Did MCPS do a sneaky thing for the magnet lotteries?

[Anonymous]

Anonymous wrote:

Anonymous wrote:All-
Read this PP's explanation. It's an excellent and really shows that either MCPS had its thumb on the scales as someone put it so well or that the whole premise that there is an achievement gap is wrong. You can't have both and have numbers come out like that.

The achievement gap is real. Look at Math 5 Performance Level 5 for various races:
https://reportcard.msde.maryland.gov/Graphs/#/Assessments/MathPerformance/2MA/5/12/3/1/15/XXXX/2019

I think that's the point. That the achievement gap is real therefore there's no way it was a completely random unweighted lottery. There are other factors like the high rate of FARMS admissions that show it's mathematically not random.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:It's pretty easy. Just look at Takoma Park placed data this year:

Asian 20.8%
Black 20.8%
Hispanic 16.0%
White 37.6%

Is this how the MAP math 85%+ scorers are distributed downcounty?

If this is the case, there is no achievement gap in math.

The state used to publish PARCC scores in five levels distributed by race. If you compare that distribution with this distribution, you will see that this is a heavily manipulated outcome.

Good points. One thing to keep in mind with the downcounty MS magnets is that from my experience very few kids from ‘W’ cluster schools elect to attend, at least I know this for Eastern. Maybe TPMS is different for Asian students who want to proceed to Blair? But in our case, coming out of CES, every parent from a ‘W’ cluster is except WJ would not consider the magnet or actually declined a seat when selected. Maybe you miss some computer science, but you still take AIM at your home school and MCPS has been increasing the number of “advanced” courses offered at home schools.

Not so sure it's all that good a point unless the state published PARCC data so that it could be locally normed otherwise it's just a very rough guess that tells us little.

The “good point” is that to have an outcome that results in a racial demographic that mirrors the county’s racial demographic for Black and Hispanic students would presume that those demographics were reflected as such in the lottery pool. Since the pool is the “top 15%” then the top 15% of student is highly racially balanced. That means that there is no achievement gap. Or it means that there is an achievement gap but MCPS put a thumb on the scale to ensure a racially balanced cohort. It cannot be both. If people want to keep saying “random lottery” then the implication is that there is no achievement gap.

Of course, they put a thumb on the scale. They did exactly what they said and used a local norm. There is no big conspiracy here. They made it perfectly clear how they did this.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:All-
Read this PP's explanation. It's an excellent and really shows that either MCPS had its thumb on the scales as someone put it so well or that the whole premise that there is an achievement gap is wrong. You can't have both and have numbers come out like that.

The achievement gap is real. Look at Math 5 Performance Level 5 for various races:
https://reportcard.msde.maryland.gov/Graphs/#/Assessments/MathPerformance/2MA/5/12/3/1/15/XXXX/2019

But if these scores were locally norned the distribution would look very different so I'm not sure this tells us anything about their admissions.

Right. The whole point of that smart PP's post. That they must have done things like local norming, extra points for ever FARMS and other things so that it wasn't a straight random lottery of everyone in the pool weighted equally.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:All-
Read this PP's explanation. It's an excellent and really shows that either MCPS had its thumb on the scales as someone put it so well or that the whole premise that there is an achievement gap is wrong. You can't have both and have numbers come out like that.

The achievement gap is real. Look at Math 5 Performance Level 5 for various races:
https://reportcard.msde.maryland.gov/Graphs/#/Assessments/MathPerformance/2MA/5/12/3/1/15/XXXX/2019

I think that's the point. That the achievement gap is real therefore there's no way it was a completely random unweighted lottery. There are other factors like the high rate of FARMS admissions that show it's mathematically not random.

It doesn't mean that at all. If there's 35% FARMS in the county and they locally norm scores at many schools with 60%-70% FARMS this is bound to happen. No mystery here. Just basic math.

[Anonymous]

We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:All-
Read this PP's explanation. It's an excellent and really shows that either MCPS had its thumb on the scales as someone put it so well or that the whole premise that there is an achievement gap is wrong. You can't have both and have numbers come out like that.

