ABC, You Don’t Know Me
Data Driven Lens on Teacher Behavior
Read by Author - Verena
I attend teacher planning and data evaluation meetings multiple times a week. I listen to deep discussion about trends in student data. I spent an entire meeting recently observing teachers move students around in Tier II intervention groups, as if they were plotting the NBA draft. This student should be put in this group because their data says they have high fluency but low comprehension. That student should be put in that group because of their task avoidant behavior.
They poured over benchmark fluency scores, accuracy percentages, comprehension grade placement numbers, statewide summative assessment ranks, and anecdotal behavioral observations. Each of the 112 students' data was neatly organized on a spreadsheet. Round and round they went for almost an hour until they settled on their ideal intervention groups. So many data points, so much deliberation, all in the name of student achievement.
As they say, Data is King, and in public education each student is decomposed and reconstructed as a pixelated version of themselves. Personal information, attendance, academic records, health information, behavioral information, special needs and accommodations, language proficiency, and parental involvement all patched together to help teachers predict student outcomes. Fortunately, there are countless edtech companies here to help disaggregate this mountain of information.
Schools lean on early warning systems, learning management systems, student information systems, and behavioral platforms all with built in analytics tools designed to visualize the student today, tomorrow, and in perpetuity. These ABCs, attendance, behavior, and course performance, are the holy grail of predicting high school completion.
Alas, something is missing…
John Hattie’s research synthesized findings from 1,400 meta-analyses of 80,000 studies involving 300 million students, into what works best in education. And, what did he determine would have “potential to considerably accelerate student achievement?”
COLLECTIVE TEACHER EFFICACY
“Collective efficacy is the perception of teachers in a school that the efforts of the faculty as a whole will have a positive effect on student learning.”
In my career, I have attended thousands of data meetings similar to the one I described. Yet, I can’t recall a single time when the conversation centered on teacher perception.
My college bestie sent me a link to a recent podcast. Given my adoration for the special guest, I tuned in immediately. Trevor, Christiana, and Ruha went around the table touching on topics like DEI, black representation, student protests, VR technology, and the possibilities and dangers of AI technology. Right there, around the 56 minute mark, Ruha put to words something that has been nagging at me.
“So many of the things in education are trying to predict whether students are going to graduate or be successful, or whether they're at risk. And they're reproducing the categories like if I say guess which students are deemed higher risk by these AI systems, right. You know who that's going to be, in my view, rather than pointing it to the students, let's figure out which adults are creating risks for these students. Let's train the AI to figure out which fields and departments are creating a hostile environment for these young people, but we never turn the lens to those who actually have the power to shape the experience. We always look at the most vulnerable and label them and stigmatize them.”
Hmmm? Sounds straightforward enough. We use data and algorithms all the time to predict what MIGHT happen to a student. Seems like it would be even easier to evaluate past data to paint a picture of what IS happening with teachers.
Easier said than done.
Although we know the effect size of teacher efficacy, from an individual teacher perspective, you’d be hard pressed to find a teacher willing to allow an automation to evaluate their impact. You’ll hear valid arguments about:
Fairness. Teachers should be allowed to to review and contest any findings that they believe to be in error.
Privacy. The importance of balancing the need for insights with respect for teachers' professional autonomy.
Consequences. Evaluations of teachers should not lead to punitive actions but rather be used as tools for growth and improvement.
Bias. Algorithms themselves can be biased based on the data they are trained on.
It is clear in education, what’s good for the goose is NOT good for the gander. We are so willing to minimize students into data points but balk at the idea of using the same methodologies to appraise the grownups or the system as a whole. Millions of dollars and countless hours go into teacher evaluation systems just for the majority to be deemed as effective—even when student achievement is abysmally low. Exactly to Ruha’s point, we have schools where students have reading proficiency below 25% and not one adult is put in an intervention group.
My ego would relish the turning of the tables. I would love to know which teacher has a history of outsized punitive action toward black and brown students. I would love to know which teacher has constantly low expectations for her IEP students. I would love to be able to say, “What the fuck are you doing? That’s the dumbest shit I’ve ever seen!”
Ego and callousness aside, we are missing a genuine opportunity to both minimize the harms of excessive data reliance exacted on students and superficial and unreliable data collection on adults. Hattie’s findings illuminate the point. Knowing what someone can’t do and predicting what they won’t achieve *unless,* doesn’t really move the needle. When the grown-ups focus on assets rather than deficits, not just of students but themselves, they create an environment less obsessed with data and more focused on fostering a holistic, supportive, and transformative educational experience. With drastic budget cuts on the horizon collective teacher efficacy offers us a free, timely, and contagious response to uncertainty.
🔗All That and a Bag of Chips
Delight № 106 - Clever sales pitches.
Delight № 107 - Meals on Wheels.
Delight № 108 - Sneaky kisses.
🔗Read This
Jambalaya: The Natural Woman's Book of Personal Charms and Practical Rituals by Luisah Teish
Free digital libraries
Libby App Thousands of public libraries offer ebooks and audiobooks (for free!) in Libby.
hoopla hoopla puts your public library at your fingertips. Anytime. Everywhere!
Used Books
ThriftBooks ThriftBooks sells millions of used books at the lowest everyday prices.
Also, check out your local thrift stores!
If the ‘Upgrade to Paid’ button is a little too much, consider a one time donation.
Use the link below to support Crayons & Code via PayPal.
Also, checkout what Conrad and I are up to on the No Opt-Out Podcast!
"Each of the 112 students' data was neatly organized on a spreadsheet"
You could have omitted the 'data' and inserted 'as a line' before 'on'. Not a grammar check, but rather a snapshot the growing view of people themselves as data (points) only. Even worse when it's young, impressionable students being converted to data points.
Thanks for shedding some light on this.