<p>According to collegeboard.com, the mean Asian SAT writing score is 512, the mean Critical reading score is 510, and the mean math score is 578. For whites, the corresponding numbers are 519, 527, 536.</p>
<p>For your information. Now can everyone quit stereotyping?</p>
<p>Check out the rest of the college board site, though, and you can find a PDF file that shows the percentiles of each race/ethnicity scoring at different levels of the SAT I v, m. When you look at that, you'll see that, for instance, the SAT score that is 99th percentile for Asians is much higher than it is for other ethnic and racial groups.</p>
<p>The data is hard to find on the College Board site, but I believe it's in the research area of the counselors' section.</p>
<p>I hope someone will give a link because I can't find it right now.</p>
<p>Northstarmom, here is the link
<a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/CR_M_%20W_PercentileRanksGenderEthnicGroups.pdf%5B/url%5D">http://www.collegeboard.com/prod_downloads/highered/ra/sat/CR_M_%20W_PercentileRanksGenderEthnicGroups.pdf</a></p>
<p>For whites,the percentiles for an 800 in CR, M, and W are 99, 99+, and 99+, respectively. For Asians, the corresponding numbers are 99, 98, and 99.</p>
<p>Of course, it is interesting to note the relatively vast drop in percentiles from 800 to 750 in the Asian group (to 97, 92, 98), as opposed to a 1% drop in the white category (to 98, 98, 99).</p>
<p>I think what this proves is that on average, the Asian scores about the same as a white student on the SAT (the mean for both is in the low 500s). Perfect scores are obtained by both races at the same low rate (~99%ile). But Asians have a higher percentage of high-but-not-perfect scorers, people who scored between 750 and 800. So I guess you could say that the Asians are more imbalanced--they have extremely high scorers mixed with average scorers, to a greater extent than the white group.</p>
<p>NSM is correct.
Look at the 750 score this way for clarity:
750 or more was scored by 3, 8 and 2% of the Asians in V/M/W sections. Corresponding numbers were 2, 2, 1% among the whites.</p>
<p>^I acknowledged that in my previous post.
"Of course, it is interesting to note the relatively vast drop in percentiles from 800 to 750 in the Asian group (to 97, 92, 98), as opposed to a 1% drop in the white category (to 98, 98, 99)."</p>
<p>Exactly. When you look at statistics, you need to take into account the other measure of central tendency - the median. I think the median score for Asians is significantly higher than that for Whites.</p>
<p>And I think the conclusion in post #4 is sound - normally the reason for a much higher median than mean is that there are a lot of outliers relative to those with average scores.</p>
<p>Agree, I should have read other posts carefully before responding.</p>
<p>Thanks for the link to the 2006 SAT breakdown by race and ethnicity!</p>
<p>You'll see the difference most in math scores when you look at the breakdown.</p>
<p>750 -- 97th percentile Asians; 98th, whites; 99th, blacks; 99th+ Mexicans & Mexican Americans.</p>
<p>700 -- 81st Asians; 94th whites; 99 blacks, 99th Mexicans</p>
<p>650 -- 68th percentile for Asians, 85th percentile for whites, 98th percentile for blacks, 96th percentile for Mexicans.</p>
<p>550 -- 40th percentile, Asians; 54th, whites; 88th blacks, 80th Mexicans</p>
<p>The mean math SAT scores per group are:</p>
<p>Asians 578; white 536; black 429; Mexican 465.</p>
<p>Interestingly, even an 800 math score is only the 98th percentile for Asians. For whites, and everyone else, that's 99+ percentile.</p>
<p>It is well worth checking out the chart.</p>
<p>Also, when considering how many students of each ethinic/racial group score in certain ranges, here are the numbers of students taking the SAT I old version in the 2006 college-bound senior cohort:</p>
<p>Asians: 138,303
whites: 825,921
blacks: 150,643
Mexicans (the largest group of US Hispanics): 64,019</p>
<p>There also are breakdowns for Puerto Ricans, other hispanics/latinos, total males, total females, and American Indians or Alaskan natives.</p>
<p>One also should consider that "Asian" covers a lot of ground. Different groups of Asians in the US (and those in the US are the most likely to take the SATs) may (as groups) have quite different educational opportunities, cultural outlooks on education and college, heritable traits, experiences in US school and society, family expectations, etc., etc.</p>
<p>For white immigrants to the US, there has been extended time (as well as a unique political and social milieu) in which the original "groups" (English, French, Irish, German, Polish, etc. ...) have to a great extent blended together.</p>
<p>So, I think it makes sense that the varying degrees of cultural "homogenity" (very broadly speaking, and with apologies for unintended stereotyping) for Asians vs. whites in the US are a contributing factor to "non-homogenous" SAT scores for Asians relative to whites.</p>
<p>Asian could classify so many different things - Russian, Indian, Chinese, or even Middle Eastern if you wanted to. I think it is wrong to classify so many different ethnicities under one label.</p>
<p>Whites also consist of many ethnicities including North Africans and Eastern Europeans.</p>
<p>Hispanics who aren't Puerto Rican or Mexican are all categorized under "Hispanics other".</p>
<p>African Americans include Caribbean and African immigrants.</p>
<p>We could go on and on and on....</p>
<p>Asian / Pacific Islander, as a category, have the "higher" percentiles described up farther in this thread. A stereotype can easily be perpetrated, however, when what might be truer generally is applied and assumed to be true for a specific person; that's prejudice. </p>
<p>And yes, I've gone and looked at all the ethnicity numbers myself from August (honest: <a href="http://www.samjackson.org/college/2006/11/05/class-of-2006-senior-sat-percentile-scores-number-crunching/%5B/url%5D">http://www.samjackson.org/college/2006/11/05/class-of-2006-senior-sat-percentile-scores-number-crunching/</a>) and I drew the same conclusion.</p>
<p>We could know more specifically if we were given the same score-by-score numerical breakdown by ethnicity (2006) that we have for male / female, but we don't.</p>
<p>That was the other interesting thing I found: since there were more girls taking the test, there were more girls with 800 section subscores, but there were still fewer girls with 2400s.</p>