I'm Dan and this is my blog. I'm a former high school math teacher and current head of teaching at Desmos. He / him. More here.


  1. anon observer student

    September 22, 2008 - 5:53 pm -

    Do they break down the same data by income/SES? I’m betting if you divide the group in two by that measure, your graph’s gap will in fact be more stark.

    We need to address the big picture.


    I see it already in a first grade class. Kid in the class, bright, already reading well. Single parent with early morning job, so kid’s up very early, tired by afternoon. Anger management problems — when parent is called after problem, swears at kid up and down, intimates that worse will be on the way.

    If he were the only kid in the school or even the grade (or even the class) with these sorts of problems, there might be a way to get him some sort of behavioral/emotional services that would be successful. But there are too many needy, attention craving children for the staff to fill them all up with anywhere near what they need. Instead, I’m sad to say that the path to his being lost is far more clearly marked than the path to success.

  2. … I need to put that same chart together for my school. Yeah, extremely depressing. Wow.

    I’d say that showing data for 7th-11th grade is a bit too wide a spread. Same goes for bundling the entire state into one graph. I mean, it shows interesting stuff, but what does it look like grade by grade, area by area? What if we add other ethnicities to the mix? Better? Worse? Care to place bets?

    Now what? The gap is there. We know this and have known this for quite some time, so this really isn’t news to any of us who have been paying attention. What do we do about it? How do we track down what’s causing this and kill it? I’m racking my brain to figure out what we can do to address this (and ANY) gap.

  3. O.k. — let’s say that just sucks all over. 30% isn’t good for any demo and 10% is horrific. So unless the tests are SO out of whack as to be completely unrealistic in their expectations, let’s all agree that the data sucks from its starting point.

    But I also find it fascinating that there is a common rise between whites and latinos. Both groups are at similar rates. What’s the significance of that? What do we make of the rise in scores? When we are talking about state wide results, is a 5% rise over five years statistically significant? What are the systemic changes, other than a renewed focus on the test itself, that would make one expect to see a rise in test scores? You are talking about a massive sample size… what can we reasonably expect on a statewide level without massive changes to the system itself?

    I’m not saying these numbers are good — they are not. But I also wonder why we expect the scores to rise without major changes in the way we construct our schools. That’s what depresses me — it’s the same question I always come back to — so many good, caring people in such a broken system, and everyone loses, especially the kids.

  4. You’re right, Chris. But in California, we hide behind those incremental gains and suggest things are fine because this year’s group of 8th graders did moderately better than last year’s.

    We don’t look at gaps.

    We don’t track cohort data.

    We dip our toes in the water of reform, but like an old lady, refuse to get our hair wet.


    You gotta look at everyone because there’s a selection bias at play. Algebra I pass rates look fantastic at 7th grade cuz it’s 20,000 of the highest achieving kids in the state. 8th grade slightly worse. 9th grade worse than that. If we just looked at kids who took Algebra I in the 11th grade, it’s low single digits.

  5. The gap is a reflection of institutionalized discrimination. It’s all of us making accommodations for students of poverty and color.

    Staff holding students accountable to higher expectations, eliminating the “pobrecito” attitude and replacing it with a “can do” attitude and addressing each individual student’s need will break the centuries long cycle of institutional racism.

    There are several districts who have figured out how to do it:

    Montgomery County Maryland

    This is a large urban district that is making tremendous progress for African American and Hispanic students.

    Another example – Atlanta Public Schools – 86% of students are African American and they are meeting and exceeding in the 80+% at all grade levels in reading and language arts


    It is being done…

  6. Claire — what is the relationship between the test taken in Maryland and Georgia and the one taken in California? I’ve grown so cynical about the way all the different districts and states measure that I don’t trust the numbers from state to state anymore.

    My modest proposal is this — let’s agree to have everyone take the NAEP (or something similar) in Reading and Math in 4th, 6th, 8th and 11th grades, and let’s call the testing done. We can admit that Social Studies and Science tests can never be skill based enough so they invariably become content tests that determine curriculum, but we can work toward getting a strong sense of skill-driven testing in Reading and Math.

    Then let’s use the test to give us a real sense of what we’re doing to teach the most important skills — literacy and numeracy, and then we can be done with it.

