iq scores and verbal-nonverbal IQ are very similar,
      85 for african-americans, verbal IQ, slightly larger than the iq-nonverbal,
      Finally we have a verbal IQ of 107-115 for Ashkenazi and 98 (?) iq-nonverbal.
      If the IQ performance is basically the average of all the tests then the average Ashkenazi intelligence is not larger than the Asian average and is only slightly higher than the average Caucasian Americans.
      Ashkenazi Jews have, besides environmental reasons (and partially genetic like nepotism), also unusual features in his model of intelligence that can result in unusual mental skills.

      Gottlieb

      Delete
    5. " It seems that the correlation between brain size and intelligence is not so great as to be in perfect symbiosis ." (Gottlieb)

      Human brain is arguablely the single most expansive item of a human body. Evolution ensures to a pinpoint accuracy (i.e. endless examples everywhere you look in nature) that there's a mission-critical reason for 1 cubic milimetre more.or 1 less. Mother Nature doesn't waste any resources to play dice.

      ~SP~

      Delete
    6. Look, from what I've read, the correlation between brain size and say, scores on cognitive tests was 0.3. Not bad, but not enough to set in a robust relationship. Mother nature does not always hit, especially with humans anthropomorphizing their habitat, since a long time that natural selection has lost its space for sexual selection. Also in nature is important to emphasize the balance of the biological entity and as a result, a large brain too much will result in charges. For all biological traits there is a spectrum (style roller coaster) where they will be neutral (start the roller coaster), advantageous (apex of the roller coaster) and disadvantageous (drop roller coaster).
      In collective groups or populations may be true even if the mother nature always make the right account of subtraction and addition, but in individual terms, a lot of mistakes can and usually happens. Natural selection, phenomenon which is no longer an imperative in modern societies functioned as a downsizing of deleterious mutations and preservation of advantageous mutations. With sexual selection, assumes that by coincidence (and not necessarily purposeful way) the individual shall select markers of good health, which often results in higher intelligence, mainly because the best physical appearance, especially in women, is based on neoteny and the same relates to the increase of the brain.


      Gottlieb

      Delete
    7. According to the Israeli press children who were tested in hebrew had a math results of 499 while children who were tested in Arabic scored only 378 . That means that the gap between Arabs and Jews is 1.2 SD (120 point). Arab students are 25% of all Israeli students.
      Consider that Ashkenazim/Mizrahim ratio among Hebrew testers is 50/50 , and that Ashkenazim typically score more 0.75 SD (75 point) then Mizrahim (e.g on the Israeli SAT) , then Ashkenazim score in math is about : 499 + 75 * 0.5 ~ 531.

      Delete
    8. And the other scores in science and reading?


      Gottlieb

      Delete
  7. Sorry, as to gaming tests, the rude word is cheating. See Education Realist on this. I don't want to believe him, but sometimes psychologists are far too trusting.

    ReplyDelete
  8. The question of sub-division can now be looked at in terms of genetic distance, so we have a metric which helps us in classificatory debates.

    ReplyDelete