Kid didn’t finish MAP-R?

[Anonymous]

Anonymous wrote:Do they definitely make sure each kid finishes? My kid (new to middle school) didn't finish his math MAP from Monday and it's Friday, and he said they still haven't told him when he'll get time to finish it.

My kid is new to middle school too. They and a number of their friends didn't finish the MAP-M test. They had MAP testing Tuesday/Wednesday of this week and haven't heard anything about when they'll get the chance to finish. This is pretty normal. I know they had to wait a few days after starting MAP tests in elementary school to get a chance to go back to it. I know my kid's school has homeroom a few times a week, which is where it seems they normally take reassessments and things. I figure that's when they'll pick it up back next week. I have also heard of students having to finish during lunch. My kid had to do that once or twice in elementary school when he still wasn't done with his MAP-M after a second period, but he would score really high so his questions tended to be a bit harder to work out and take longer to do as a result.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

"No, because MAP scores are already +/- 3 margin of error in RIT score"

This is the margin of error for an individual score based on the sample size of the question pool. The errors in this study are based on 1000s of kids and are negotiable. There is a substantial drop on average scores (4 points) in some grades. This will significantly affect average percentiles for individual kids.

What's outrageous is MCPS using a test for a purpose well outside its intended scope of applicability.

The fact remains that MCPS is using that for that purpose and that this will affect significant number of children in unknown way. Pretending that averages are all that matters is ridiculous.

If everyone drops 4 points, the percentiles don't change.

4 points is an average drop. doesn't mean that everybody remain in the same spot.

also, these percentiles come from a different population using a different test. you can end up with the appearance of children dropping on the percentile scale where there is not drop in actuality. some will have spring test to refer to but some won't (because they came from a different district).

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:My kid would often take multiple sessions. They were a 99%ile scorer pretty often. The test goes and goes and goes until it feels the kid is out of their depth.

Just as a heads up, NWEA changed their algorithms this year on the MAP tests. Their algorithms now emphasize grade level standards and curriculum. This means that if I child is in sixth grade for instance, they're going to get A LOT of sixth grade questions. More so than before where it was just straight up ability/knowledge of the student with no limit to being given higher grade level questions. Because of this focus on grade-level material, there is less opportunity for children who generally test higher to get tougher, above grade level questions, resulting in lower scores.

My child had their MAP-R test this week. They normally score 98-99th percentile. They said the screen to log in noticeably said "grade 6" with everyone's names when it never was linked to a grade level before. The test they said was noticeably easier, yet their score dropped about ten points. Everyone in the school was upset, many were crying, and it was the talk of lunch because everyone's scores dropped dramatically (many had drops of 17 points, over 20 points, etc). These kids were those who were scoring very high on MAP-R, had been in CES and/or enrichment the previous year.

It looks like the algorithm change also impacts MAP-M. My child has not finished, but they score 99th percentile and they said the questions were a lot easier than they normally are, and that so far, they were all 6th grade level questions or thereabouts (pythagorean theorem, interquartile ranges, dividing fractions, etc). They kept waiting to get trigonometry questions, like they got last spring, but it hasn't gotten any harder yet. So we anticipate that math scores will go down as well.

Just be prepared if your child's score drops, as it isn't necessarily their effort as much as the algorithm changes now being used by NWEA.

https://njgifted.org/understanding-nwea-maprit-scores/
https://ecragroup.com/2024/05/02/nwea-map-growth-assessment-changes-for-the-2024-2025-school-year/#:~:text=NWEA%20has%20changed%20its%20item,by%20the%20current%20test%20version.

Well that's lovely for those of us who are hoping to get in the lottery for CES this year....

It won't affect the lottery, because the lottery is already set to a cutoff that isn't far above grade level.

True for low SES schools but not all. The cutoff to qualify for the lottery was the 95th percentile a couple years ago at our Bethesda ES.

I think you may have been misinformed. The only data was have is from the MPIA request put in by the MCPTA Gifted and Talented Committee and that showed that the cut-off at the lowest FARMS schools was 92nd percentile for MAP-R and 93rd percentile for MAP-M. That's, again, at the absolute lowest FARMS schools in MCPS.