    (And Killian — I hear you, but the devil is, of course, in defining what real reform looks like.)

  7. @ Killian: CT, too, at least in my district. And my mom’s district. And my friends’ districts.

    @ Chris: First good idea w/r/t testing I’ve heard in a while. Granted, I’ve been sticking my fingers in my ears and going “la la la” recently, but still.

  8. Maybe because algebra are not very Hispanic culture friendly?
    I don’t think its because teachers treat a Latino and a White kid differently and resulted that gap. It’s just their culture might not put such an emphasis on academics.
    It’s a cultural thing I believe. If that plot get to show Asian, it’s going to be something different.
    It particularly showed White instead of Asian on that graph because that could encourage the thinking of “institution racism”, if it’s Asian, the graph would lost it’s power. even if Asian people have higher percentages.

  9. Put African Americans on that graph and you’ll see another achievement gap. AA’s score below Latinos for the most part.

    Dan, you might be interested in the book Courageous Conversations About Race by Singleton. Our district is using that as a spring board to actually attempt to close the achievement gap. So far, we’ve come up with the fact that students are painfully aware that the face at the front of the classroom is white, even if we don’t notice it ourselves.

  10. 1% per year improvement is what we see here in PA, too. Despite (or perhaps because of) all the turnover of administration, the focus on review, the siphoning of time and money away from science, art, music, etc., improvement marches on at it’s own steady, dismal pace. Sure there are districts that are doing much better, but the only meaningful number in the long term is the overall progress. Either those few districts that excel are anomalies that are not replicable, or no one’s learning anything in this process.

    As Chris points out, the tests are not comparable from state to state, however, improvement rate should tell you something more comparable than absolute scores. Richard Elmore made this point at an NSF meeting two years ago – look for the schools/districts with the steeper slope, not for those with the better score.

    Mgccl, Latino kids are absolutely treated differently. Steve Kramer at U Md analyzed data and sorted kids by ability and then compared kids of equal propensity – students of color typically took 1-2 fewer math courses in high school that their equally well-prepared white peers, a difference that fully accounted for their lower proficiency. That’s not a cultural factor, it’s an advising and placement issue. This study is not published yet, but I believe has been submitted.

    Dan, I jut discovered this blog today – thanks for sharing!

  11. Mgccl,

    You’re kidding, right? Algebra not culturally “friendly” to Latino kids?

    There IS institutionalized racism. We give less of everything that matters in education — less reasources, less qualified teachers, less need-based instruction, less positive expectations of success — to those who already start off with less, and then explain away the predictable gaps in achievement that result as saying that tests or subject matter or whatever are not culturally friendly? Or that an entire ethnic group devalues education?

    Oh, please.

  12. I agree that Singleton’s book might be a worthwhile professional development conversation starter. My district in Illinois hired his Pacific Education Group to lead our efforts and for two years we have worked hard to make significant changes. There’s a lot we still need to do, but I can see some changes in my teaching staff. I hope that’s only the beginning and will grow even in those who embraced the changes sooner than others.

  13. I agree it’s terrible that one bunch of people are scoring far less than another bunch, when the criteria has nothing to do with ability.

    However, for me the important question here is, *why* the gap? Is it all explained away by income or their parents’ education – or is the gap above and beyond all other factors, in which case the system is institutionally racist?

    In the first case (it relies on income, or parents’ education), then to an extent that’s ‘fair.’ Western society whole-heartedly endorses the idea that your industry can improve life for you *and your children.* If we don’t like that, we’ve got to look further than just the schooling system.

    If it’s above and beyond all other factors, though, then you’re right – it’s unacceptable and we need to move heaven and earth to address it. Otherwise we risk a vicious cycle, where fewer Latino achievers now means fewer Latino role models in the future.

  14. First I think you need to focus on what causes the gap. There are probably several variables which are in play here, many already mentioned (SES, institutionalized racism, ESL/ELL, etc.).

    Then we need to look at whether, as educators and citizens, we can eliminate those variables. Many times they’re a little over our heads, or would require serious societal transformations (We’d all like to root out the (sometimes) subtle systemic racism that exists in our schools, but it’s not going to be a quick & easy fix).

    Since many of these variables (whether for good or ill) are ingrained in our society, the question should be asked, “What extra can we provide for this group that’s facing an uphill struggle?”

    For me, this speaks to the idea of the school as being a community service center in high minority, high poverty areas.

  15. @Chris – In British Columbia we have a provincial test given to grades 4, 7, 10 each year. The results are then sent to each school and to the parents. At the school level the results are used to plan school wide literacy and numeracy yearly goals, classroom instruction etc… But then a wonderful think-tank, The Fraser Institute, gets a hold of these results and prints a list that rates schools against each other in the province which causes a flurry of Union backlash, parental letters to the editors…Beware what you wish for!

    @Ben – our most notorious inner city school is a community service centre in its area and this has seemly proven helpful on a social level, but it hasn’t improved it much academically. Those “extras” you talked about are being put into place here too. Here it is for our Aboriginal students. We have had town-hall style meetings with Aboriginal parents and school board officials, and teachers. The District has hired an Aboriginal Principal and there is talk of an Aboriginal/Non-aboriginal dual-trac school or separate school. We also have Aboriginal curriculum. Too soon to say whether it is helping, but my concern is it may create the “them” and “us” mentality. Not exactly helping institutional racism. What’s the solution? Maybe just the courage to keep trying and adjusting as we work our way along this journey.
    @Dan – I’m glad this data exists too. What if we focused on our little corner of the world and taught our best to our students making the most significant difference possible for them – is it possible bit by bit the achievement gap may close? Ah, maybe I’m still too idealistic. PS: I would have given that hypothetical student the “A” too! You made my entire day over that and my ADHD son thanks you too. Apologies for the length of this comment.

  16. @Dan: Yeah, it is discouraging. I find the way to deal with it is to ignore it when it emotionally bothers me, and just focus on helping the kid in front of me. I do my best to only read huge statistics near the end of the week. As teachers, we can’t get into this mentality that we have to fix all of the world’s problems because that’s what the graph is—a huge set of super–larger than your classroom/school/district problems. That one graph represents a multitude of social/educational issues that cause it. As teachers, we gotta just give the kids we got the best we can deliver. Otherwise, we’d go insane.

  17. All,
    Can we at least agree that the word “data” is a plural noun? “Data are…” not “data is…”
    “These data….” not “this data…”

    We probably won’t agree, but I can try…

  18. Awright, not to protest too much or anything, but I sat there proofreading, pushing publish, fully aware of the error. I just don’t know at what point this becomes one of those “above par / below par” malapropisms where it’s just better to roll with common usage. Apologies, anyway.

  19. @Chris I couldn’t agree with you more on using a national leveled test. Our district is using the ACT series of tests for 8th – 11th grade which helps our students be college ready. Oregon has implemented the PSAT for all 10th graders across the state – again helping students ready themselves for college entry. I believe some students who may never think of themselves as college worthy will be given the opportunity to see that they are.

    While state tests cannot be compared to one another – I still think we should pay attention to these two districts – Montgomery County and Atlanta as they’ve made a difference in many kids’ lives.

    @Dan – thanks for having the courage to bring up a very uncomfortable subject. We need to continue working through the issue until we find the solutions.

  20. I’m currently reading Tony Wagner’s The Global Achievement Gap and his suggestion seems apropos here. Each of these assessment tools is imperfect in its measurement of student preparation for adulthood and productive citizenry. I wonder what would happen if we followed his suggestion and all took the NAEP or something like it and used that as a common baseline for all data. At least it would give us the opportunity to define success in common ways and enable school districts to share successful strategies in a more meaningful way than in the current hodge podge collection of disconnected assessments across state lines.

  21. from Grammarphobia.com:

    “R.I.P. It’s time to admit that data has joined agenda, erotica, insignia, opera, and other technically plural Latin and Greek words that have become thoroughly Anglicized as singular nouns taking singular verbs. No plural form is necessary, and the old singular form, datum, can be left to the Romans. (Media, it seems, is going the same way, though it’s not there yet. Ask again in a few years.)”

  22. The first sentence should have read: The system, public schools, and teachers have little or nothing to do with the *problem* unless they are surrounded by supplementary programs.