We had statisticians at work who said we could not ask a grantee to explain racial differences if the number for the group was less than 30. That was a rule of thumb. I noticed, however, that our statisticians did not always follow their own rules.
I don’t know about Compton and Inglewood but in SF teachers are prevented or discouraged from disciplining students. Called racists if they do. Part of the popularity of charters in the Black community is that charters will suspend or expel students.
I found most of this article very interesting and helpful but you loose me when you speak about genetic variants and education. There is no logic in saying that a Latino couple who grew up in poverty and had no access to high school let alone college, no parents or other educated figures in their lives, didn’t achieve many educational years due to their lack of good genetic variants and so didn’t pass them on to their kids. Logic suggests other lacks were more important in their lack of education. Nor do we have as clear an understanding of genes and their roles as your imply. It’s also puzzling that you leave out even the possibility of racism. It’s one thing to say there are these other plausible possibilities etc. (helpful!) and another to cover various possibilities without once mentioning that studies find that we all still hold stereotypes and assumptions and it’s inevitable that many teachers and school staff might hold strong negative views towards certain races.
Hi Yanina, I encourage you to read "The Genetic Lottery" by Kathryn Paige Harden, which was my source for that paragraph. It examines these issues in far more detail than I could ever manage and changed my thinking in several ways. Eliminating the structural/environmental/cultural issues that are holding some students back is a proper goal of policy but she pointed out that this will increase the salience of genetic issues, not reduce them. If all plants don't get enough sunlight, none of them will thrive. But if they do get enough sunlight, the ones with the right genes will grow much taller than the others. When women were prevented from pursuing post-secondary education, the link between a mother's years of education and her children's was much weaker than it is today. The heritability of 'years of education' has increased because the environmental constraints have been weakened.
There is one data point that shows how well a school is doing educating it's students (and possibly this data can also be used to compare districts.) Core Growth Data (see link below) compares how much students grow on SBAC from year to year, while controlling for student demographics (such as ESL and SED as referenced in your article but unfortunately not parent education level.) Before the pandemic, many large CA districts participated including Oakland (my district,) SFUSD and Los Angeles.
Oakland made their Core Growth data public (see link below.) It showed some interesting and unexpected results. For example, my kids school (low-ESL, low-SED) showed high proficiency rates on SBAC, but low Core Growth. This means that compared to other non-ESL, non-SED students, our student grew less each year. Additionally, you can find high-ESL, high-SED schools showed low proficiency rates on SBAC, but high Core Growth. This school was able to move their students further in one year (and in some cases year after year) than schools serving similar populations.
I'm not sure if SFUSD still participates in Core Districts, and I don't know if they ever made their data public, but it would be interesting to see. (Let me know if you want to talk further about this data.)
Hi Josh, I've never seen the core growth data and would be happy to talk about it to learn more. You may or may not be aware that CDE is planning to release its own growth data later this year. I'm eagerly looking forward to the data release for precisely the reasons you mention.
Yes. I am excited to see that the state will be publishing Growth Data alongside Proficiency Data. I think it is important to view both when thinking about school quality or school accountability.
The CDE Growth Data will not control for student demographics, like Core Growth Data does. Considering your arguments in the article, I would assume you would be in favor of considering student demographic information when analyzing and comparing assessment data.
Let chat about it sometime. I am curious if you will be able to dig up the SFUSD Core Growth Data. Or, if you will be able to compare Core Growth Data to upcoming CDE data to see how those two metrics differ. Email me and we can set up a time to chat: josh.lee.cline@gmail.com
I agree with cheesemonkey. Add to that disciplinary polices that allow disruptive students to remain in the classroom. If you remove a smart Black child from bad classroom learning environment they would perform at much higher levels. Do Black students that addend schools with more White students perform better?
There are districts where White and Asian students are less than 1% of the entire student population where Black students do much better than they do in San Francisco. Look at Compton Unified (14% Black) and Inglewood Unified (28% Black) in Los Angeles. Yes, peer effects matter and classroom discipline does too, but I find it hard to believe that those factors are better for Black students in Compton and Inglewood than in SF. Chronic absenteeism is, however, lower in those districts than in SF, which is evidence in favor of chessmonkey's point.
Thanks for another great analysis, Paul. One factor you didn't mention that I've been investigating is the role of chronic absenteeism. I'm wondering if there could also be some correlations between proficiency levels and chronic absenteeism. I wonder what the data show about this.
You're right. I did some posts on chronic absenteeism more than a year ago and I'm pretty sure I looked at this at the time. Do a search and you should be able to find the posts.
One issue is that it's hard to untangle cause and effect. Being chronically absent will undoubtedly hurt a child's education. I would even bet that high chronic absenteeism rates in a class will harm children in the class who are not chronically absent because the teacher will spend so much time trying to bring the absentees up to speed when they return. On the other hand, one major reason kids miss school is that they are not doing well. Is the absenteeism a symptom or a cause? If the absenteeism is a symptom then one could argue that it is the district's fault, at least in part, and I was trying to keep this post focused on things that are clearly out a district's control.
We had statisticians at work who said we could not ask a grantee to explain racial differences if the number for the group was less than 30. That was a rule of thumb. I noticed, however, that our statisticians did not always follow their own rules.