I'm not misinformed - you just have limited data. Based on my first-hand experience with my DC, the cutoff was 95th percentile for MAP-R the year DC took the test as a 3rd grader. DC scored in the 94th percentile and this disqualified DC from the lottery pool based on DC's locally normed score, per the multiple individuals I spoke to at DC's school and central office. I have no idea what's in the MPIA data that was reviewed, but perhaps it's a limited data set or restricted to a particular year, or not complete.

You are correct that the MCPS responses to the MPIA request applied only to that year's FARMS groupings and locally normed 85th percentiles. One can make some general assumptions about a school's likely FARMS rate grouping and that the percentiles were similar, but MCPS, without providing specifics, acknowledges that each year there are changes both to the FARMS groupings and the locally normed 85th percentiles based on the FARMS data/MAP scores seen from that year.

[Anonymous]

Can someone explain if the questions are easier why the scores are dropping?

[Anonymous]

Anonymous wrote:Can someone explain if the questions are easier why the scores are dropping?

Yea it doesn’t make sense for the high achievers. From NWEA’s research it’s the lower achievers who end up lower here because they’re less likely to see items from significantly below grade level…. It does seem odd that folks are complaining that their high flyers are struggling

[Anonymous]

Anonymous wrote:

Anonymous wrote:Can someone explain if the questions are easier why the scores are dropping?

Yea it doesn’t make sense for the high achievers. From NWEA’s research it’s the lower achievers who end up lower here because they’re less likely to see items from significantly below grade level…. It does seem odd that folks are complaining that their high flyers are struggling

High achievers are not getting the most difficult (advanced) questions which are worth the most points. Therefore, the highest scores are getting lower. But they are still using the old tables, resulting in misleading percentiles.

[Anonymous]

I always thought my 3rd grader is a great reader, probably not because she scores 202 today on MAP R. She can't make it to the lottery with that score.

[Anonymous]

Anonymous wrote:I always thought my 3rd grader is a great reader, probably not because she scores 202 today on MAP R. She can't make it to the lottery with that score.

What score does qualify?

In any case I think it’s winter MAP, not fall, that counts….

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone explain if the questions are easier why the scores are dropping?

Yea it doesn’t make sense for the high achievers. From NWEA’s research it’s the lower achievers who end up lower here because they’re less likely to see items from significantly below grade level…. It does seem odd that folks are complaining that their high flyers are struggling

High achievers are not getting the most difficult (advanced) questions which are worth the most points. Therefore, the highest scores are getting lower. But they are still using the old tables, resulting in misleading percentiles.

Thank you, PPs, for bumping my question.

These high achievers don’t sound like they had a solid understanding of some earlier math concepts.

[Anonymous]

Anonymous wrote:I always thought my 3rd grader is a great reader, probably not because she scores 202 today on MAP R. She can't make it to the lottery with that score.

She sounds like a great reader but the test is about comprehension.

That score looks like it would be above local normed 85 percentile, but I’m not in front of that percentile converter. A nice PP always reposts that link and I saw it on one of these recent threads.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Can someone explain if the questions are easier why the scores are dropping?

Yea it doesn’t make sense for the high achievers. From NWEA’s research it’s the lower achievers who end up lower here because they’re less likely to see items from significantly below grade level…. It does seem odd that folks are complaining that their high flyers are struggling

High achievers are not getting the most difficult (advanced) questions which are worth the most points. Therefore, the highest scores are getting lower. But they are still using the old tables, resulting in misleading percentiles.

Thank you, PPs, for bumping my question.

These high achievers don’t sound like they had a solid understanding of some earlier math concepts.

I don’t think you understand what the PP wrote.

[Anonymous]

Anonymous wrote:

Anonymous wrote:I always thought my 3rd grader is a great reader, probably not because she scores 202 today on MAP R. She can't make it to the lottery with that score.

She sounds like a great reader but the test is about comprehension.

