Lesson: Political Math (Racial abilities, Bush, misleading statistics)
Summary: President Bush claimed that the statistics show decrease in the racial achievement gap as a result of his policies.
Applying real statistics to the data shows no such decrease. The achievements of blacks and whites could have increased or the passing score could have became easier to achieve - but the black-white gap measured correctly is about exactly the same.
The Harvard M.B.A. graduate Bush either does not know basic statistics or is intentionally misleading the public.
In an address to the Urban League in July 2003 (http://www.whitehouse.gov/news/releases/2003/07/20030728-3.html), President Bush observed:
Nearly half a century after Brown versus Board of Education, there's still an achievement gap in America. On the most recent National Assessment of Educational Progress, on the reading test, 41 percent of white 4th graders were proficient and better readers, but only 12 percent of African-Americans met that standard. That means we've got a problem.
. . .
I know setting high standards works. I know measuring and using the measurement system as a way to diagnose problems so you can focus on the problems works. In my state (Texas), 73 percent of the white students passed the math test in 1994, while only 38 percent of African-American students passed it. So we made that the point of reference. We had people focused on the results for the first time -- not process, but results. And because teachers rose to the challenge, because the problem became clear, that gap has now closed to 10 points. Because every child can learn, you've just got to focus the attention and the resources when necessary. Accountability tells you what's going right and it tells you what's going wrong and it shows you where the emphasis needs to be.
We're having the same results in North Carolina. In states that measure, you'll find that the achievement gap is closing dramatically.
NC percent passing both reading and math at Level III. Grades 3-8
% white % black point gap
1992-93 63.4 30.1 33.3
1993-94 66.2 32 34.2
1994-95 68.9 34.5 34.4
1995-96 70.9 36.2 34.7
1996-97 72.7 38.4 34.3
1997-98 76.7 44.7 32
1998-99 79.2 48.5 30.7
99-2000 80.2 49.6 30.6
2000-01 82 52 30
2001-02 84.4 56.6 27.8
2002-03 88.8 66.9 21.9
Source: NC Department of Public Instruction
TX percent passing TAAS math exams. Grades 3-8 and 10
% white % black point gap
1994 73.3 38.1 35.2
1995 79.2 43.8 35.4
1996 85.0 55.0 30.0
1997 89.5 64.1 25.4
1998 91.9 70.5 21.4
1999 92.5 72.8 19.7
2000 93.6 77.0 16.6
2001 95.1 81.9 13.2
2002 96.5 86.5 10.0
Source: Texas Education Agency
Let's concentrate on two years from each sample - earliest and latest dates.
North Carolina shows a 34% decrease in "gap" - from 33.3 to 21.9 points.
Texas shows 72% decrease in "gap" - from 35.2 to 10.0 points.
What does that "point gap" mean? If you wanted to measure the performance gap in jumping between a team of toddlers and a NBA team, you could measure how many of them would jump over a 1 foot-high plank. If 100% of NBA players and 90% of toddlers jump over, you have a gap of 10 points. If you lower the plank to 1/2 foot or train the toddlers to jump higher, 95% of toddlers will be able to jump over the plank and the gap will decrease to 5 points. It will be halved!
Does it seem like a good way to measure performance gap? Unless you are president Bush, you would probably be interested in the difference (gap) between the average height of a jump for each team.
So let's try to estimate the difference in average test score from the data available, converting percentiles into standard deviations (SD):
NC percent passing both reading and math at Level III. Grades 3-8
% white % black pt gap SD wht SD blk SD gap Adj SD gap
1992-93 63.4 30.1 33.3 -0.36 0.5 0.86 0.90
2002-03 88.8 66.9 21.9 -1.22 -0.44 0.78 0.92
TX percent passing TAAS math exams. Grades 3-8 and 10
% white % black pt gap SD wht SD blk SD gap Adj SD gap
1994 73.3 38.1 35.2 -0.62 0.3 0.92 0.99
2002 96.5 86.5 10.0 -1.81 -1.1 0.71 0.91
I assume that a standard deviation (variability) of black population is 0.9 of the standard deviation of the white population - from other sources but the difference between 1 and 0.9 would not change results much. As expressed in black SDs, the difference between white and black average scores was for NC:
1992-93 0.90 black SDs - a score of an average white was better than that of 81.6% blacks
2002-03 0.92 black SDs - a score of an average white was better than that of 82% blacks
For Texas:
1994 0.99 black SDs - a score of an average white was better than that of 84% blacks
2002 0.91 black SDs - a score of an average white was better than that of 82% blacks
That means that in Texas the gap decreased by only 2% (not 72%) and increased in NC by 0.5%.Of course margn of error in my rough calculations is about that much, so we can consider that there was no change. When you have two populations with distinct normal distributions, the shift in a treshhold would cause non-linear effect on percentile rations. It's just a statistical trick - not necessarily an indication of any change.
One thing is clear - the gap was not reduced. If you increased the difficulty of the test so that 73.3% white texas students passed it, 38% of black texas students will pass it just like they did in 1994. If the test was made hard enough for 63.4% white NC students to pass, 30% black NC students would pass it just like in 93.
I would not speculate whether there was real performance improvement of both whites and blacks on the test or the test was "dumbed down" - I have no data for that though I have my suspicions. I would not speculate whether nature or nurture, genetics or racism caused the persistent gap in average abilities between races. The only claim I make is that the gap exists and is exactly the same as it was.
Politicians lie. You learn economics - you discover that they lie about economics. You learn history - you discover that they lie about history. You learn natural sciences - you discover that they lie about environment, biology, technology, etc. You learn math - you discover that they lie about math.
When their lips are moving, they are lying. If you are not intelligent and educated enough to understand what they are talking about, you are lost.
miko