I was going to write a long and very detailed post about societal discussions being based on statistics which are inaccurate or inaccurately applied. I was going to use as the first example, the belief that Black American youth education is dominated by the charge of "acting white" despite the studies showing the "acting white" idea is suspect. I have a category just for "Acting White" here. If that isn't good enough for you, type the following into Google and have at it:
- "Acting white" fryer torelli
- "Acting white" phillip cook jens ludwig
I was also going to use as an example, the results of a racial profiling poll done in 2001 that "showed" Blacks support racial profiling of Arabs in airports. This is the post I made after I tried to contact Zogby International and The Gallup Organization about the poll. (Ignore the USENET flame war).
Guess who wrote to Zogby International and The Gallup Organization?
Guess who asked to know what questions were asked, the responses, the demographics of the poll respondents, the times the poll was done, the areas of the country the poll respondents were from, age of the respondents, and whether or not the respondents knew anyone who was killed or injured in the attacks?
Guess who responded?
Not Zogby, but Gallup.
Do you think I got all of the information I asked for?
No, I did not.
Do you think I got some information?
Yes, I did.
[ Note to Trudogg. The numbers you see here are different from what I sent you. That is because the person emailed me again with the _CORRECT_ numbers ]
So, Kaiju? Do you want to read what I have?
Well, here we go: :=D
Number of people polled: 1032
Number of Blacks in the poll: 71
Percentage of Blacks in the poll: 6.88%
Percentage of Blacks in the U.S. population: 13%
Oh, look what the numbers show us. Is a sample size of 71 Blacks accurate? How about a sample size of 6.88% vs 13%?
For grins, here is a link about error in sample size:
Look at the tables and explanation on sample size. And look at the cut off. A sample size of 100.
I was going to use that post as an example of the inaccuracies that result from stating poll results broken down by race when the poll itself didn't use methods to "over sample" racial groups to get a proper poll universe.
I was going to write about the affirmative action "discussion" which a large part of it is based on the graduation rates of Blacks in college. One of the things I've always stated was the statistics being used never accounted for what percentage of drop outs did so because of funding issues or those who actually left one school and transferred to another school. I wrote that because I noticed that instead of saying "dropped out," many people were saying the Black students did not graduate from the institution they entered in 4 or 5 years. I was going to point to this post:
According to a study by Nellie Mae, the largest nonprofit provider of federal and private education loan funds in this country, 69 percent of African Americans who enrolled in college but did not finish said that they left college because of high student loan debt as opposed to 43 percent of white students who cited the same reason.
Do you see how that information can drastically alter the discussion?
I was going to write about The Washington Post poll about Black Men. ( I have a category just for The Washington Post "Black Men" series). I pointed out in one of my blog entries:
- Six in 10 black men said their collective problems owe more to what they have failed to do themselves rather than "what white people have done to blacks." At the same time, half reported they have been treated unfairly by the police, and a clear majority said the economic system is stacked against them.
- More than half said they place a high value on marriage -- compared with 39 percent of black women -- and six in 10 said they strongly value having children. Yet at least 38 percent of all black fathers in the survey are not living with at least one of their young children, and a third of all never-married black men have a child. Six in 10 said that black men disrespect black women.
- Three in four said they value being successful in a career, more than either white men or black women. Yet majorities also said that black men put too little emphasis on education and too much emphasis on sports and sex.
- Eight in 10 said they are satisfied with their lives, and six in 10 reported that it is a "good time" to be a black man in the United States. But six in 10 also reported they often are the targets of racial slights or insults, two-thirds said they believe the courts are more likely to convict black men than whites, and a quarter reported they have been physically threatened or attacked because they are black.
- Black men said they strongly believe in the American Dream -- nine in 10 black men would tell their sons they can become anything they want to in life. But this vision of the future is laden with cautions and caveats: Two-thirds also would warn their sons that they will have to be better and work harder than whites for equal rewards.
I was going to point out that I never see this poll being cited in commentary about Blacks and their thoughts about things happening in some aspects of Black America. Do you think a poll like that would affect the discussion?
I was going to write all of that plus more, but DS 2.0 is throwing a fit because he's sleep deprived, I have household laundry to do, and I'm waiting status on something work related that is of importance. Besides, I've already written about that stuff and P6 gets off a decent rant related to the same idea so you should probably read it and think about it. If that's not enough, Dr. Spence writes about this as well. So, I guess my commentary isn't needed. ;)