
219: On Grading on 100-Point Scales
An AI-created episode based in a human-composed blog post: On Grading on 100-Point Scales
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

An AI-created episode based in a human-composed blog post: On Grading on 100-Point Scales
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
You know that feeling, right? Oh, I think we all do. Yeah, you're... You're staring at a screen or a piece of paper, and your heart is just sinking into your stomach. Because you're agonizing over a single number. Exactly. You're looking at an 89%, and you are just torturing yourself over the fact that it is not a 90%. It's the absolute worst. It is. Because that one tiny percentage point, I mean, that's the difference between an A and a B. Right. Between, you know, a celebration and a total disappointment. And it feels so absolute, like it's mathematically certain. Yeah, we tend
to treat that single digit as this, like, objective truth about our intelligence or our effort. Right, like it's a permanent label. Exactly. It carries just an incredible amount of psychological weight for something that, you know, on the surface seems so microscopic. Well, today we are doing a deep dive into a genuinely eye -opening article by Dr. Gary Ackerman. From Hack Science Education, right? Yes, exactly. It's called On Grading on 100 -Point Scales. And what Ackerman does here is essentially, well, he pulls the rug out from
under that 89%. He really does. He breaks down the fundamental ways we quantify human knowledge, and he demonstrates why the standard 100 -point scale, the one practically every single person listening has experienced, is just scientifically flawed. It's mathematically flawed, too. Right. Our mission today is to prove why that 1 % difference you are stressing over might actually be a complete mathematical illusion. Yeah, and to really understand where the 100 -point scale fails, if we connect this to the bigger picture, we have to look at
a foundational error of 20th century psychometrics. Oh, wow. OK. Going back a bit. Yeah. Ackerman actually points us toward evolutionary biologist Stephen Jay Gould. Oh, right. His book, The Mismeasure of Man. Exactly. Landmark work. Gould identified this. core logical fallacy that early scientists made when they first attempted to put a hard number on the human mind. Right. It basically comes down to a concept called reification. OK, let's unpack this. What exactly does reification mean in the context of intelligence and testing?
So reification is our human tendency to take an abstract, highly complex concept and mentally convert it into a hard, singular, physical thing. Give me an example. Well, think about human intelligence. It's not a single, tangible in your brain that we can just take out and put on a scale, right? Right, obviously. It's this incredibly vast, fuzzy ecosystem of memory, analytical reasoning, creativity, adaptability. It's a million different things all working together. Exactly. But early
researchers, they decided they needed a way to rank people efficiently. So they reified intelligence. They made it a thing. Yes. They designed tests, generated a single overarching number like an IQ score, and then operated under the assumption that this number was a literal physical quantity of smartness inside a person's head. Wow. So they basically mistook the map for the territory. Oh, that's a brilliant way to put it. Like they thought the shadow on the wall was the actual object they were trying to measure. Yeah, they assumed that because they could produce a number, that number represented a distinct, perfectly
objective reality. And people just bought into it. Totally. And that assumption gave institutional weight to a fundamentally flawed measurement system, because once you convince people that an abstract concept can be reduced to a precise number, well, policies are built around those numbers. Right. Like funding gets distributed based on those scores. Exactly. Students get tracked into different career paths and immense pressure is put on achieving specific numerical thresholds. OK. But but wait, let me push back on this a little bit. Sure. Go for it. We aren't
talking about 1920s intelligence tests here. If a teacher gives a Tuesday morning pop quiz on, say, cell biology, they aren't claiming to measure a student's innate soul or overall human intelligence rate. They just want to know if the student memorized the function of the mitochondria. The powerhouse of the cell. Exactly. Right. So doesn't narrowing the scope of the assignment bypass all that historical baggage? Like, why assume a standard classroom assignment suffers from the exact same flawed science? Well, because