That score looks like it would be above local normed 85 percentile, but I’m not in front of that percentile converter. A nice PP always reposts that link and I saw it on one of these recent threads.

The converter is nationa norm, not local norm

https://sites.google.com/view/nweapercentilecalculator

For local norm you need the FARMS percentile chart

[Anonymous]

Anonymous wrote:Interesting… here is NWEA’s analysis of the new algorithm — https://www.nwea.org/uploads/Research-MAP-Growth-with-enhanced-item-selection-algorithm-updates-on-score-compatibility_NWEA_Research_Guide.pdf

They find minor differences in the RIT and not the dramatic shifts PPs are reporting

Applying the concordance table
Changes in the rate of growth in math relative to the 2020 norms mean that these norms are less
applicable to math scores derived from MAP with EISA and will provide less meaningful context for student
achievement and growth. Therefore, we strongly recommend that districts leverage the concordance
table to support decision-making that is consistent with historical decisions based on the 2020 norms. The
concordance tables should be used until new norms are released in 2025. There are two categories of use
cases for applying the concordance table: adjusting practice and adjusting data.
Adjusting practice. This category of use cases involves adjusting cut scores, such as those used for program-
placement decisions. Given the increases in winter and spring scores, using unadjusted cut scores results
in underidentifying students for intervention services and overidentifying students for talented and gifted
programs.6 By using the concordance study to adjust these cut scores in math, partners can make decisions
about student placement that are more consistent with past decisions, ensuring that students receive the
appropriate level of support. Similarly, the cut scores used for determining which students are on track for
proficiency, based on NWEA linking studies, should also be adjusted based on the concordance table.
Adjusting data. The second use case category involves using the concordance study to adjust individual
scores. It is important to note that the concordance table cannot be applied to aggregate data or
directly to growth metrics. Instead, it must first be applied to individual scores, which can then be used
to recalculate growth metrics at the student, grade, school, or other grouping levels. This ensures those
metrics are not upwardly biased and is strongly recommended when normative growth metrics are used
for evaluation or decision-making. Additionally, adjusting data is necessary when scores will be used in
longitudinal research that spans the transition period

[Anonymous]

NWEA are terrible communicators, but here's the key point.

They are going to prefer to give harder problems on current grade level standards, instead of problems at next grade level up. (And the same idea on the other direction at the low end)

https://connection.nwea.org/s/nwea-news/17-state-expansion-of-the-enhanced-item-selection-algorithm-MCDOB2VK7YABD3LLL5ILP3OZRLCM?language=en_US

How does the enhanced item selection algorithm choose items differently compared to previous versions of MAP Growth tests?

A: These tests will have three significant changes to the item selection algorithm.

(1) Grade level:
Currently, the MAP Growth item selection algorithm emphasizes adaptivity and test content. It creates a test closely matched to a student’s ability level, without factoring in information about a student’s grade
The enhanced item selection algorithm is designed to prioritize items that are closer to a student’s grade level when selecting items on the test. This new algorithm finds an item that is best matched to the student’s ability, grade, and the test content needed to fulfill the test blueprints

waterfall table on item selection

(2) Instructional-Area weighting :
Currently, the MAP Growth item selection algorithm uses equal weighting for each of the instructional areas on the test. This means that the algorithm tries to select an equal number of items from each instructional area
The enhanced item selection algorithm has the ability to adjust the weighting for each of the instructional areas, enabling tests to provide more items in certain instructional areas
(3) Additional content factors:
The enhanced item selection algorithm allows for additional content factors to be included in item selection process, such as Aspects of Rigor for mathematics

[Anonymous]

Here's the concordance table to convert between old and new scores.

PP was right. The off-the-charts 99+%ile scores, from kids who hyper accelerated but didn't learn grade level material thoroughly, get much lower scores on the new test, which poses hard grade-level problems instead of easy higher level "exposure" topics (like complex numbers or radicals)

For sub 99%ile, the differences ae minor.

https://www.nwea.org/uploads/MAP_Growth_Fall-Winter-Spring_Concordance_NWEA.xlsx