The achievement gap is real. Look at Math 5 Performance Level 5 for various races:
https://reportcard.msde.maryland.gov/Graphs/#/Assessments/MathPerformance/2MA/5/12/3/1/15/XXXX/2019

But if these scores were locally norned the distribution would look very different so I'm not sure this tells us anything about their admissions.

Right. The whole point of that smart PP's post. That they must have done things like local norming, extra points for ever FARMS and other things so that it wasn't a straight random lottery of everyone in the pool weighted equally.

They said they were using local norming and ever-FARMS, though. That was in the information provided. They also considered whether a child had an IEP in some cases.

[Anonymous]

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

I've been on this board for a long time, with kids in the GT programs, and I've never seen anything like that.

[Anonymous]

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

My kid who is in the TPMS magnet told me today that one of their classmates who said they scored in the 230s in 5th barely broke 200 on the fall map in 6th. I'm no expert but suspect they'd be better off in Math 6 than magnet math.

[Anonymous]

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

Do you have data for this? I recall that 2 years ago we were looking at locally normed CogAT, not MAP tests. It is important to realize that the local norms didn't include the full MCPS population, just those who were identified and tested for admission. So yes, a 97th percentile nationally was lower in the local norming because the local norming population had already been pre-screened and already represented only the top half of the population. I do not recall any instances where 75th percentile nationally was 90th percentile locally, and I don't believe that is true; in fact, I don't remember any reports of a national percentile being dramatically lower than a local percentile, but maybe I'm forgetting.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:It's pretty easy. Just look at Takoma Park placed data this year:

Asian 20.8%
Black 20.8%
Hispanic 16.0%
White 37.6%

Is this how the MAP math 85%+ scorers are distributed downcounty?

If this is the case, there is no achievement gap in math.

The state used to publish PARCC scores in five levels distributed by race. If you compare that distribution with this distribution, you will see that this is a heavily manipulated outcome.

Good points. One thing to keep in mind with the downcounty MS magnets is that from my experience very few kids from ‘W’ cluster schools elect to attend, at least I know this for Eastern. Maybe TPMS is different for Asian students who want to proceed to Blair? But in our case, coming out of CES, every parent from a ‘W’ cluster is except WJ would not consider the magnet or actually declined a seat when selected. Maybe you miss some computer science, but you still take AIM at your home school and MCPS has been increasing the number of “advanced” courses offered at home schools.

I don't think this parent actually has a student at Eastern because if you did you'd know the buses from BCC, Whitman and Churchill area are completely full and there are many kids going to Eastern not just TPMS and Blair. If you look at the directory, a very large number are Potomac or Bethesda addresses.

Because I apparently have infinite time on a Friday afternoon, here is the breakdown of the top 5 zip codes for every child in the directory:

20910 - 33
20910 - 19
20902 - 17
20854 - 14
20895 - 13

So, the first three zip codes are Silver Spring and Takoma Park. The next is Potomac and the last is Kensington. That's great, and I have a lot of admiration for the kids willing to ride the bus all the way from Potomac. But it's actually pretty balanced and certainly the numbers above *feel* true to me as a two-time Eastern magnet parent who knows other parents and whose kids hang out with other Eastern kids. A lot of them are thankfully right in the neighborhood.

[Anonymous]

Anonymous wrote:

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

Do you have data for this? I recall that 2 years ago we were looking at locally normed CogAT, not MAP tests. It is important to realize that the local norms didn't include the full MCPS population, just those who were identified and tested for admission. So yes, a 97th percentile nationally was lower in the local norming because the local norming population had already been pre-screened and already represented only the top half of the population. I do not recall any instances where 75th percentile nationally was 90th percentile locally, and I don't believe that is true; in fact, I don't remember any reports of a national percentile being dramatically lower than a local percentile, but maybe I'm forgetting.

Here's the actual thread from 2 years ago (my kid was part of this group, so I remembered the thread) -- scores were reported starting on page 9 or 10: https://www.dcurbanmom.com/jforum/posts/list/120/792860.page
I flipped through about 8 pages and there were no reports of a national percentile being lower than a local percentile.

[Anonymous]

Anonymous wrote:

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

Do you have data for this? I recall that 2 years ago we were looking at locally normed CogAT, not MAP tests. It is important to realize that the local norms didn't include the full MCPS population, just those who were identified and tested for admission. So yes, a 97th percentile nationally was lower in the local norming because the local norming population had already been pre-screened and already represented only the top half of the population. I do not recall any instances where 75th percentile nationally was 90th percentile locally, and I don't believe that is true; in fact, I don't remember any reports of a national percentile being dramatically lower than a local percentile, but maybe I'm forgetting.

I have a child who attended a high poverty school, so I saw the "local normed" cogat and the national cogat for the same raw score. They were the same. So unless someone comes out of the woodwork to say that there was a huge difference between the locally normed number for a rich school and the national percentile, then I don't think we can trust PP's narrative.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:All-
Read this PP's explanation. It's an excellent and really shows that either MCPS had its thumb on the scales as someone put it so well or that the whole premise that there is an achievement gap is wrong. You can't have both and have numbers come out like that.

The achievement gap is real. Look at Math 5 Performance Level 5 for various races:
https://reportcard.msde.maryland.gov/Graphs/#/Assessments/MathPerformance/2MA/5/12/3/1/15/XXXX/2019

I think that's the point. That the achievement gap is real therefore there's no way it was a completely random unweighted lottery. There are other factors like the high rate of FARMS admissions that show it's mathematically not random.

I agree the achievement gap is real but this theory kind of breaks down when you realize they selected students based on local norms.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:We did see 2 years ago that there was a big difference in local norms. A child who scored 97th percentile in MAP would only be 90th in the wealthy band but in the low SES band you could score 75th percentile and actually be 90th percentile.

Do you have data for this? I recall that 2 years ago we were looking at locally normed CogAT, not MAP tests. It is important to realize that the local norms didn't include the full MCPS population, just those who were identified and tested for admission. So yes, a 97th percentile nationally was lower in the local norming because the local norming population had already been pre-screened and already represented only the top half of the population. I do not recall any instances where 75th percentile nationally was 90th percentile locally, and I don't believe that is true; in fact, I don't remember any reports of a national percentile being dramatically lower than a local percentile, but maybe I'm forgetting.

I have a child who attended a high poverty school, so I saw the "local normed" cogat and the national cogat for the same raw score. They were the same. So unless someone comes out of the woodwork to say that there was a huge difference between the locally normed number for a rich school and the national percentile, then I don't think we can trust PP's narrative.

At our W feeder the locally normed Cogat was 10 points higher.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:It's pretty easy. Just look at Takoma Park placed data this year:

Asian 20.8%
Black 20.8%
Hispanic 16.0%
White 37.6%

Is this how the MAP math 85%+ scorers are distributed downcounty?

If this is the case, there is no achievement gap in math.

The state used to publish PARCC scores in five levels distributed by race. If you compare that distribution with this distribution, you will see that this is a heavily manipulated outcome.

Good points. One thing to keep in mind with the downcounty MS magnets is that from my experience very few kids from ‘W’ cluster schools elect to attend, at least I know this for Eastern. Maybe TPMS is different for Asian students who want to proceed to Blair? But in our case, coming out of CES, every parent from a ‘W’ cluster is except WJ would not consider the magnet or actually declined a seat when selected. Maybe you miss some computer science, but you still take AIM at your home school and MCPS has been increasing the number of “advanced” courses offered at home schools.

I don't think this parent actually has a student at Eastern because if you did you'd know the buses from BCC, Whitman and Churchill area are completely full and there are many kids going to Eastern not just TPMS and Blair. If you look at the directory, a very large number are Potomac or Bethesda addresses.

Because I apparently have infinite time on a Friday afternoon, here is the breakdown of the top 5 zip codes for every child in the directory:

20910 - 33
20910 - 19
20902 - 17
20854 - 14
20895 - 13

So, the first three zip codes are Silver Spring and Takoma Park. The next is Potomac and the last is Kensington. That's great, and I have a lot of admiration for the kids willing to ride the bus all the way from Potomac. But it's actually pretty balanced and certainly the numbers above *feel* true to me as a two-time Eastern magnet parent who knows other parents and whose kids hang out with other Eastern kids. A lot of them are thankfully right in the neighborhood.

There is no way places like Takoma or SS could possibly have more students than the W's!!!!!