the mechanism of the error is exactly the same. being applied to a narrower domain. Even on a Tuesday morning pop quiz, the teacher is assuming that their specific instrument, like a multiple -choice test or short essay, is perfectly capable of capturing the internal cognitive state of the student. Oh, I see. But what is that quiz actually measuring? I mean, is it measuring pure comprehension of cell biology? Or is it measuring, like, the student's test -taking strategy? Or their reading speed. Yes. Or how well they slept
the night before. Or even how well they can parse the subtly ambiguous phrasing of question number four. Right. So the test isn't a pure conduit to their knowledge at all. Not even close. It's heavily filtered through all these external variables. Exactly. And this is where Ackerman introduces this brilliant, really tangible analogy to explain why this matters. Oh, the ruler analogy. I love this part. Yeah. He asked us to just step away from the abstract idea of knowledge and look at how we measure physical things in the real world. Right, just a standard everyday ruler
you might have sitting on your desk. Yeah, standard United States ruler. So Ackerman points out that U .S. rulers are divided into these ridiculously difficult to differentiate fractions of inches. Like down to one sixteenth of an inch. Right. Compare that to the metric system with millimeters, which is way easier, but the fundamental constraint is identical either way. Which is what's fascinating here. The physical marks printed on that ruler dictate your absolute limit of precision. Right, you cannot measure anything more accurately than the closest sixteenth of an inch because your
tool physically lacks the capability to see anything smaller. Exactly. And this brings us to the concept of bins, which is crucial for understanding how measurement actually works. Bins, right. Break that down for us. So imagine you were handed this massive pile of random tiny objects, like paper clips, screws, pebbles, whatever. OK, I'm picturing it. And all of them are less than one inch long. Your task is to measure them and sort them by length using only that standard English ruler. OK, so I'm measuring tiny pebbles with a plastic school ruler. Right. And because your
tool only goes down to one sixteenth of an inch, you are forced by the laws of physics to sort all of those distinct items into a maximum of 16 distinct bins. So if I pick up two slightly different pebbles and their true lengths differ by less than a sixteenth of an inch. They both go into the exact same bin. Because the ruler can't tell them apart. They have to go in the same bin. Let's look inside, say, the two sixteenths bin. The items sitting in that specific tray are not mathematically identical. Right. One pebble might be a microscopic fraction longer.
Another might be slightly shorter. But you are forced to group them together and treat them as equals because your instrument simply cannot tell the difference. Wow. This is the foundational rule of all measurement from carpentry to quantum physics. Every single measurement relies entirely on the reliable precision of the instrument. So if I'm understanding this, if you want to know which of those pebbles in the 216 spin is truly longer, you can't just keep squinting at the same cheap plastic ruler. No, you need a completely different tool, like a digital caliper
designed for submillimeter precision. Right. Using a standard test to measure a student's brain is like trying to measure a microscopic cell with a plastic school ruler. The tool just isn't built for that level of detail. That is a phenomenal way to illustrate the mismatch. You have an objective tool. yielding a highly specific number, but you're applying it to a subjective, multifaceted domain where that number is completely meaningless. And here's where it gets really interesting. Oh, yes. If we accept that a physical ruler has hard, undeniable limits
that force us to group slightly different objects into the exact same bin, what happens when we apply that exact same constraint to an invisible metric like human thought? It completely falls apart. That is where Ackerman applies the bin concept directly to the 100 -point grading scale. And that's where the whole system just starts to look mathematically absurd. It really does. Because when an educational system utilizes a 100 -point scale, they're making a massive unspoken claim. A huge claim. They're claiming that a classroom assignment, like an essay, a project,
a test, has enough reliable scientific precision to sort students into 100 distinct bins of comprehension. Which requires an astronomical leap in logic. We just established that a physical ruler struggles to accurately sort physical objects into 16 bins. Exactly. Yet we are supposed to believe that a history essay can accurately sort the invisible complexities of a teenager's brain into a hundred different categories. The instrument is simply not that precise. It is a blunt instrument being wielded as a surgical scalpel. I love that. A
