Curtin: four meeting timeline on discussion of accountability
last month, this month, either February/March for regulations to go out, either May/June to vote regulations
weighting of indicators
"this can get a little in the weeds"
"trying to keep this at an appropriate level for today"
need to decide:
- weighting in school percentile calculation
- weighting of "meeting targets"
- weighting "all students" vs. "lowest performing" students
first change: dropping grade 9 passing as an accountability indicator
"specifically talking about the indicators on the left, and not the individual measures within each indicator"
the measures within each indicator will be weighted equally (for example: English, math, science weighted equally within achievement)
to Peyser Q: are weighed equally, not by number of test takers
Peyser: that cohort of students is disproportionately included in achievement
West: science is not in the student growth percentile
Curtin: current measure of achievement to growth is 3 to 1
in current system in elementary, achievement + growth = 100%
ratio between achievement and growth will no longer add up to 100%
increasing weight of growth in system increases the differentiation of schools whose achievement is the same
want to be sure we're getting schools that are both low achieving and low growing
HOWEVER, "a couple of cautions about increasing the value the use of growth in the system"
growth is a normative measure: takes a group of kids who have like test score histories, and looks at how they did on the next test
"if everybody did great on that test the next year, somebody's going to be in the 99th percentile and somebody's going to be in the 1st percentile"
"some might say that 'the school percentile is a normative measure and you have no problem doing that!'"
but it looks at all of the measures together
"You want to be careful on norming a norm"
increasing the value of growth decreases the value of the third grade assessment, because third grade has no growth
increasing value of growth decreases value of science as science has no growth
Craven asks for review how vote in June 2014 changed weighting
Curtin: vote in June 2014 weighted achievement v growth 3 to 1
"is what's used in that charter cap calculation, also"
have that conversation again because it also ties into the charter cap calculation
Wulfson: use of growth has increased over time: was simple achievement, then 4 to 1, now 3 to 1
"trying to find a good balance"
"each of these indicators, each of these weights represents a different value"
Craven: impact on urbans; "Because Brockton had experienced a lot of growth, we were unable to open a charter school there one year"
West: "I see a strong argument for bringing growth into a school accountability system"
over time
national time to incorporate growth over time
"most states are going more aggressively in that direction" even half and half
Peyser: can't answer how this plays out over time
"I don't want to make it all about a smell test and where we think schools and districts ought to be" but if it's consistent
growth to standard approach
Curtin: we'd be having a different conversation if we were talking about a growth to standard approach
weighting needs to be flexible as not all schools have an EL subgroup
more measures in high schools; actual weighting of achievement and growth is going to count more in elementary as there are more indicators for high school
recommendation is to maintain 3 to 1 weighting
Note also newess of measure; we might be coming back to you and suggesting an alternative ratio
Have made an assumption that no indicator should be less than 10%
had some discussion of if 20% is enough for graduation rate; there is an additional required identification of low graduation rates under ESSA
"this does put achievement below 50% in the calculation" for high school
Curtin: did a lot of extensive work on validity work of SGP on 10th grade measure
graduation rate being included as an achievement measure more than a growth measure
Curtin: when we talked about setting targets for schools based on where they're starting
more points if you succeed a target?
It's hard to do this without a second year of assessment data
equal intervals between categories of performance against meeting targets
Declined: 0
No change: 1
Improved: 2
Met target: 3
Exceeded target: 4
and they just lost me in talking about "having room"...
West: having a bit of trouble thinking about the equal interval measure, as I don't know how we're setting targets for these schools
"not only relative, but a school-specific school accountability system"
"not going to be useful for parents" if schools are meeting targets
Curtin: how the improvers did
"not just looking at statewide averages...of those that improved, what if we looked at what the 50th percentile did"
"providing that level of expectation on the new assessment to our schools"
West: might not be a good overall quality measure
Curtin: where the strengths of the school lie
going to have a designation of where the school's strengths lie
provides information how the schools are doing at meeting those targets
"not looking at a statewide average; only looking at those who improved"
Stewart: quite an exercise
trying to get an override, but system isn't complete
Curtin: report card "that will include many more indicators"
"paint a more complete picture to accompany a school accountability measure"
Wulfson: "the complexity is deliberate, because the field has asked for more than how your kids did on MCAS"
Curtin: not only how the school is doing for meeting targets for all students, but also lowest performing students, controlling for transience, for students who have been there for more than a year
proposing to do a 50/50 split
"deliberate: we really want schools to focus on the lowest performing students"
Wulfson: continuing to have feedback with our stakeholder groups
and possibly back next month
"specifically talking about the indicators on the left, and not the individual measures within each indicator"
the measures within each indicator will be weighted equally (for example: English, math, science weighted equally within achievement)
to Peyser Q: are weighed equally, not by number of test takers
Peyser: that cohort of students is disproportionately included in achievement
West: science is not in the student growth percentile
Curtin: current measure of achievement to growth is 3 to 1
in current system in elementary, achievement + growth = 100%
ratio between achievement and growth will no longer add up to 100%
increasing weight of growth in system increases the differentiation of schools whose achievement is the same
want to be sure we're getting schools that are both low achieving and low growing
HOWEVER, "a couple of cautions about increasing the value the use of growth in the system"
growth is a normative measure: takes a group of kids who have like test score histories, and looks at how they did on the next test
"if everybody did great on that test the next year, somebody's going to be in the 99th percentile and somebody's going to be in the 1st percentile"
"some might say that 'the school percentile is a normative measure and you have no problem doing that!'"
