High Schools for Equity: Data and background on school selection
Last week we published a critique of a recent publication by Stanford professors Linda Darling-Hammond and Diane Friedlaender. The critique was titled Stanford study's "success" story questioned. A great deal of discussion ensued on the list and in the comments here, including a detailed response from the study's authors.
I am now happy to republish the full text of their response here with their permission. You can find the original version here:
November, 2007
Based on a conference presentation releasing a new report last week, Caroline Grannan has raised a number of questions about the rationale for our selection of June Jordan School for Equity in a study of schools succeeding with African American and Latino students in California. We are sure that her concerns arise from the same motive as our own — a desire to support strong education for children of color in California. However, our data sources and interpretations are different. We used data on achievement that are disaggregated by race and class, and data directly from school records on student transfers and college-going which are more accurate than what seems to be available to Ms. Grannan. For those who are interested, below we provide information that explains why we selected June Jordan for the study and more descriptive information on the school's context.
A central difference in our perspectives is that Ms. Grannan compares data on overall API scores and college-going rates for a number of schools without disaggregating the data by race. We selected our schools in 2005-06 based on their successes in high school graduation, college-going and achievement for low-income students and students of color. We looked at schools in comparison to others in their neighborhoods. For 2005-2006, June Jordan — with a population that was 70% African American and Latino — had an overall API base score of 605, which was lower than that of some nearby schools with very different populations, such as Balboa, with a population that was 61% white and Asian, and Burton (54% white and Asian), but higher than that of other comprehensive high schools serving more similar students in the district, such as Mission High School with an API score of 563 (61% African American and Latino students), John O'Connell High School (81% African American and Latino) with an API score of 553.
More pertinent for our purposes is the fact that June Jordan's API scores for African American and Latino students were considerably higher than those of all of the other surrounding schools, including Balboa and Burton. (See table below).
Table 1: API scores of nearby schools, 2005-2006
June Jordan had substantially higher English Language Arts (ELA) CST scores for African American and Latino students than all of these schools by a considerable margin, often showing proficiency rates 10 to 20 percentage points higher for similar students. (See Table 2.)
Table 2: ELA CST scores proficient or above of nearby schools, 2005-2006
June Jordan also had substantially higher pass rates on the CAHSEE exit exam in 10th grade in ELA than all of these other schools — for all students as well as for African American and Latino students (See Table 3.) JJSE's ELA passage rate of 72% far surpassed the next closest ELA passage rate, which was Balboa's, at a pass rate of 53%. JJSE surpassed all the other schools except for Burton in overall pass rates in mathematics, and surpassed all — including Burton — in math pass rates for African American and Latino students.
Table 3: CAHSEE passage rates of nearby schools, 2006
Our data analysis also examined graduation rates and college-going rates. With access to meticulously coded individual student data, we used the state metric for calculating a graduation rate from the school, which adjusts for transfers and counts dropouts. (The state had not yet done this calculation, because JJSE was just completing its first graduating class.) Using the state calculation method (described in our report), we found that JJSE had a graduation rate of 95%, with two students from the cohort still in school finishing a fifth year of high school.
We did track losses of students that occurred when June Jordan moved locations from its original site on the San Francisco State University campus to its current site, far from many of its initial students' neighborhoods. We also tracked, in our longer case study, the number of students who did not graduate but are still at the school completing credits, expected to graduate next year. We collected data that tracked the transitions of students into and out of the school (including where those who left transferred to and their likelihood of graduation). Finally, we attended to changes in the student composition of the school over time, which can influence a variety of outcomes.
Finally, we documented, with individual student data, the fact that 73% of the school's graduates were admitted to four-year colleges and 95% to 2- or 4-year colleges. Although Ms. Grannon questions whether the June Jordan's college-going rate is noteworthy, this four-year college-going rate is nearly three times the state average, for a school with many more students of color, and its overall college-going rate is comparable to that of Balboa, which has fewer than half as many African American and Latino students.
As these comparisons are made, it is worth noting that JJSE has the highest percentage of African-American students of any high school in San Francisco Unified School District, except for the "second chance" continuation schools. In 06-07, JJSE's student population was 37% African American, ISA's was 35%, and all others were below 20% — Marshall 19%; Mission 18%; Wallenberg 18%; Burton 15%; Balboa 13%; SOTA 12%; O'Connell 11%; Washington 8%; Galileo 7%; Lincoln 7%; and Lowell 3%.
