PRESS RELEASE
For Immediate Release | May 22, 2018
Middle School Selection is an incredibly important decision for children and families. My own family went through this process at the end of the 2016-2017 school year, and it is fresh in my mind. There are many reasons for preference of one middle school or the other. The purpose of this statement is not to address the reasons for having a preference of one school or the other, but rather to explain how we must proceed in light of the inability to balance enrollment with a choice approach. I believe that when the School Committee voted to use what at the time was called the ‘partial lottery’ that this methodology is precisely what was intended and certainly what I had meant when I voted to proceed in this fashion.
Choice of words matters and for this reason I have stopped using the word lottery. Lottery by definition has winners and losers. While it could certainly be said that this word is still appropriate when you consider that students are either winning by getting their choice of middle school, or losing, by not getting their choice, it is hard to not also see the win/lose paradigm being applied to the schools themselves, thus my change in language here to randomization.
Our middle schools are great, and equally great. Like every human endeavor there is variation and of course some variation is good and some is not. If either of our middle schools has a shortcoming then we should work to identify it and fix or improve that shortcoming. I know of no reliable assessment that shows either of our middle schools are better than the other on a macro scale.
Efforts to understand how to be fair required clear statements of principle and goal:
Principle: All elementary schools shall send half of their rising 5th graders to each middle school.
Goal: Ensure transparency and improve confidence that selection has been entirely fair.
This is a technical document. It is technical because I believe we cannot be transparent without providing the details of what the numbers mean and how we have arrived at them.
Thank you to the McGlynn Elementary School rising 5th graders and their families for reducing the scope of this challenge - without their over-selection of the McGlynn Middle the number of students requiring randomization to a non-selected school would be much greater.
Respectfully,
Paul Ruseau
Member, Medford School Committee
2018-2019 Middle School Placement Randomization Proposal
|
# rising 5th graders |
% of all rising 5th graders |
# needed at Andrews Middle |
# selected Andrews Middle |
# to change to McGlynn Middle |
# selected McGlynn Middle |
# ideal to each middle school 2 |
|
|
Brooks |
65 |
22% |
41 |
59 |
18 |
6 |
33 |
|
Columbus |
58 |
20% |
37 |
37 |
20 |
29 |
|
|
Roberts |
83 |
28% |
53 |
57 |
4 |
26 |
42 |
|
McGlynn 1 |
88 |
30% |
16 |
16 |
72 |
44 |
|
|
total rising 5th graders |
294 |
147 |
169 |
22 |
124 |
147 |
1 Due to over-selection of the McGlynn Middle by McGlynn Elementary students we can honor this over-selection as it reduces the number of students from the other schools that will be sent to a non-selected school.
2 The number of students to go to each middle school as a percentage of rising 5th graders in the event McGlynn Elementary students had not over-selected McGlynn Middle.
NOTE: Randomization will exclude students that are not eligible. A student is not eligible if he/she will have a sibling at the Andrews Middle, or have programmatic placement requirements that are not available equally at both middle schools.
Formulas used above:
% of all rising 5th graders = (Number of rising 5th graders at a school) / (total rising 5th graders).
# needed at Andrews Middle = (% of rising 5th graders at a school) * ((total rising 5th graders / 2) - (McGlynn Elementary rising 5th graders going to Andrews Middle)) / 70%. The 70% is an adjustment for the 30% of rising 5th graders from the McGlynn Elementary because the McGlynn Elementary has over-selected for the McGlynn Middle. The McGlynn Elementary students have taken 49% of the McGlynn Middle spots reducing the need for greater assignment from the other schools to a non-selected school.
# ideal to to each middle school = (% of all rising 5th graders at a school) * (total rising 5th graders) / 2.
Recommendations for how to proceed in 2018-2019
Three of our elementary schools will require a randomized placement to determine which students that have selected Andrews Middle School will be placed at McGlynn Middle School. It is recommended the following steps be taken, in order, to ensure fairness and transparency: 1) A randomized placement will take place. This process must be publicly observable and use a method that is trusted. There are numerous methods for selecting from a list of values, such as https://www.random.org/lists/.
I. One school at a time, in no particular order, the eligible student IDs will be placed into the list for each sending elementary school, Randomize will be pressed once and the values will be randomized. These lists will be saved off. All sending schools with eligible students should have this completed before continuing to step II.
-
The school with the largest number of students that must be randomized to McGlynn Middle should have their students selected until they reach the number of the next highest school. Example: Brooks needs to send 18 students. The top 14 students from their list are selected as going to the McGlynn Middle. Then the next student is selected from the Roberts. Then the next student is selected from the Brooks. Then the next student is selected from the Roberts, and so on until the number of students that require reassignment has been reached. In reality this means 15 additional students from the Brooks will go to the McGlynn Middle, and 0 additional student from the Roberts will go to the McGlynn Middle, and 0 additional students from the Columbus will be moved to the McGlynn Middle. This is important - all of the schools are going through the randomization - but in practice, since the other schools have sent sufficient students based on the proportionality of their rising 5th graders it is only the Brooks that will be sending additional students to the McGlynn Middle.
-
This process shall be recorded on video, and members of the public shall be able to observe. If student IDs are not allowed under privacy laws, then a mapping table shall be created and saved for reference and inspection - this table must be created and attested to by at least one School Committee member prior to the randomization being performed. This entire process shall be performed in one session of time with all observers present (even though the public will not be able to view the mapping table).
2) When letters of school assignment are sent out the families receiving the notices that have been selected into a non-selected school should include a letter explaining this process used, where they may view or request viewing of the randomization process, and an acknowledgement that this process is not ideal and the School Committee will be working with the new Superintendent to find ways to prevent such unfavorable situations in the future.