I don’t know about Compton and Inglewood but in SF teachers are prevented or discouraged from disciplining students. Called racists if they do. Part of the popularity of charters in the Black community is that charters will suspend or expel students.
I found most of this article very interesting and helpful but you loose me when you speak about genetic variants and education. There is no logic in saying that a Latino couple who grew up in poverty and had no access to high school let alone college, no parents or other educated figures in their lives, didn’t achieve many educational years due to their lack of good genetic variants and so didn’t pass them on to their kids. Logic suggests other lacks were more important in their lack of education. Nor do we have as clear an understanding of genes and their roles as your imply. It’s also puzzling that you leave out even the possibility of racism. It’s one thing to say there are these other plausible possibilities etc. (helpful!) and another to cover various possibilities without once mentioning that studies find that we all still hold stereotypes and assumptions and it’s inevitable that many teachers and school staff might hold strong negative views towards certain races.
Hi Yanina, I encourage you to read "The Genetic Lottery" by Kathryn Paige Harden, which was my source for that paragraph. It examines these issues in far more detail than I could ever manage and changed my thinking in several ways. Eliminating the structural/environmental/cultural issues that are holding some students back is a proper goal of policy but she pointed out that this will increase the salience of genetic issues, not reduce them. If all plants don't get enough sunlight, none of them will thrive. But if they do get enough sunlight, the ones with the right genes will grow much taller than the others. When women were prevented from pursuing post-secondary education, the link between a mother's years of education and her children's was much weaker than it is today. The heritability of 'years of education' has increased because the environmental constraints have been weakened.
There is one data point that shows how well a school is doing educating it's students (and possibly this data can also be used to compare districts.) Core Growth Data (see link below) compares how much students grow on SBAC from year to year, while controlling for student demographics (such as ESL and SED as referenced in your article but unfortunately not parent education level.) Before the pandemic, many large CA districts participated including Oakland (my district,) SFUSD and Los Angeles.
Oakland made their Core Growth data public (see link below.) It showed some interesting and unexpected results. For example, my kids school (low-ESL, low-SED) showed high proficiency rates on SBAC, but low Core Growth. This means that compared to other non-ESL, non-SED students, our student grew less each year. Additionally, you can find high-ESL, high-SED schools showed low proficiency rates on SBAC, but high Core Growth. This school was able to move their students further in one year (and in some cases year after year) than schools serving similar populations.
I'm not sure if SFUSD still participates in Core Districts, and I don't know if they ever made their data public, but it would be interesting to see. (Let me know if you want to talk further about this data.)
https://edsource.org/2019/how-california-measures-academic-success-is-changing-at-some-of-the-states-largest-districts/620361
https://dashboards.ousd.org/views/COREGrowth_15676333862790/Summary?iframeSizedToWindow=true&%3Aembed=y&%3AshowAppBanner=false&%3Adisplay_count=no&%3AshowVizHome=no&%3Aorigin=viz_share_link
Hi Josh, I've never seen the core growth data and would be happy to talk about it to learn more. You may or may not be aware that CDE is planning to release its own growth data later this year. I'm eagerly looking forward to the data release for precisely the reasons you mention.
Yes. I am excited to see that the state will be publishing Growth Data alongside Proficiency Data. I think it is important to view both when thinking about school quality or school accountability.
The CDE Growth Data will not control for student demographics, like Core Growth Data does. Considering your arguments in the article, I would assume you would be in favor of considering student demographic information when analyzing and comparing assessment data.
Let chat about it sometime. I am curious if you will be able to dig up the SFUSD Core Growth Data. Or, if you will be able to compare Core Growth Data to upcoming CDE data to see how those two metrics differ. Email me and we can set up a time to chat: josh.lee.cline@gmail.com
I agree with cheesemonkey. Add to that disciplinary polices that allow disruptive students to remain in the classroom. If you remove a smart Black child from bad classroom learning environment they would perform at much higher levels. Do Black students that addend schools with more White students perform better?
There are districts where White and Asian students are less than 1% of the entire student population where Black students do much better than they do in San Francisco. Look at Compton Unified (14% Black) and Inglewood Unified (28% Black) in Los Angeles. Yes, peer effects matter and classroom discipline does too, but I find it hard to believe that those factors are better for Black students in Compton and Inglewood than in SF. Chronic absenteeism is, however, lower in those districts than in SF, which is evidence in favor of chessmonkey's point.
Thanks for another great analysis, Paul. One factor you didn't mention that I've been investigating is the role of chronic absenteeism. I'm wondering if there could also be some correlations between proficiency levels and chronic absenteeism. I wonder what the data show about this.
You're right. I did some posts on chronic absenteeism more than a year ago and I'm pretty sure I looked at this at the time. Do a search and you should be able to find the posts.
One issue is that it's hard to untangle cause and effect. Being chronically absent will undoubtedly hurt a child's education. I would even bet that high chronic absenteeism rates in a class will harm children in the class who are not chronically absent because the teacher will spend so much time trying to bring the absentees up to speed when they return. On the other hand, one major reason kids miss school is that they are not doing well. Is the absenteeism a symptom or a cause? If the absenteeism is a symptom then one could argue that it is the district's fault, at least in part, and I was trying to keep this post focused on things that are clearly out a district's control.