Did the Takoma MS magnet got MORE white this year?

[Anonymous]

PP again. Here's one more write-up of a useful study:

http://blogs.edweek.org/teachers/teaching_now/2016/03/bias.html

"Published in the journal Economics of Education Review, the "Who Believes in Me?" study was compiled to investigate how teachers form expectations for students, whether those expectations are systematically biased, and whether they are affected by racial differences.

The findings are based largely on data from the Educational Longitudinal Study of 2002, an ongoing study following 8,400 10th grade public school students. For the survey, two different math or reading teachers, who each taught the same student, were asked to guess how far that one student would go in school.

The findings show that with white students, evaluations from both teachers were about the same. But for black students, white teachers had lower expectations than black teachers."

[Anonymous]

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

[Anonymous]

Anonymous wrote:PP again. Here's one more write-up of a useful study:

http://blogs.edweek.org/teachers/teaching_now/2016/03/bias.html

"Published in the journal Economics of Education Review, the "Who Believes in Me?" study was compiled to investigate how teachers form expectations for students, whether those expectations are systematically biased, and whether they are affected by racial differences.

The findings are based largely on data from the Educational Longitudinal Study of 2002, an ongoing study following 8,400 10th grade public school students. For the survey, two different math or reading teachers, who each taught the same student, were asked to guess how far that one student would go in school.

The findings show that with white students, evaluations from both teachers were about the same. But for black students, white teachers had lower expectations than black teachers."

bottom line recs are biased and this should be based on objective criteria

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

Do you know they used inconsistent test scores or is this just an assumption?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

Do you know they used inconsistent test scores or is this just an assumption?

This data hasn't been released so this is purely speculation and unfounded.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

Do you know they used inconsistent test scores or is this just an assumption?

This data hasn't been released so this is purely speculation and unfounded.

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

[Anonymous]

Anonymous wrote:

Anonymous wrote:Absolutely meaningful because these people are trained teachers. They understand the curriculum and have many hours to observe the child in an academic setting.

There is subjectivity in many fields/processes (grant selection, academic publishing, college apps, job application, the arts, etc etc), we don't throw out the evaluation because of subjectivity.

There should be a breadth of recs to avoid discrimination if one reviewer is overly subjective.

They are also the ones who really do have a sense of the relative performance, maturity, work ethic, and aptitude of each student. It is not a parent who only sees her child and maybe a handful of child's friends.
We loved our pre-schools report cards and commentary; they really knew our child and had suggestions on EC classes or things to work on each trimester.

Yes, this.

No one is saying recs are perfect, but they should be a part of an evaluation. If studies show there is racial bias, consider them different for students of color and work on dealing with bias. But, ignoring the evaluation of the person/people who we pay to evaluate our kids academically is not effective. Tests are biased. Should we throw out tests? How do we evaluate? We need a variety of data and the wisdom to interpret it.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

Do you know they used inconsistent test scores or is this just an assumption?

This data hasn't been released so this is purely speculation and unfounded.

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

but a "group" is not an individual.

[Anonymous]

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I think people want easy answers, and there aren't any.

It would be EASY to just use CogAT or a similar test to "fine tune" peer groups, but that assumes CoGAT is the right tool for identifying gifted kids, and/or kids with a large amount of potential.

We all know that the smartest kids we went to school with are not necessarily the most successful today, and, as young people get older, working hard and staying focused becomes almost as important as having "gifts."

Additionally, there are multiple ways to be "gifted" and giftedness is often present in some areas and not others. This means that tests like the CogAT will be "spiky" for some kids, which makes an easy solution hard.

If you take the kid who has a 99% across the board but leave behind the kid who has a 99.99999% on one subtest but a 75% in another, does that serve the second kid?

These are HARD questions, and we haven't even touched the ways in which the tests themselves are culturally biased.

I know we all want a hard and fast rule, but that's not how life works.