blunt instrument. Yeah. When a teacher gives one student an 89 % and another student a 90%, the institutional implication is that the test has the power to accurately differentiate a literal 1 % difference in total human understanding. But what is that 1 % difference in reality? I mean, it is almost never a true reflection of 1 % less knowledge about the subject matter. Never. But wait, let me push back on this idea that grading is just a blunt, subjective guess. What about rubrics? A teacher isn't just staring at a history essay and pulling an 89 out of thin
air, right? Not usually. Right. Modern education relies heavily on these highly detailed rubrics. You might get 10 points for a strong thesis, five points for grammatical accuracy, 10 points for historical citations. And five points for structural flow, sure. Yeah. So doesn't breaking the assignment down into those micro components mathematically justify the precision of an 89 versus a 90? It absolutely seems like it does. which is exactly why rubrics have become the gold standard in education today. But if we examine
the actual mechanism of what is happening, rubrics often just hide the subjectivity behind a larger wall of numbers. Hide it. How? Well, let's look at that 10 points for a strong thesis category you mentioned. What is the objective, physical, undeniable difference between an 8 out of 10 thesis and a 9 out of 10 thesis? Well, it's still a judgment call by the person reading it. Exactly. It relies entirely on the evaluator's interpretation in that specific moment. I mean, is the teacher tired? Did they just read five poorly written
essays in a row, making your average essay look brilliant by comparison? Oh, that happens all the time. Did you use a transitional phrase they personally find clunky, even if it is grammatically correct? You are taking highly subjective, qualitative judgments. assigning them arbitrary numerical values, and then adding those values together. The final sum, that 89%, looks like hard, unassailable data. But the foundation it is built on is still entirely subjective. So the rubric gives the
illusion of objectivity. Yes. It makes the student and the parent feel like the grade was calculated rather than felt. Precisely. And just like the pebbles in the 2 -16 spin on our ruler, The student with the 89 % and the student with the 90 % almost certainly belong in the exact same bin of comprehension. They basically know the same amount of stuff. Yeah, they both generally grasp the material to the same degree. The rubric graded essay just lacks the calibration to reliably tell them apart. But if we, as a society, kind of implicitly know
this, like... If educators know, deep down, that an 89 .4 % and an 89 .5 % are functionally representing the same level of comprehension, which they do, then why do we push this system so far? Because it's not just whole numbers anymore. Oh, don't get me started on the decimals. Right. Schools are actively using decimals. A student doesn't just get a 90, they get a 90 .5. Yep. By adding that decimal point, aren't they claiming their test can sort students into a thousand different bins? Isn't that just the total illusion of precision?
One thousand bin scale. It is the absolute definition of an illusion of precision. It's crazy. A ninety point five percent implies you can divide human learning on a given Tuesday into one thousand distinct rankable categories. Which is impossible. There is not a classroom assessment on earth no matter how rigorous the rubric. that can justify that level of granularity. So why do we keep doing it? I mean, it seems to come down to institutional momentum. We aren't calculating grades to the second decimal point because it is scientifically sound. We are doing it because our digital infrastructure
demands it. Right. Greeting software platforms like Canvas or Blackboard, they're programmed by default to spit out decimal averages. It's entirely systemic. Yeah. Once a hydrospecific metric is introduced into a bureaucratic system, it just takes on a life of its own. A decimal point feels scientific. It feels unarguable. Oh, I see. So if a parent challenges a grade, handing them a spreadsheet that calculates to an 89 .4 % gives the institution like a mathematical shield. Exactly. Furthermore, higher education
relies on this data. College admission algorithms require fractional GPAs to automatically filter out thousands of applications. So we are locked into this illusion because the massive machinery we have built requires data to function. Even if everyone involved knows that data is fundamentally flawed. So what does this all mean? If everyone, the teachers creating the rubrics, the software calculating the decimals, the students agonizing over the fractions, if everyone is pouring all this time and emotion into arguing over a tenth of a percentage point? What is the hidden cost?
