It Could be Culture, part I (The NAEP Black-Mixed-White gap)

[Edit: I made a few corrections. Two variables were mislabeled in my discussion: Student-identified race is DRACE, School identified race is SDRACE. There were no errors in the data analysis.]

(There will be no part II if I am unable to come up with a coherent hereditarian explanation for the results)

(This is dissertation material people. I just want to dirty it up with my smelly racist paws before some SWPL gets to it.)

I again looked at NAEP math and reading

Method

For grades 4 and 8, I Identified mixed race students 4 ways. which after some consideration, I decided were the most reliable:

1) Student-identified as “Black,” “White,” and “non-Hispanic”
(In NAEP Explorer, cross tab BB21201 “Black”, BA21201 “White”, and DRACE “Two or more races”)

Pros: High face validity, relatively large Ns
Cons: Questionable reliability (the population identified could be an aggregate of monoracial students who cross identify)

2) Student-identified as “Black,” “White,” and school identified as “(non-Hispanic) Mixed”
(In NAEP Explorer, cross tab BB21201 “Black,” BA21201 “White”, and SDRACE “Two or more races”)

Pros: High face validity, Higher reliability (eliminates both monoracial White and Black students as a confounder)
Cons: Small Ns

3) Student-identified as “Black” and “White” and school identified as “(non-Hispanic) Black”
(In NAEP explorer, cross tab BB21201 “Black,” BA21201 “White,” and SDRACE “(non-Hispanic) Black”

Pros: High face validity, intermediate reliability (eliminates monoracial White students as a confounder)
Cons: small Ns

4) Student-identified as “Black” and “White” and school identified as “(non-Hispanic White”
(In NAEP explorer, cross tab BB21201 “Black,” BA21201 “White,” and SDRACE “(non-Hispanic) White”

Pros: High face validity, intermediate reliability (eliminates monoracial Black students as a confounder)
Cons: small Ns

I did this for all years for which there was relevant information. Since small sample sizes precluded, I was unable to do this for grades 12.

Results

Self-identified mixed race individuals (1) fell intermediate to the parental races in 4th and 8th grade for all years. However, only for grade 8, did self-identified mixed race individuals who were also school identified as being mixed (2) consistently significantly fall intermediate to the parental races. (See the two graphs on the left.)

In all cases self-identified mix race individuals who were school identified as being monoracial Black (3) or monoracial White (4) fell in between the parental races. In all cases, those who were school identified as Black (3) performed worse than those who were school identified as White (4). (See the two graphs on the right.)

Interpretation

Mixed race (1) and (2)

The most reliable method for identifying mixed race students is that which uses both student and school report. (2) above. 4th grade mixed race students identified thusly do not behave in accordance with a hereditarian model. 8th grade students do. The results are odd, since we are dealing with overlapping populations. The 2003, 2005, and 2007 4th grade sample is supposedly a representative sample of the same population from which, respectively, the 2007, 2009, and 2011 8th grade sample came from. The mixed race 4th grade increase, relative to Whites, between 2005 and 2007 should have show up as an 8th grade increase, relative to Whites, between 2009 and 2011.

Problem 1. It’s not clear how to best interpret these results. One could interpret the lack of difference between 4th grade mixed race students ( method 2) as indicating the absence of genetic difference between Whites and Blacks and then attribute the 8th grade results to the effects of culture (e.g., the Black-cultural undertow). Alternatively, one could interpret the increase in difference between mixed race students and whites as an effect of gene expression.

Mixed race (3) and (4)

Presumably, self-identified mixed race students who were school identified as being White (4) outperform self-identified mixed race students who were school identified as being Black (3) because these groups are confounded, respectively, with White and Black monoracials.

Problem 2. It’s not clear, though, why (3) outperform Self and School identified Blacks or why (4 ) underperform Self and School identified Whites. Mixed race self-identity has a clear effect on tests scores.

A cultural theory could possibly explain the above. It could be argued that identification with White and Black culture leads respectively to (3) outperforming Blacks and (4) underperforming Whites. Let’s call this the “Black cultural undertow hypothesis.” Accordingly, Black identity/culture confers a disadvantage to Whites and White identity/culture confers an advantage to Blacks.