I agree with a lot of what you say. I think that for the CES programs it definitely makes sense to include as many children who show "promise" even if it is just in one area. The middle and high school programs are really programs for kids who are not just gifted but also high achieving. They are very challenging academic programs with a heavy work load and high teacher expectations. A "spiky" kid or a kid who is really into one subject but is not otherwise engaged in school would have a tough time. There are some kids like this who are so exceptional that there is no other place in the school system for them and they should be in the magnet programs but this is a relatively small group of kids.
I am not sure that is what happened this year. It sounds like a lot of 99+ kids were passed over not because there were spiky kids in other schools who deserved those spots but simply because of where they lived.

Great question. But didn't the county do the same with the magnet acceptance process?

Took a lot of kids who they single-handedly defined as "strong potential" based on inconsistent test scores and home schools without peer cohort.

Do you know they used inconsistent test scores or is this just an assumption?

This data hasn't been released so this is purely speculation and unfounded.

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

but a "group" is not an individual.

oh yes, I keep hearing this argument, but when it's pointed out that MCPS is looking at "group", ie cohort, for the "individual" high achieving student from the west side, well, that's ok to use group statistics then.

You want to use group statistics to have a different admittance criteria to magnets (ie not enough URM representation), but you don't want to use group statistics to show which group has a higher probability/statistics to be represented in the magnets.

In other words, only use those statistics when it favors your side but not when it doesn't. You can't have it both ways.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

? you just made the argument for me regarding "peer cohort" vs individual performance, and that they are using location as a proxy for race.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

? you just made the argument for me regarding "peer cohort" vs individual performance, and that they are using location as a proxy for race.

Where did you get your information? Do you have a credible source inside the Central Office?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

? you just made the argument for me regarding "peer cohort" vs individual performance, and that they are using location as a proxy for race.

Where did you get your information? Do you have a credible source inside the Central Office?

This is crazytown.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

? you just made the argument for me regarding "peer cohort" vs individual performance, and that they are using location as a proxy for race.

Where did you get your information? Do you have a credible source inside the Central Office?

The MS that were most affected by the "peer cohort" nonsense are the ones with low URM and FARMs. It's not really that hard to connect those dots.

It's kind of like how the Rs removed the SALT Deductions. It's not hard to connect the dots that they took away this deduction to hurt the coastal blue states, , which disproportionately has high state income and property tax.

I don't think MCPS did this deliberately to go after wealthy Asian/white students. However, it just so happens that many of these students are wealthy white/asian. In this case, race and location are closely linked together, like how high state and property tax are closely linked with liberal coastal elites.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

MCPS's own statistics shows which group has high test scores and which don't. Apply probability and statistics to the numbers. If the admittance breakdown to magnets is out of whack with MCPS's own statistics, then that means they are not accepting top performers, and indeed, MCPS has admitted that they look at peer cohort. There is no question about this fact.

But that's not true if there truly is a deficit of seats compared to highly qualified students. If, for example, you have 100 Asian students evaluated and 20 score at whatever your target threshold is (99% or whatever raw score) and you have 100 of some other race evaluated and only 5 score at your SAME target threshold, but you have 10 seats in the program, you can offer those seats to the 5 students of some other race and to 5 of the Asian students and everyone admitted will have met the same threshold for high performance. There will just be 15 Asian kids who also met the same threshold but didn't get admitted. And if those 15 all go to the same school, they will hopefully be a peer cohort in class with each other encouraging each other to excel. Obviously no real life situation is as simple as a stripped-down example, but if there are not enough seats for all of the highly able students (which everyone seems to agree is the case) then it is entirely possible for the student body selected not to mirror the racial percentages of the entire pool (either of MCPS students or of MCPS students who score X on any particular metric) and for each admitted student to still be eminently qualified and not erode the quality of the program at all.

? you just made the argument for me regarding "peer cohort" vs individual performance, and that they are using location as a proxy for race.

Where did you get your information? Do you have a credible source inside the Central Office?

Why did MCPS put most of the magnets in the lower income (and predominately URM) schools if not because of using location (and race) as the reason?

If MCPS can use location and race as a qualifier for putting in magnets programs in those places, why wouldn't they use location/race as a qualifier for students to be admitted in those magnets?