Some debate in Chevy Chase (DC) on significance of latest NAEP scores …

On a local DC list-serve for the region where I last taught (and also went to Junior High School), I posted this:

==========================================================

Those of us with kids in Chevy Chase – DC, either now, in the future, or in the past, have seen many changes in education here in DC, especially since 2007, when the elected board of education was stripped of all powers under PERAA and Chancellor Rhee was appointed by Mayor Fenty.
[I personally went to Junior High School here at Deal back in the early 1960s, taught math in DCPS from 1978 to 2009, including 15 years at Deal (much to my surprise) and my own kids went K-12 in DCPS, graduating from Walls and Banneker, respectively]
Was mayoral control of schools in DC a success? Is the hype we have all heard about rising test scores for real?
We now have statistics from  NAEP* for about two decades, and we can compare scores for various subgroups before and after that 2007 milestone.
Did Black students make faster improvements after PERAA than beforehand? Nope. To contrary: their scores were inching up faster *before* 2007 than they have been doing since that time.
Did Hispanic students make faster improvements under the reformers? Nope, again.
How about students whose parent(s) didn’t graduate high school, and/or those who finished grade 12 but either never went to college or else didn’t earn a degree – surely they did better after Rhee, Henderson et al. took over? Again, no.
Then what group of students in Washington DC *did* make more progress on the NAEP after the Reformers took over?
You guessed it, I bet:
White students, and students with parents who earned a college degree.
Amazing.
Guy Brandenburg
*National Assessment of Educational Progress
======================================================================
Another person contested my assessment and wrote the following:
=======================================================================
The NAEP is cross-sectional data, i.e. it does nothing to adjust for changes in composition of test-takers over time, which is why Steve Glazerman refers to comparisons of NAEP scores over time as “misNAEPery” [https://ggwash.org/view/ 31061/bad-advocacy-research- abounds-on-school-reform] and I have referred to the same thing as “jackaNAEPery” [https://www.urban.org/urban- wire/how-good-are-dcs-schools] .
There has been a dramatic, even shocking, compositional change since 2000 in births across the city, entering cohorts of students, and exit rates from DC schools and the city.
Most noticeably in NW, better educated parents are substantially more likely to have kids in DC, enroll them in DC public schools, and stay past 3rd grade.
Any analysis of test score change needs to grapple with that compositional change.
But more importantly, the compositional change itself is a policy outcome of note, which the DC Council and Mayor have an interest in promoting.
The only evidence one should accept must *at minimum* use longitudinal data on students to compute *learning* as opposed to static achievement, e.g. this analysis of 2008 school closures:
A lot of other things happened 1996-2008 of course, including a rapid expansion of charters, a shrinking proportion of DC residents attending private schools, etc.In 2008 alone, a lot of Catholic schools closed, and some converted to public charter schools.
During this time, we also had a voucher program that produced some gains early on, and then began to lower test scores relative to public options:
All of this is not to say DCPS and charter schools shouldn’t serve less advantaged students better than they do–obviously they should! But the evidence is nuanced, and DC has made huge gains across the board since the 1990’s that make attributing any changes to policy rather than shifting population composition problematic at best.
Interestingly, the NAEP data explorer [https://www. nationsreportcard.gov/ndecore/ xplore/nde]does not report scores for white 8th graders in 1990, 1992, and 1996, presumably because too few were tested. I.e. the means by race show a lot of  “‡ Reporting standards not met.
[I personally attended DCPS (Hyde, Hardy, and School Without Walls) 1976-1989, have 2 children currently in Deal and SWW.]
Austin Nichols
========================================================================
I wrote a response to Nichols, but it hasn’t been posted yet, and might never be:
========================================================================
My previous reply got lost somewhere in cyberspace.
If looking at long-term trends in the NAEP and TUDA is ‘misnaepery’ or ‘jacknaepery’, as Mr Austin would have us believe, then the entire NAEP bureaucracy has been doing just that. (In fact, an entire branch of the National Center for Education Statistics is devoted to, yes, Long Term Trends: https://nces.ed.gov/nationsreportcard/ltt/ )