Our study was completed before state test scores were published for 2006-07. However, we are aware that the scores for JJSE went down that year, as its population grew by nearly 50%, from 254 to 371. In that year, the share of African American and Latino students served at JJSE also increased from 70% to 73%. We are also aware that two of the comparison schools — Balboa and O'Connell — lost so many African American students between the two years that they no longer have an African American subgroup whose scores are counted. The lower scores that incoming students brought with them to JJSE in their first year in the school will be a new challenge for the staff. We believe the staff should be evaluated on how they raise student achievement in the years to come, and that all analyses of school progress and accomplishment should be sensitive to these issues of student populations and migration that are essential to interpreting how schools are serving students.
Respectfully,
Linda Darling-Hammond and Diane Friedlaender,
School Redesign Network at Stanford University
I am now happy to republish the full text of their response here with their permission. You can find the original version here:
High Schools for Equity: Data and background on school selection
November, 2007
Based on a conference presentation releasing a new report last week, Caroline Grannan has raised a number of questions about the rationale for our selection of June Jordan School for Equity in a study of schools succeeding with African American and Latino students in California. We are sure that her concerns arise from the same motive as our own — a desire to support strong education for children of color in California. However, our data sources and interpretations are different. We used data on achievement that are disaggregated by race and class, and data directly from school records on student transfers and college-going which are more accurate than what seems to be available to Ms. Grannan. For those who are interested, below we provide information that explains why we selected June Jordan for the study and more descriptive information on the school's context.
A central difference in our perspectives is that Ms. Grannan compares data on overall API scores and college-going rates for a number of schools without disaggregating the data by race. We selected our schools in 2005-06 based on their successes in high school graduation, college-going and achievement for low-income students and students of color. We looked at schools in comparison to others in their neighborhoods. For 2005-2006, June Jordan — with a population that was 70% African American and Latino — had an overall API base score of 605, which was lower than that of some nearby schools with very different populations, such as Balboa, with a population that was 61% white and Asian, and Burton (54% white and Asian), but higher than that of other comprehensive high schools serving more similar students in the district, such as Mission High School with an API score of 563 (61% African American and Latino students), John O'Connell High School (81% African American and Latino) with an API score of 553.
More pertinent for our purposes is the fact that June Jordan's API scores for African American and Latino students were considerably higher than those of all of the other surrounding schools, including Balboa and Burton. (See table below).
Table 1: API scores of nearby schools, 2005-2006
JJSE | Balboa | Mission | O'Connell | Burton | |
| API Base Scores | 605 | 672 | 553 | 553 | 663 |
| API African Am. | 525 | 479 | 448 | n/a | 476 |
| API Latino | 611 | 575 | 513 | 549 | 569 |
June Jordan had substantially higher English Language Arts (ELA) CST scores for African American and Latino students than all of these schools by a considerable margin, often showing proficiency rates 10 to 20 percentage points higher for similar students. (See Table 2.)
Table 2: ELA CST scores proficient or above of nearby schools, 2005-2006
JJSE | Balboa | Mission | O'Connell | Burton | |
All Students | |||||
| CST ELA 9th Gr | 32% | 47% | 19% | 21% | 39% |
| CST ELA 11th Gr. | 29% | 33% | 17% | 14% | 31% |
Af. Am. Students | |||||
| CST ELA 9th Gr. | 22% | 15% | 6% | 10% | 6% |
| CST ELA 11th Gr | 23% | 20% | 17% | 5% | 14% |
Latino Students | |||||
| CST ELA 9th Gr. | 46% | 27% | 13% | 19% | 25% |
| CST ELA 11th Gr. | 26% | 20% | 13% | 14% | 13% |
June Jordan also had substantially higher pass rates on the CAHSEE exit exam in 10th grade in ELA than all of these other schools — for all students as well as for African American and Latino students (See Table 3.) JJSE's ELA passage rate of 72% far surpassed the next closest ELA passage rate, which was Balboa's, at a pass rate of 53%. JJSE surpassed all the other schools except for Burton in overall pass rates in mathematics, and surpassed all — including Burton — in math pass rates for African American and Latino students.
Table 3: CAHSEE passage rates of nearby schools, 2006
JJSE | Balboa | Mission | O'Connell | Burton | |
All Students | |||||
| ELA Passed | 72% | 53% | 44% | 43% | 51% |
| Math Passed | 58% | 56% | 39% | 39% | 67% |
Af. Am. Students | |||||
| ELA Passed | 69% | 41% | 40% | 37% | 34% |
| Math Passed | 44% | 30% | 16% | 28% | 38% |
Latino Students | |||||
| ELA Passed | 69% | 39% | 36% | 41% | 55% |
| Math Passed | 57% | 41% | 31% | 39% | 50% |
Our data analysis also examined graduation rates and college-going rates. With access to meticulously coded individual student data, we used the state metric for calculating a graduation rate from the school, which adjusts for transfers and counts dropouts. (The state had not yet done this calculation, because JJSE was just completing its first graduating class.) Using the state calculation method (described in our report), we found that JJSE had a graduation rate of 95%, with two students from the cohort still in school finishing a fifth year of high school.
We did track losses of students that occurred when June Jordan moved locations from its original site on the San Francisco State University campus to its current site, far from many of its initial students' neighborhoods. We also tracked, in our longer case study, the number of students who did not graduate but are still at the school completing credits, expected to graduate next year. We collected data that tracked the transitions of students into and out of the school (including where those who left transferred to and their likelihood of graduation). Finally, we attended to changes in the student composition of the school over time, which can influence a variety of outcomes.
Finally, we documented, with individual student data, the fact that 73% of the school's graduates were admitted to four-year colleges and 95% to 2- or 4-year colleges. Although Ms. Grannon questions whether the June Jordan's college-going rate is noteworthy, this four-year college-going rate is nearly three times the state average, for a school with many more students of color, and its overall college-going rate is comparable to that of Balboa, which has fewer than half as many African American and Latino students.
As these comparisons are made, it is worth noting that JJSE has the highest percentage of African-American students of any high school in San Francisco Unified School District, except for the "second chance" continuation schools. In 06-07, JJSE's student population was 37% African American, ISA's was 35%, and all others were below 20% — Marshall 19%; Mission 18%; Wallenberg 18%; Burton 15%; Balboa 13%; SOTA 12%; O'Connell 11%; Washington 8%; Galileo 7%; Lincoln 7%; and Lowell 3%.
Our study was completed before state test scores were published for 2006-07. However, we are aware that the scores for JJSE went down that year, as its population grew by nearly 50%, from 254 to 371. In that year, the share of African American and Latino students served at JJSE also increased from 70% to 73%. We are also aware that two of the comparison schools — Balboa and O'Connell — lost so many African American students between the two years that they no longer have an African American subgroup whose scores are counted. The lower scores that incoming students brought with them to JJSE in their first year in the school will be a new challenge for the staff. We believe the staff should be evaluated on how they raise student achievement in the years to come, and that all analyses of school progress and accomplishment should be sensitive to these issues of student populations and migration that are essential to interpreting how schools are serving students.
Respectfully,
Linda Darling-Hammond and Diane Friedlaender,
School Redesign Network at Stanford University
Labels: Charters
3 Comments:
Odd that Linda Darling Hammond never mentioned the most obvious reason why she would include June Jordan in her study - that is, that the school's co-founder Matt Alexander had previously been one of her interns at Stanford, and even co-wrote at least one other study of small schools with her. Here is the link to it.
Redesigning Schools: What Matters and What Works. 10 Features of Effective Small Schools. By Linda Darling-Hammond with Matt Alexander and Donielle Prince
http://www.srnleads.org/resources/publications/10_features.html
You can even see a photo of Matt Alexander on the cover. Given that Linda Darling Hammond already knew he subscribed to all of her beliefs, and that his school was founded on the principles put forth in the study they co-wrote, it would have been more surprising if she had NOT included June Jordan in the current study.
But surely any study which is supposed to be unbiased should have at least mentioned the author's previous relationship with the co-founder of one of the schools being studied?
A poster on the sfschools listserve -- apparently a June Jordan staffer or parent -- posted this comment in defense of the study:
"LDH [Linda Darling-Hammond] has been and continues to be a proponent of, advocate and spokesperson for SSD (Small Schools by Design). Her study would, therefore, reflect this, wouldn't it?"
This raises the obvious question of whether the study was PR or impartial academic research. If it's PR, fine, but it needs to be labeled as such. PR masquerading an impartial study is dishonest.
Although the posters raise some worthwhile issues to consider regarding full disclosure, it still feels like they miss the point of the school's promising achievements. That is, JJSE seems to be doing what none of the comprehensive high schools have been able to do: make progress toward closing the achievement gap when one looks at the disaggregated data by race. Students of color, students who live in poverty and students who are learning English are the least well-served state-wide and nationally. So, while I would also want to see better results, I am intrigued by the model that by some measures is working for students for whom schooling, as currently constituted, has been a failure. At a time when more young men of color are more likely to drop out than earn a high school diploma, these results should be investigated, replicated if they prove successful and these results should be more widely and positively reported, IMHO.
Other sites and articles of interest:
The Sentencing Project
Tackling the Black, Brown Male Crisis
Post a Comment
<< Home