It’s not clear how such a hypothesis can be reconciled with a cultural explanation for the gap between Whites and Mixed race 8th graders (by 2) and the lack of a gap between Whites and Mixed race (by 2) 4th graders (above). As noted, such a hypothesis contends that the presence of a gap between Whites and mixed race (by 2) students in eight grade is due to cultural effects. A corollary to this is that the absence of a gap between Whites and mixed race (by 2) students in the fourth grade is due to the absence of cultural effects. If there are no race cultural effects in grades 4, then a cultural hypothesis can not explain the grade 4 results for (3) and (4).

To investigate further, I compared (3) and (4) to two other classifications for 4th and 8th graders:

5) Student-identified as “Black” (monoracial) and school identified as “White” (monoracial)
(In NAEP explorer, cross tab DRACE and SDRACE)

6) Student-identified as “White” (monoracial) and school identified as “Black” (monoracial)
(In NAEP explorer, cross tab DRACE and SDRACE)

The following hierarchy was found for 8th graders:

1) White
2) W/B-Self, White-School
3) Black-Self, White-School
W/B self, Black-School
5) White-Self, Black School
6) Black

For 4th graders

1) White
2) W/B-Self, White-School
3) W/B-Self, Black-School
4) Black-Self, White-School
5) Black
White-Self, Black-School

Grade 8: School classified Blacks who self classified as White perform significantly better than School classified Blacks who self classified as Black. They tend to perform slightly worse than their school classified Black peers who self-classified as Mixed.

This is not fully consistent with the above cultural explanation. Self-identification as White alone does not provide an advantage relative to self-identification as Mixed.

School classified Whites who self classified as Black perform significantly worse than School classified Whites who classified as White. They tend to perform slightly worse than their school classified White peers who self-classify as Mixed.

This is consistent with the above cultural explanation.

Grade 4: School classified Blacks who self classified as White perform the same as School classified Blacks who self classified as Black. They tend to perform significantly worse than their school classified Black peers who self-classified as Mixed.

This is inconsistent with the above cultural explanation. Self-identification as White provides not advantage relative to self-identification as Mixed or Black.

School classified Whites who self classified as Black perform significantly worse than School classified Whites who classified as White. They tend to perform slightly worse than their school classified White peers who self-classified as Mixed.

This is consistent with the above cultural explanation.

Overall, while “Black cultural undertow hypothesis” can explain the pattern of performance for school identified Whites (via a wigger effect) it can not alone explain the pattern for school identified Blacks. Moreover, as discussed above, “Black cultural undertow hypothesis” is unable to account for the either the grade 4 results for (3) and (4) or the grade 4 results for (1) and (2). See above.)

How then do we account for the results?

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

8 Responses to It Could be Culture, part I (The NAEP Black-Mixed-White gap)

  1. JL says:

    How did you calculate math+reading scores? Did you do it manually or does the NAEP explorer allow you to do that?

  2. JL says:

    Presumably, self-identified mixed race students who were school identified as being White (4) outperform self-identified mixed race students who were school identified as being Black (3) because these groups are confounded, respectively, with White and Black monoracials.

    I’d guess that the difference between (3) and (4) is mostly that those mixed-race individuals who have more white (black) ancestry are more likely to be school-identified as white (black).

    I think the significant thing in those first four graphs is that individuals whom schools identify as mixed-race perform better than other mixed-race individuals in both grades 4 and 8. For example, looking at grade 4 math scores (2011), students who self-identify as black (BB21201), white (BA21201), and multiracial (DRACE10) average 240, but those who self-identify as black (BB21201) and white (BA21201) and are identified as multiracial by schools (SRACE10) average 247. However, the latter must be a subset of the former, so why do schools tend to report only the higher-performing self-identified multiracials as multiracial? (Self-identified white monoracials average 249, school-identified white monoracials 250.)

    • Chuck says:

      “I think the significant thing in those first four graphs is that individuals whom schools identify as mixed-race (2) perform better than other mixed-race individuals (1, 3, 4) in both grades 4 and 8.”

      Yes, this is significant.

      There is a hypothetical population of “true mixed race” individuals. And we are trying to identify them. Some methods are more reliable than others. It could plausibly be argued that (2) above — those who self-identify as black (BB21201) and white (BA21201) and are identified as multiracial by schools (SRACE10) [This should read SDRACE] — is a more reliable index of “true mixed raceness” than (1, 3, or 4). It could then be argued that there is little difference in grade 4 between “true mixed race” individuals and Whites. A case against an additive genetic hypothesis could then be made. So what would be an alternative? Why might (2) not be representative of the “true mixed race” population? And how might we get a better estimate of this population using the data we have? (Unfortunately we can’t decompose the school mixed race category like we can the student one.)