It’s a laughable idea that we could just use the tests chosen by DCPS and later by OSSE and administered every year, to tell how good DC public or charter schools are, over time. First of all, the tests administered here have changed dramatically. Back in the 1990s it was the CTBS. Then it was the SAT-9, developed by a different company. Then it was the DC-CAS, again, a different vendor. Now we have the PARCC produced by yet another vendor. We also know that in the past there has been major fraud with these tests, committed by adults, in order to gain bonuses and keep their jobs. We also have no way of comparing DC with any other city or state using those tests, since only a handful of states even use the PARCC and for all I know, their cut scores and questions might be different from what we use here in DC.
The idea of measuring median student improvement from year to year might appear to have some merit, until you talk to students and teachers involved. You discover that many of the older students see no reason to take the tests seriously; they bubble in, or click on, answers as fast as possible, without reading the questions, in order to be free to leave the room and go do something else. Any results from that test are simply unreliable, and it is simply not possible to tell whether DC education policies have improved over time based on the PARCC, DC-CAS, SAT-9, or CTBS, no matter what sort of fancy statistical procedures are employed.
With the NAEP, on the other hand, there has never been any suggestion of impropriety, and the same agency has been devising, administering, and scoring these tests for decades. We have no other nation-wide test that has been systematically given to a random sample of students for any length of time.
Obviously the 4th or 8th graders who took the NAEP in 2017 were not the same ones who took it in 2015. (Duh!) However, we do in fact have a record of NAEP scores in every state and DC since the 1990s, and they are also broken down by lots of subgroups. Obviously DC is gentrifying rapidly, and there are more white students in DCPS than there were 10 or 20 years ago. If we trace the various subgroups (say, African-American students, or Hispanics, or students whose parents didn’t finish high school, or whatever group you like), you can watch the trends over time in each subgroup. However, Mr Austin does inadvertently raise one valid point: since the proportion of black students in DC is decreasing, and the proportion of white students with college-educated parents is rising, then the natural conclusion would be that this gentrification has *inflated* overall scores for 4th and 8th grade students in DC (and DCPS), especially since 2007. Which is more evidence that ‘reform’ is not working. Not evidence that we should throw the scores out and ignore them completely.
Those trends show something quite different from what Mayor Bowser keeps proclaiming. For one thing, if you look at the simple graphs that I made (and you can examine the numbers yourselves) you can see that any improvements overall in DC, or for any subgroups, began a decade before the ‘reformers’ took over DC schools. ( see https://bit.ly/2K3UyZ1 to begin poking around.) Secondly, for most of the subgroups, those improvements over time were greater before Rhee was anointed Chancellor. Only two groups had better rates of change AFTER Rhee: white students, and those with parents with college degrees – the ones that are inflating overall scores for DC and DCPS during the last decade.
I would note also that the previous writer’s salary is paid by one of the Reform organizations supported by billionaires Gates and Arnold. You can look at the funding page yourself ( page 3 at https://urbn.is/2II1YQQ ). I suspect that when ‘reform’ advocates say not to look at our one consistent source of educational data, it’s because they don’t like what the data is saying.
Guy Brandenburg
Advertisements

The URI to TrackBack this entry is: https://gfbrandenburg.wordpress.com/2018/05/25/some-debate-in-chevy-chase-dc-on-significance-of-latest-naep-scores/trackback/

RSS feed for comments on this post.

2 CommentsLeave a comment

  1. CTBS -> SAT-9 -> DC–CAS -> PARCC??

    Of course there are published techincal manuals discussing everything including comparisons of the various tests and the succeding tests all were administered in parallel with the previous one, right?

    Why does the expression measuring with an elastic band come to mind?

    Like

    • Elastic band it is. After a number of years with any given test, DC (like other entities) finds that their scores increase, as teachers figure out what needs to be done to get their students’ scores to go up. Then, when the new test was adopted, those scores took a nose dive. This happened repeatedly. Roller coaster or rubber band.

      Like


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 )

Google+ photo

You are commenting using your Google+ 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 )

w

Connecting to %s

%d bloggers like this: