Bacon's Rebellion

Democracy Thrives in Sunlight

Using Big Data to Bolster Student Performance

—

by

James A. Bacon
in ,

Principal Tina McCay with students at Goochland Elementary.
Principal Tina McCay with students at Goochland Elementary.

At the beginning of the 2013-2014 school year, 4th grade teachers at Goochland Elementary School were setting academic goals for their students, with an eye to their performance in the Standards of Learning (SOL) exams. With support from the Virginia Department of Education (VDOE) and an innovative data system implemented in Goochland, the school had been provided historical data and predictive analysis of their new students’ standing. Teachers were alarmed to find that nine students were on a predicted trajectory to fail their reading SOLs.

In a 4th-grade class of 55 children, failure by nine students constituted an unacceptably high percentage.  “We refused to accept the outcome as inevitable,” said Principal Tina McCay. “The data gave us a peek into one possible future, but it also gave our team time to develop a strategy to reverse the trend and set our students on a solid path to achievement and success.”

Teachers decided that the targeted students would benefit most by the participation in small instructional groups. “As a result of the small groups, students who previously were crying from frustration suddenly became engaged and confident,” said teacher Krystle Demas. “It was exciting to witness. Just to see that spark in their eyes and a return of the excitement and passion for learning was so rewarding!”

The result: All nine students passed the SOL at the end of the school year. Said McCay, “This extraordinary success might never have happened without real time access to data at each step of the process.”

Over the past decade, the education data industry has churned out new tools for schools and teachers to analyze data and see trends that would have been overlooked in the past.  “Data can be used to help educators tailor curricula, identify at-risk students, customize classroom learning and improve their students’ college readiness,” said Bethann Canada, director of the Office of Educational Information Management at VDOE.

Building on Goochland’s success, VDOE is partnering with the state’s Center for Innovative Technology (CIT) with the goal of transforming itself from a static, one-way collector of data into a provider of support and service for data-driven decision making across the state.

In an elaborate consultation process, the data implementation team engaged more than 400 individuals across Virginia in focus group activities. Ninety-seven percent of Virginia’s school divisions took part. The end product was a strategy with two components: (1) Technology and Integration and (2) Professional Development and Division Support. Providing the data would not be enough. Educators needed to know what to do with it.

“It’s more than just having access to the data that’s important, it’s knowing what to do with it,” said Paul McGowan, vice president of consulting services at CIT. “Or, as some focus group participants explained, “Even if we’re able to run reports, a lot of teachers say, ‘now what?’ Many of us don’t know what to do with the data once we have it.”

Following up on the successful first phase of the project, the project team expects to roll out a similar capability statewide in the next year. Phase II is focusing on implementing the technology and integration solution and building a new Education Data Professional Development Center.

“We hope to attract, persuade and retain support for data use and to persuade all K-12 stakeholders to include data as an integral component of their work and educational plans and intervention strategies,” said Canada. “Generating a viable solution will take time and hard work, but will bring numerous dividends in the form of customized learning, stronger curricula, identifying and aiding at-risk students, and much more.”

(This is a condensed version of an article released by the Goochland Public Schools, Virginia Department of Education and the Center for Innovative Technology.)

Share this article

ADVERTISEMENT

(comments below)

ADVERTISEMENT

(comments below)

Comments

6 responses to “Using Big Data to Bolster Student Performance”

  1. larryg Avatar
    larryg

    The big question is what is the data they are talking about and where did it come from?

    my suspicions are it comes from testing and assessment of the students themselves and to me that proves just how important testing really is when we seem to now have a bunch of folks who are opposed to testing….

    second – testing and analyzed data does you no good what-so-ever – if the school system does not set aside the resources needed to address the deficits the testing and assessments have revealed.

    this should not be rocket science for anyone.

    it works for all students – regardless of their demographics. It just turns out to be hugely important for at-risk, economically disadvantaged kids who tend to have more deficits and deficits in key areas that suffocate learning if not diagnosed AND remediated.

    It works not only in Goochland – and congrats to Goochland. It works wherever it is tried.. it’s the basic framework for the Federal Title 1 program – i.e. identify the deficits – provide the right resources to address the deficits.

    Finally – give DOE credit – for realizing that money is needed and coming up with the grant money and the school system for applying for the grant.

    Now it will be up to county taxpayers and state legislators if they choose to provide the necessary future funding to carry the program forward – as grants, by definition are pilot programs .. “seed” grants which allow something to be tried but then rely on the locality – once the program proves it works – to fund it.

    and as we have seen with other programs like Head Start and even Title 1 – some would rather argue that it’s the parents fault and bad genes and having to pay “extra” to help these kids is unfair and punishes taxpayers to be paying for things that should not be required!

    1. Hill City Jim Avatar
      Hill City Jim

      Not bad genes…different genes!

      1. larryg Avatar
        larryg

        I’d go along with “different”.

        the “good” part about the concept of “big data” for k-12 education is substantial in that:

        1.- there is an acceptance/acknowledgment/agreement that we need a standardized way to measure – with the emphasis on the word “measure” which means you really do want to know what each kid has learned (or not) – no matter the parent or genes.. just measure.

        2 – measure enough things with enough granularity that you really do know what specific areas the kids are behind in.

        3. a real educational plan for addressing the deficits

        4. a commitment of funds and staffing – the right kind of staffing to deal with the specific deficits…per the plan.

        if what it takes to do the above is to call it “big data” – awesome!

        so of course, being the contrarian I am thought to be – I’d like to see the Koch Brothers and ALEC get into this game from the voucher school side of things.

        set up some schools with “grants” from Koch and company that will emulate, compete with, and outdo the public schools with it’s use of “big data”.

        How about it ?

        How about Heritage Action, CATO and the GMU Mercatus Center set up their own private sector “big data” program using ‘good’ non-union teachers and prove that the private sector can equal or beat Goochland and deserves tax dollars to do the job “better”?

        I’d LOVE to see a genuine competition between the private sector and the public sector on K-12 education – a legitimate apples-to-apples competition and what better way that to use the same academic standard – the SOLs or Common Core – NAEP – pick your standard … and have at it – but do use a standard.

        I’ll bet that one of the very first things you’ll notice is that the private schools are going to throw all of their efforts into core academic only – just as Europe and Asia do – and in doing so – they’re going to demonstrate that one of the problems with public schools is their soup-to-nuts approach that while satisfying to parents who want “more” than just a core academic education for their kids – will cause folks to think about good core academic education at less cost as an option where “add-ons” can be purchased for extra fees by the parents.

        then, perhaps we can really deal with the truth of the matter – not what we want to believe – but what really is.

        I believe that if k-12 focuses solely on core academic ONLY that we’d see less expensive schools as well as better overall results.

        and I further believe that most voucher/choice schools would focus almost entirely on core-academic unless more dollars were proffered.

        In other words, If you gave a private school 10K to educate a kid and the ONLY thing you really measured and held them accountable for was the SOLs – they’d achieve that goal .. and .. make a profit.

        okay .. so now .. others here can call me full of crap – (which is fine) but do explain why!

        😉

  2. Hill City Jim Avatar
    Hill City Jim

    I have been a little reluctant to post this but here goes. And only because JB has blogged about VDOE and big data.
    Recently a taxpayer in Loudon County filed a FOI against VDOE and then won in court the right to see data that measures how well students are progressing. We all know SOLs tell us what the student knows but this new data SGP, Student Growth Percentile, tells us how well that student is progressing against similar students in the state. VDOE states “A student growth percentile complements a student’s SOL scaled score and gives his or her teacher, parents and principal a more complete picture of achievement and progress. A high growth percentile is an indicator of effective instruction, regardless of a student’s scaled score.” Conversely, to that I can assume that a low growth percentile is an indicator of ineffective instruction.
    My friend, John, @ CrankysBlog has done a fabulous job of sifting through the weeds and has compiled a very good explanation of the background of the effort to obtain the “secret” SGPs along with a look at Richmond City scores. From this link you should be able to find lots of background info that should you choose will point you in the right direction to get your school’s SGP. But hurry, the teacher’s union is trying very hard to both discredit and remove the data from the public’s eyes. I mean, who want to know where and who the good educators are?
    http://calaf.org/?p=982

    1. larryg Avatar
      larryg

      interesting stuff and thanks for posting..

      re: ” A high growth percentile is an indicator of effective instruction, regardless of a student’s scaled score.” Conversely, to that I can assume that a low growth percentile is an indicator of ineffective instruction.”

      Maybe I’m missing something but say two kids have the same SOL and a year later one falls behind on the score… as opposed to one doing better than expected.

      I’m not yet fully convinced that these numbers are reflective ONLY of instruction – effective or ineffective..

      some of this depends on where they started initially – were they both “passing” SOL or “proficient” or behind?

      if a kid is behind in 3rd grade reading but in a year, he gains gains one year but still end up behind – the same amount – does that mean effective instruction?

      if a kid is behind in reading in the 3rd grade – and does not get remediated – it’s possible, it’s likely that he’ll fall even further behind – even if the teacher is teaching effectively for all the other kinds – that are not behind.

      I think there is more to this.. although I’m totally in favor of using data analytics – correctly….

      and I’ll admit there may be some not-wonderful teachers that may need to seek other opportunities but at the end of the day – when a child has deficits – a “good” teacher who does well with students on grade level – is not necessarily going to do “good” with a child that is behind – if he does not get additional help. At that point – where there are more than one teacher involved, how do you determine “effective” vs “ineffective”?

  3. larryg Avatar
    larryg

    I actually think “big data” analytics will teacher’s friends if the teacher is assigned a classroom for which she/he simply does not have the resources to cope with. The data will tell the tale and in many cases it will be that the teacher is not the culprit but the administration and it’s allocation of resources and staffing.

    the idea that any teacher should be able to teach any assigned class regardless of the kids individual abilities and the aggregate class abilities and if they fail that they are “bad” teachers protected by the union – that idea is going to get blown away by data analytics which will tell the true tale.

    nine times out of ten – what happens is that 1/2 or 3/4 of the class progresses as expected but some kids will not unless they get special help. Should those kids have even been put in a class with others of higher capabilities in the first place?

    the question is – who is responsible for making sure ALL the kids in the class -achieve the standards ?

    if any teacher, even super-good teachers are given a class that has more challenges than she/he can handle – how does that teacher not get blamed and labelled as “bad”? If a brand new teacher right out of college gets put in a poor neighborhood school where 80% of the class in at-risk, economically-disadvantaged, and on free lunch – and 80% of the class fails to achieve the benchmarks – what safeguards are in place to recognize that putting a brand new teacher in that situation does not make that teacher a “bad” teacher – but rather will call into question the competence and efficacy of the administration.

    My view is that scapegoated teachers – when given their due process – are going to use the data analytics to defend themselves and in doing so are going to shift the focus of the failures more towards the real culprits – the administration.

    I predict that “Big Data” is going to be “Big Trouble” for any school system that does not do what Goochland is doing. It’s not going to be about the teachers – it’s going to be about the school leadership and administration.

    Further – let’s do this with private/voucher/choice schools also.. let’s get a unified level playing field that focuses on the mission and leadership and not the folks following orders.

Leave a Reply

Read More