First School Quality Working Group Meeting Tomorrow Night
One of the key recommendations by the EAC was the formation of a committee to look at quality in BPS schools going forward. Superintendent Johnson agreed to this and the School Quality Working Group (should we pronounce it squig?) will have its first meeting tomorrow at 6pm at City Hall, room 801. The group will be chaired by Hardin Coleman who is dean of the BU school of Education and was co-chair of the EAC and by Meg Campbell who is a School Committee member and Director of the Codman Academy charter school.
I have been asked to serve on the committee, but I don't know who else (other than the chairs) is on the group. I have been told that, initially, the group will be working on a quality definition and benchmarks.
I'd very much like feedback from readers on how we should be defining and measuring quality in our schools. The EAC came up with what I thought was a pretty good definition of quality, but most of the items in their definition couldn't be measured using available data. What would you add to their list and how would you measure these factors? BPS used their "MCAS snapshot" as a way to measure academic quality during the student assignment reform process. Is this the best way to measure academic performance? Should we try to account for the fact that BPS schools have a wide variety of student populations? If so, how might we do so?
Please post your feedback in the comments.
As much as test scores do not fully reflect the quality of a school I think it is important that a quality measurement rely primarily on hard numbers that are not susceptible to manipulation. With that in mind it seems like MCAS really needs to be the primary input, along with any other surveys etc were data is collected in a complete and thorough manner (I don't know what they may be offhand).
However, if numbers like MCAS are used I think it is really essential that they be indexed to the demographics and background of the student population, i.e. performance by subgroup against the subgroup average. It seems like to me that what you are trying to measure by a quality school is its ability to improve the education of the students there from whatever point they entered that school. With that in mind perhaps scores at the highest grade of the school should also be given more weight then right when kids enter.
People may complain that this is reinforcing low expectations, but in the end of the day if a school is doing a good job with a challenging population that seems like a more impressive and valuable accomplishment then doing an average job with a less challenging population. However, if you just do raw MCAS scores the school that is working with the more challenging population may never be able to be considered a "quality" school, which seems rather unfortunate.
Another positive side benefit of this might be to help reduce the effect of people chasing the schools that have the best test scores, which has a lot to do with who goes there. It seems like it would be really unfortunate if the quality measurement creates a system were people chase schools far away because of raw test scores when there might be other schools closer to them that are actually producing more improvement.
This is along the lines of what I'm thinking. I do worry about doing anything that sets low expectations. One thing I'm interested in is how much each student improves each year. The idea is that no matter where you are at the beginning of the year we expect you to progress a certain amount. In theory, this could even account for some schools having advanced work. There is a measure called Student Growth Percentile that looks at this. However, since it's a percentile, it really looks at performance against everyone else in the state. I think we need an absolute measure.
Have you taken a look at the
Have you taken a look at the state DESE DART (District Analysis and Review Tool)?
It is designed to address similar questions and provides quite a bit of data at the school level as well as for districts (not so relevant for you.) There is also a built in way to compare schools with other "similar" schools, taking into account type of school, total enrollment, the percentage of low-income students enrolled, the percentage of ELL students enrolled, and the percentage of special education students enrolled.
While DART probably won't address all the measures of quality you want, the DESE has gone to a bit of work thinking about these issues, and it would seem like a decent place to start.
Thanks for posting, this is an interesting tool. I think the hard part is to figure out how to compare dissimilar schools.
The bigger picture / New approach to participation
I have to hope quality benchmarks are just one part of the agenda, one working group. They really need to start with the big picture and the process.
It also seems they could do some very different things with this, than with the assignment process. Build on that, but do something far less closed, and far less … old school. The assignment process discussion was inherently contentious, high stakes, zero sum. This isn't.
Putting a part of this process online could reach and involve a lot more people, and have a lot more participation. It could also in effect distribute the work, to look at things at the school level, to provide a level of insight that goes far beyond the broad brush level of analysis done for the assignment process.