What type of MAP-M are 5th graders getting?

[Anonymous]

Anonymous wrote:So are the scale scores between the different levels of the MAP supposed to be more or less equivalent? There’s the one they take in K-2, the one they take 3-5 and then 6+… if a kid scores say a 250 on any of these is that score “supposed” to mean the same thing?

Yes, with the caveat that MAP scores are extremely lossy. The overall score is an average of 4 subscores, and it is not a single-level test. (Contrast against IAAT which is specifically targeted at Algebra 1)
It uses statistical woo to arrive at a statistical "level".

In practical terms, this means that someone who has learned more math but is sloppy gets the same score as someone who has learned less math but is more careful and precise, despite the fact that those two students have differing needs for their near-term math education.

For MAP-M 3-5 vs 6+, it means that sloppy students who already know some HS Algebra and HS Geometry lose points on 3-5 for sloppiness but don't get points for Algebra and Geometry.

Careful students who only know Math 6 get points for accuracy but don't lose points for not knowing Algebra.

This doesn't matter much for Algebra placement, but explains why many students see a score drop when they switch to 6+ test.

[Anonymous]

How many kids are in the pool for TPMS lottery?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

Thank you. Still, what ballpark raw score are we talking about in W feeders? I know they are not all the same, either.

250.

NWEA reports that "proficient" (barely passing, preparation for community college, not UVA/VT engineering) in Algebra is 238.

But even among students who already score exactly 238 in prior spring, 30% of them have a lower (non-proficient) score a year later after taking Algebra!

250 in prior year spring is nearly certain predictor of Proficient 238 at end of Algebra year. (And remember, Proficient is the minimum, not the goal.)

https://www.nwea.org/resource-center/guide/75475/When-are-students-ready-for-algebra-1_NWEA_guide.pdf/

Low FARMs and low moderate FARMS groups perform similarly (within one point) so the cutoff will be similar. 250 I believe is roughly 99 percentile and low FARMs schools cutoff is unlikely to be that RIT.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

Thank you. Still, what ballpark raw score are we talking about in W feeders? I know they are not all the same, either.

250.

NWEA reports that "proficient" (barely passing, preparation for community college, not UVA/VT engineering) in Algebra is 238.

But even among students who already score exactly 238 in prior spring, 30% of them have a lower (non-proficient) score a year later after taking Algebra!

250 in prior year spring is nearly certain predictor of Proficient 238 at end of Algebra year. (And remember, Proficient is the minimum, not the goal.)

https://www.nwea.org/resource-center/guide/75475/When-are-students-ready-for-algebra-1_NWEA_guide.pdf/

Low FARMs and low moderate FARMS groups perform similarly (within one point) so the cutoff will be similar. 250 I believe is roughly 99 percentile and low FARMs schools cutoff is unlikely to be that RIT.

Not true at all. We were at a high farms school and multiple kids got 250 or higher.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

You sound very knowledgeable but there is no where in MCPS where the top 15 percent are scoring above 99th percentile. That’sa ridiculous statement.

I did say the cutoff can be rather high, but I did not say that there was or wasn't an MCPS grouping locally normed 85th percentile above the 99th national norm percentile. I said the paradigm would prevent anyone at the 99th percentile nationally from being excluded from the lottery pool.

DCUM, of course, may suggest that there is such a distribution at their school.

Direct quote from your post “even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. ”

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

Thank you. Still, what ballpark raw score are we talking about in W feeders? I know they are not all the same, either.

250.

NWEA reports that "proficient" (barely passing, preparation for community college, not UVA/VT engineering) in Algebra is 238.

But even among students who already score exactly 238 in prior spring, 30% of them have a lower (non-proficient) score a year later after taking Algebra!

250 in prior year spring is nearly certain predictor of Proficient 238 at end of Algebra year. (And remember, Proficient is the minimum, not the goal.)

https://www.nwea.org/resource-center/guide/75475/When-are-students-ready-for-algebra-1_NWEA_guide.pdf/

Low FARMs and low moderate FARMS groups perform similarly (within one point) so the cutoff will be similar. 250 I believe is roughly 99 percentile and low FARMs schools cutoff is unlikely to be that RIT.

Not true at all. We were at a high farms school and multiple kids got 250 or higher.

Yeah, in that magical place where everyone knows everyone’s scores accurately. Unless you are a teacher?

[Anonymous]

Anonymous wrote:

Anonymous wrote:So are the scale scores between the different levels of the MAP supposed to be more or less equivalent? There’s the one they take in K-2, the one they take 3-5 and then 6+… if a kid scores say a 250 on any of these is that score “supposed” to mean the same thing?

Yes, with the caveat that MAP scores are extremely lossy. The overall score is an average of 4 subscores, and it is not a single-level test. (Contrast against IAAT which is specifically targeted at Algebra 1)
It uses statistical woo to arrive at a statistical "level".

In practical terms, this means that someone who has learned more math but is sloppy gets the same score as someone who has learned less math but is more careful and precise, despite the fact that those two students have differing needs for their near-term math education.

For MAP-M 3-5 vs 6+, it means that sloppy students who already know some HS Algebra and HS Geometry lose points on 3-5 for sloppiness but don't get points for Algebra and Geometry.

Careful students who only know Math 6 get points for accuracy but don't lose points for not knowing Algebra.

This doesn't matter much for Algebra placement, but explains why many students see a score drop when they switch to 6+ test.

So how many fifth graders in MCPS get 250+ in the Fall? There are around 12,000 taking the test, right? If this is 250 is close to locally normed 85th percentile, it would mean about 1800 had this score? Could this really be true? I find that hard to believe.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

Thank you. Still, what ballpark raw score are we talking about in W feeders? I know they are not all the same, either.

250.

NWEA reports that "proficient" (barely passing, preparation for community college, not UVA/VT engineering) in Algebra is 238.

But even among students who already score exactly 238 in prior spring, 30% of them have a lower (non-proficient) score a year later after taking Algebra!

250 in prior year spring is nearly certain predictor of Proficient 238 at end of Algebra year. (And remember, Proficient is the minimum, not the goal.)

https://www.nwea.org/resource-center/guide/75475/When-are-students-ready-for-algebra-1_NWEA_guide.pdf/

Low FARMs and low moderate FARMS groups perform similarly (within one point) so the cutoff will be similar. 250 I believe is roughly 99 percentile and low FARMs schools cutoff is unlikely to be that RIT.

Not true at all. We were at a high farms school and multiple kids got 250 or higher.

Multiple kids your kid knows, multiple kids in their grade, or classroom or....?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:So are the scale scores between the different levels of the MAP supposed to be more or less equivalent? There’s the one they take in K-2, the one they take 3-5 and then 6+… if a kid scores say a 250 on any of these is that score “supposed” to mean the same thing?

Yes, with the caveat that MAP scores are extremely lossy. The overall score is an average of 4 subscores, and it is not a single-level test. (Contrast against IAAT which is specifically targeted at Algebra 1)
It uses statistical woo to arrive at a statistical "level".

In practical terms, this means that someone who has learned more math but is sloppy gets the same score as someone who has learned less math but is more careful and precise, despite the fact that those two students have differing needs for their near-term math education.

For MAP-M 3-5 vs 6+, it means that sloppy students who already know some HS Algebra and HS Geometry lose points on 3-5 for sloppiness but don't get points for Algebra and Geometry.

Careful students who only know Math 6 get points for accuracy but don't lose points for not knowing Algebra.

This doesn't matter much for Algebra placement, but explains why many students see a score drop when they switch to 6+ test.

So how many fifth graders in MCPS get 250+ in the Fall? There are around 12,000 taking the test, right? If this is 250 is close to locally normed 85th percentile, it would mean about 1800 had this score? Could this really be true? I find that hard to believe.

The locally normed score is based on the norming group. MCPS says they group schools with similar FARMS rates together. (This is in the faq on this websitehttps://www.montgomeryschoolsmd.org/curriculum/specialprograms/middle) They separate into several norming groups - a PP says there are 5? That means the top 15% of the scores are in the 85th percentile for that group. So, if the group of low farms schools has 1,000 kids, 150 kids receive the 85th percentile or above. Another group might have 5,000 kids, so 750 kids would be in the 85th percentile or above. Each group will have a different cut off for the 85 percentile based on the actual scores students receive. It will also change year to year, as they norm it each year. Probably why they don't publish it because while the percentile stays the same, the associated RIT score is different for each grouping of schools each year.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:The 3-5 and 6+ tests, on average, are supposed to result in the same RIT score and the same percentile. However, the individual variation expected is among the reasons that MCPS central decided to count both the 4th-grade spring MAP-M and 5th-grade fall MAP-M (whichever locally normed percentile is higher) when evaluating those in accelerated math for the criteria-based magnet MS pool eligibility and recommended auto-placement in AMP7+/AIM in 6th, whichever is offered (each leads to Algebra in 7th).

This tends to compensate for the decision to administer the 6+ test to that population in 5th, which is consistent with what MAP is supposed to be used for -- evaluation of teaching effectiveness across populations and identification of possible strengths/weaknesses to guide teachers' approach to meet students' individual needs (not so much as a spot test without other evaluative tools for placement decisions).

MS placement in AMP7+/AIM is not limited to this centrally-identified population. Grades and later standardized test performance, along with the input of the elementary teacher/math specialist, typically are taken into consideration as MS classes are planned late in the school year for the following one. There is some variation among middle schools in what qualifies, often due to the need to keep relatively consistent/manageable class sizes, but there usually is also some flexibility in placement.

So what is the cutoff for the criteria-based magnet eligibility?

85th percentile, locally normed by FARMS-rate grouping (5 of those, as of last report, from High-FARMS to Low-FARMS). If all the group-FARMS elementaries have a distribution where it takes an RIT score of x to be at the 85th percentile among all the scores in that grouping (not nationally), then x is the locally normed 85th percentile. Sort of, that is. MCPS goes back to the national norms, finds the percentile there that most closely matches the strict 85th percentile RIT score in the FARMS-rate grouping and then establishes the RIT score from that national percentile as the 85th percentile local norm for lottery placement purposes.

As might be expected, the cutoff can be rather high at low-FARMS schools, where the cohort typically has extra opportunities for exposure to content. That last bit, though, going back to the national percentiles, ensures that anyone hitting 99th percentile nationally is not excluded, even if the strict locally normed 85th percentile might have been at an RIT score above the national 99th. (Avoiding reliance on expected statistical variation at the extremes.)

It also can vary from year to year, as can the FARMS rate group of a particular school. MCPS has not made the actual cutoffs public since a response to a request from MCCPTA a couple of years back, but parents have been able to make rough guesses after the fact by comparing scores of who got in at their school. The understanding from presentations to the BOE is that the cutoff becomes the locally normed 70th percentile for individual students receiving services (FARMS, EML, IEP or 504).

The idea is to try to identify students with ability/need for more. That used to be done with CogAT, an abilities-related test, but they had to abandon that during the pandemic and eventually went with MAP. MAP is largely an exposure-related test, and there's info on the NWEA website about not using MAP for identification/placement purposes without an ability-related measure (NWEA is the organization from which MAP originates). There also is reasoning there for using local norms, though that is often discussed by school instead of by school grouping (the same thoughts, roughly, should apply, however).

This lottery/placement paradigm was supposed to be reviewed, but, while there have been internal discussions/decisions from year to year (such as changing to use both the end of 4th and beginning of 5th MAP-M scores for those in accelerated math to adjust for using the 6+ test for those students), resourcing a formal review has been a challenge.

Thank you. Still, what ballpark raw score are we talking about in W feeders? I know they are not all the same, either.

250.

NWEA reports that "proficient" (barely passing, preparation for community college, not UVA/VT engineering) in Algebra is 238.

But even among students who already score exactly 238 in prior spring, 30% of them have a lower (non-proficient) score a year later after taking Algebra!

250 in prior year spring is nearly certain predictor of Proficient 238 at end of Algebra year. (And remember, Proficient is the minimum, not the goal.)

https://www.nwea.org/resource-center/guide/75475/When-are-students-ready-for-algebra-1_NWEA_guide.pdf/

Low FARMs and low moderate FARMS groups perform similarly (within one point) so the cutoff will be similar. 250 I believe is roughly 99 percentile and low FARMs schools cutoff is unlikely to be that RIT.

Not true at all. We were at a high farms school and multiple kids got 250 or higher.

Yeah, in that magical place where everyone knows everyone’s scores accurately. Unless you are a teacher?

Kids talk

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:So are the scale scores between the different levels of the MAP supposed to be more or less equivalent? There’s the one they take in K-2, the one they take 3-5 and then 6+… if a kid scores say a 250 on any of these is that score “supposed” to mean the same thing?

Yes, with the caveat that MAP scores are extremely lossy. The overall score is an average of 4 subscores, and it is not a single-level test. (Contrast against IAAT which is specifically targeted at Algebra 1)
It uses statistical woo to arrive at a statistical "level".

In practical terms, this means that someone who has learned more math but is sloppy gets the same score as someone who has learned less math but is more careful and precise, despite the fact that those two students have differing needs for their near-term math education.

For MAP-M 3-5 vs 6+, it means that sloppy students who already know some HS Algebra and HS Geometry lose points on 3-5 for sloppiness but don't get points for Algebra and Geometry.

Careful students who only know Math 6 get points for accuracy but don't lose points for not knowing Algebra.

This doesn't matter much for Algebra placement, but explains why many students see a score drop when they switch to 6+ test.

So how many fifth graders in MCPS get 250+ in the Fall? There are around 12,000 taking the test, right? If this is 250 is close to locally normed 85th percentile, it would mean about 1800 had this score? Could this really be true? I find that hard to believe.

The locally normed score is based on the norming group. MCPS says they group schools with similar FARMS rates together. (This is in the faq on this websitehttps://www.montgomeryschoolsmd.org/curriculum/specialprograms/middle) They separate into several norming groups - a PP says there are 5? That means the top 15% of the scores are in the 85th percentile for that group. So, if the group of low farms schools has 1,000 kids, 150 kids receive the 85th percentile or above. Another group might have 5,000 kids, so 750 kids would be in the 85th percentile or above. Each group will have a different cut off for the 85 percentile based on the actual scores students receive. It will also change year to year, as they norm it each year. Probably why they don't publish it because while the percentile stays the same, the associated RIT score is different for each grouping of schools each year.

I understand the procedure, but those are the details. The bottom line is that, on each school bus there is, like a dozen kids that belong to Mensa. I just find that hard to believe. I went to an extremely selective magnet in HS and I knew kids who were below 98th percentile. Suddenly, nobody is. It blows mind if true.

[Anonymous]

Sorry for being unclear. 250 is for Algebra readiness, not for magnet lottery.

[Anonymous]

TPMS Magnet lottery is probably about 240.
If your kid is in a higher FARMS school, it might have a lower cutoff, but this kid will have a real bad time if they attend.

[Anonymous]

Anonymous wrote:Sorry for being unclear. 250 is for Algebra readiness, not for magnet lottery.

It doesn't matter. I mean, it matters, and thank you for clarifying, but you are still basically saying that, at low, maybe low medium FARMS, top 30% kids (1 in 3 kids, people) are in top 2% nationally (what used to be mensa cutoff). Right? That's just insane.

[Anonymous]

Anonymous wrote:TPMS Magnet lottery is probably about 240.
If your kid is in a higher FARMS school, it might have a lower cutoff, but this kid will have a real bad time if they attend.

How does the lottery work? What happens to seats that are declined?