      The finding that (3) over-perform Blacks and (4) underperform Whites is somewhat informative. It could be that (1) includes a number of true Black and true White monoracials (i.e., monoracial Blacks and Whites who happened to check both “Black” and “White.”) The effect would be that the scores of (1) fall in between those of “Whites” and “Blacks,” thus giving the illusion of intermediate mixed race performance. (Though, in reality, if this were the case, one would not expect the scores to fall in the middle, as they do, as monoracial Whites are 4 times as numerous as monoracial Blacks. As a result, in the above scenario, we would expect 4 times as many monoracial Whites to (mis)check both the “Black” and “White” box as monoracial Blacks and, so, expect the average score of (1) to be skewed to the White end.) Such a “sampling” explanation, though, can not explain the findings with (3) and (4), since (3) presumably excludes monoracial Whites and (4) presumably excludes monoracial Blacks (via school identification in both cases).

      It could be though that (3) and (4) result from a cultural effect, which is why I compared them with (5) and (6).

      Anyways, the point is that I looked at (3-6) because of the odd results of (2) for 4th graders. I’m still trying to account for this. As usual, I would appreciate your thoughts on this.

    • Chuck says:

      JL,

      I had a few of the labels wrong. Sorry for the confusion. The above should have read:

      “1) Student-identified as “Black,” “White,” and “non-Hispanic”
      (In NAEP Explorer, cross tab BB21201 “Black”, BA21201 “White”, and DRACE “Two or more races”)

      2) Student-identified as “Black,” “White,” and school identified as “(non-Hispanic) Mixed”
      (In NAEP Explorer, cross tab BB21201 “Black,” BA21201 “White”, and SDRACE “Two or more races”)”

      Student identified race is called DRACE (or DRACE10 in 2011). School identified race is called SDRACE (or SRACE10 in 2011). The results are the same though.

    • Chuck says:

      JL,

      I looked into this some more. Here’s a screen shot of my excel file for 2011 4th grade math:
      https://lesacreduprintemps19.wordpress.com/?attachment_id=235 As I calculated sample size from s.e. and SD, the numbers are not exact. In the bottom left of the screen (3), you’ll notice that of all self-identified B/Ws (BB21201 on BA21201), only a quarter are school classified as being “Mixed.” And these have the highest scores by far.

      In the top left corner (1 and 2), you’ll notice that there are about 2000 school classified mixed kids and 1700 self classified mixed kids. On the middle right side (4), you’ll notice, which is rather interesting, that there are only 500 kids which are both self and school classified as being mixed. This implies that there’s a low correspondence between self and school “mixed” classification.

      As you can see in the middle and top right side (4), a large number of school classified “mixed” 4th graders self-classify as Black (400) or White (217). I broke down school classified mixed race by self classification on the bottom right (5).

      I’ll have to think about this a little…

  3. Chuck says:

    JL,

    I need some feedback. What about this idea:

    One problem that we’re having is that we don’t know the school identified mixed race scores. We can’t directly get into that black box, but we can indirectly. See number (5) in my linked excel screen shot below. (And ignore the typos.) We can cross tab school identified mixed race with self identified race. If we just make the assumption that school identified mixed race individuals who self identified as monoracial X are multiracial individual of X and some other parentage, we can estimate these scores based on racial intermarriage rates.

    For example, in grade 4 2011 Math, there were 402 school classified multiracial students who self-classified as Black. Assuming these individuals are Black + some other race, and knowing that 86% or so of Black mixed race, non-hispanic unions are Black and White, we can infer that 349 are mixed Black and White, non-hispanic. (The school classified Mixed race data is in exclusion of hispanics.) Doing the same for school classified multiracial students who self-classified as White, we get 62 mixed Black and White. (I’m using the Nytimes 2011, mixed marriage data, so the numbers might be somewhat off, but not horribly — ideally we would use 2001 rates, as these kids were 10 in 2011) Now we can add these to the self-identfied Black Whites who are school identified as mixed. When doing so, we get a n-weighted average of 240, relative to a White and Black average of 249 and 224 respectively.

    What do you think about the method?

  4. Pingback: Color Differences: Corrections and Further Analysis. Part 2 | Human Varieties

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s