The 30-Second Diagnostic Every Teacher Needs

Welcome to the show everybody! I'm Renata Salas, here with Colin Whitfield. And Colin, I want you to picture this. It's second period, 9:15 AM. You've just spent fifteen minutes explaining a concept. You look out at twenty-eight blank middle school faces, and you ask, "Does everyone understand?" One kid in the front row—usually named Marcus or Sophia—nods enthusiastically. So you say, "Great!" and you move on to the independent practice. And then [laughs]... the ENTIRE lesson collapses because actually, twenty-seven of them had absolutely no idea what you were talking about. Ah, yes. The classic "illusion of competence." You've basically outsourced the formative assessment of twenty-eight human beings to Sophia in the front row. And look, we've all done it, but it completely ignores the central provocation from Dylan Wiliam: "If nothing has changed in long-term memory, nothing has been learned." "If nothing has changed in long-term memory..." That is such a high bar. Because in the moment, as a teacher, a nod FEELS like a change in memory. It feels like learning. Right, but it isn't. Which is why in September 2015, Wiliam published a piece in *ASCD Educational Leadership* called "Designing Great Hinge Questions." And this is the method we want to unpack today. A hinge question is not a warm-up. It's not a discussion prompt to get the vibes going. It is a 30-second diagnostic check designed to reveal—incontrovertibly—whether your students are actually ready for the next twenty minutes of instruction. Thirty seconds. That's the part that I think catches people off guard. When we talk about formative assessment in my coaching sessions, teachers INSTANTLY picture grading exit tickets during their lunch period. But a hinge question relies on an immediate read of the room. You're using ABCD cards, or having them hold up fingers, or using something like Plickers. Exactly. And Wiliam provides a very specific threshold for this: the 80/20 rule. You put the question up. If 80 percent of the class gets it right, and the 20 percent who got it wrong chose completely scattered answers... you move on. You can catch those stragglers later. But if that 20 percent is bunched—meaning they all picked option C—[urgently] you have a crisis. They all picked option C. Yes! Because that means option C isn't just a random guess. It means there is a highly specific misconception that has infected a quarter of your room. And Jennifer Gonzalez over at *Cult of Pedagogy* has the best phrase for this. She calls it a "moment of contingency." The hinge is the moment your lesson plan legally has to bend to student reality. A moment of contingency. I like that. It strips away the ego of the lesson plan. Totally. Like in an ELA classroom, a classic hinge point is the shift from identifying a metaphor to actually analyzing its effect. If they can't accurately spot the metaphor in the text on a quick ABCD card check... what am I DOING asking them to write a paragraph about its thematic impact? I have to stop and re-teach. But this brings us to the actual mechanics of writing these questions, which is where the genius of Wiliam's approach really sits. He gives this classic 7th-grade math example. The students have to find the area of a semicircle. You give them four options: A, B, C, and D. One is correct. But the other three—the distractors—are engineered so that if a student picks one, you know *exactly* what cognitive error they just made. [leaning in] Wait, break that down for me. What are the errors they're making on a semicircle? Right, so Option B might be the answer you get if you use the diameter instead of the radius. Option C might be what you get if you calculate the whole circle and forget to divide by two. Option D is what you get if you mix up the formula for area with the formula for circumference. So when twelve kids hold up the card for Option C... you don't just know they're wrong. You know EXACTLY why they're wrong. They forgot to divide by two. Okay, so as an instructional coach, this is where I have to play the "bad cop." Because writing a multiple-choice question where every single distractor maps perfectly to a specific cognitive error? That takes an hour. And my teachers roll into the parking lot at 7:30 AM, print their materials at 7:42, and the bell rings at 8:00. The "hour-long problem" is why teachers end up just winging it and asking, "Does everyone get it?" Which is a completely fair critique of the research. In a laboratory, you have all afternoon to write the perfect distractor. In a public school in Chicago, you have thirteen minutes. Exactly. But—[hesitates] and I am gently skeptical of tech fixes usually—this is the exact use case where AI is a LIFELINE. You don't ask ChatGPT to write your lesson plan. You use it as a diagnostic engineer. You type in this specific prompt: "You are an expert in formative assessment. I am teaching 7th grade math, and the key concept is calculating the area of a semicircle. Write me one hinge question following Dylan Wiliam's rules. Provide one correct answer and three distractors that map to common student misconceptions, and tell me what misconception each distractor represents." "Tell me what misconception each distractor represents." That's the crucial parameter. Because it forces the model to justify the wrong answers, rather than just generating random numbers. Right! [snaps fingers] And it spits it out in ten seconds. You project it on the board, and suddenly you have a high-leverage diagnostic that used to take an hour to build. Which allows you to hit the ultimate checklist for this method. Every student must respond simultaneously—no opting out. The whole process takes under two minutes. And the data forces an immediate go/no-go decision. You either forge ahead, or you fix the diameter-radius confusion right then and there. It's just such a massive upgrade over "Does everybody get it?" So, try it tomorrow morning. Use that prompt, test the room, and let us know if your 80/20 split surprises you.