but it looks at all of the measures together
"You want to be careful on norming a norm"
increasing the value of growth decreases the value of the third grade assessment, because third grade has no growth
increasing value of growth decreases value of science as science has no growth
Craven asks for review how vote in June 2014 changed weighting
Curtin: vote in June 2014 weighted achievement v growth 3 to 1
"is what's used in that charter cap calculation, also"
have that conversation again because it also ties into the charter cap calculation
Wulfson: use of growth has increased over time: was simple achievement, then 4 to 1, now 3 to 1
"trying to find a good balance"
"each of these indicators, each of these weights represents a different value"
Craven: impact on urbans; "Because Brockton had experienced a lot of growth, we were unable to open a charter school there one year"
West: "I see a strong argument for bringing growth into a school accountability system"
over time
national time to incorporate growth over time
"most states are going more aggressively in that direction" even half and half
Peyser: can't answer how this plays out over time
"I don't want to make it all about a smell test and where we think schools and districts ought to be" but if it's consistent
growth to standard approach
Curtin: we'd be having a different conversation if we were talking about a growth to standard approach
weighting needs to be flexible as not all schools have an EL subgroup
more measures in high schools; actual weighting of achievement and growth is going to count more in elementary as there are more indicators for high school
recommendation is to maintain 3 to 1 weighting
Note also newess of measure; we might be coming back to you and suggesting an alternative ratio
Have made an assumption that no indicator should be less than 10%
So that looks like this #MAEdu pic.twitter.com/P8mQ9avR1A— Tracy O'Connell Novick (@TracyNovick) January 23, 2018
I'll stick nicer versions of them in laterAnd like this:#MAEdu pic.twitter.com/XkDzZ4sYVS— Tracy O'Connell Novick (@TracyNovick) January 23, 2018
had some discussion of if 20% is enough for graduation rate; there is an additional required identification of low graduation rates under ESSA
"this does put achievement below 50% in the calculation" for high school
Curtin: did a lot of extensive work on validity work of SGP on 10th grade measure
graduation rate being included as an achievement measure more than a growth measure
Curtin: when we talked about setting targets for schools based on where they're starting
more points if you succeed a target?
It's hard to do this without a second year of assessment data
equal intervals between categories of performance against meeting targets
Declined: 0
No change: 1
Improved: 2
Met target: 3
Exceeded target: 4
and they just lost me in talking about "having room"...
West: having a bit of trouble thinking about the equal interval measure, as I don't know how we're setting targets for these schools
"not only relative, but a school-specific school accountability system"
"not going to be useful for parents" if schools are meeting targets
Curtin: how the improvers did
"not just looking at statewide averages...of those that improved, what if we looked at what the 50th percentile did"
"providing that level of expectation on the new assessment to our schools"
West: might not be a good overall quality measure
Curtin: where the strengths of the school lie
going to have a designation of where the school's strengths lie
provides information how the schools are doing at meeting those targets
"not looking at a statewide average; only looking at those who improved"
Stewart: quite an exercise
trying to get an override, but system isn't complete
Curtin: report card "that will include many more indicators"
"paint a more complete picture to accompany a school accountability measure"
Wulfson: "the complexity is deliberate, because the field has asked for more than how your kids did on MCAS"
Curtin: not only how the school is doing for meeting targets for all students, but also lowest performing students, controlling for transience, for students who have been there for more than a year
proposing to do a 50/50 split
"deliberate: we really want schools to focus on the lowest performing students"
Wulfson: continuing to have feedback with our stakeholder groups
and possibly back next month
No comments:
Post a Comment