That's the real issue. What are we actually losing? This brings us to what I consider the most profound and vital insight from Dr. Ackerman's analysis. The cost of this illusion is what he terms wasted cognition. Wasted cognition. Yes. He describes the fallout of this system as widespread consternation. It generates immense anxiety, friction, and wasted effort across every single level of the educational ecosystem. Wasted cognition. That implies that brain power is like a finite resource being actively
drained away. Because it is. Cognitive load theory teaches us that human cognition, our working memory, our attention span, our capacity to process and synthesize information. It is strictly finite. We only have so much to give. We only have so much cognitive bandwidth to utilize on any given day. When a teacher spends hours mathematically defending a .5 deduction on a rubric, or when a student burns through their emotional and mental energy, stressing over it an 89 .4%. Which happens constantly. That cognitive energy is effectively
destroyed, is consumed by the bureaucracy of the grade. And because it's gone, it cannot be used for its actual intended purpose. Which is building new knowledge. Exactly. Engaging in deep learning. It is a massive energy leak in the system. Huge. By forcing everyone to focus on the artificial boundaries of the 1000 bins, we are literally draining the brain power that could have been used to actually master the material. And it fundamentally alters the psychology of the student, too. How so? Well, when the system emphasizes the microscopic difference between
bins, it destroys intrinsic motivation. Oh, I can see that. Students stop asking, how can I understand this historical event more deeply? Or how can I improve my writing style? Right. Instead, their entire cognitive focus shifts to gaming the system. Yes. They ask, what is the absolute minimum requirement to scrape together an extra 0 .5 % so I can cross the threshold into the A minus bin? Wow. The pursuit of knowledge is replaced by the pursuit of accounting. That's a sad but accurate way to frame it. Yeah. And
it completely fractures the dynamic between the teacher and the student. It creates a deeply adversarial relationship. It's no longer the teacher and the student teaming up to conquer the material. No, it's the student fighting the teacher for a fraction of a point. Which destroys the learning environment itself. Completely. As Ackerman points out, learning requires trust. It requires a psychologically safe environment where a student feels comfortable taking intellectual risks. Right, making mistakes, experimenting with new ideas. But you cannot foster that kind
of environment if every single misstep, every slight deviation from the rubric, is permanently recorded and calculated to the nearest decimal point. That's so true. When a student or parent challenges an unjustified grade and the teacher is forced by the system to defensively double down on their flawed instrument. It just breeds resentment. Yeah. It creates friction that actively interferes with the positive relationships absolutely necessary for teaching to happen at all. Man, when you lay out the historical baggage, the
physical limits of measurement, and the psychological toll of wasted cognition, the hundred point scale really does sound like a massive institutional failure. It's hard to unsee it once you realize how it works. But I want to make sure we are conveying Dr. Ackerman's actual conclusion accurately to you, the listener, because the ultimate takeaway from this deep dive isn't a call to, you know, burn down the school system or eliminate all standards. No, absolutely not. Ackerman is not arguing that we should abolish grading entirely.
Right. Measurement is a necessary component of growth. We need feedback to know if we're improving. whether we are learning algebra or training for a marathon. We need to know where we stand. Yes. The argument is not against measurement. It is against false precision. OK. So the goal is to adopt grading systems that are justified, defendable, and that actually contribute to your learning rather than actively hindering it. Exactly. We have to stop pretending our plastic rulers are digital microscopes. It's the perfect metaphor. If an instrument can only accurately sort students
into broader categories, like excellent, satisfactory, or needs improvement, then the grading scale should just reflect those broad categories. It is about aligning the precision of the scale with the reality of the tool, moving away from the 100 -point scale toward broader, more holistic evaluation methods. Yeah. It reduces the adversarial friction. It frees up cognitive load. And it redirects everyone's energy back toward actual learning. So whether you are a student navigating a degree program, a professional prepping for
a corporate performance review, or just someone trying to learn a new language for the sheer joy of it, it is always worth asking yourself if you are measuring your own progress with a faulty ruler. Right. Are you agonizing over the difference between an 89 and a 90 when the tool you're using to measure your worth can't even tell the difference? It's a huge perspective shift. It really requires a massive shift in how we view ourselves and our institutions. And it leaves us with a rather profound challenge. Well, if we accept the central premise of Ackerman's work, if we accept that human knowledge, learning,
and curiosity simply cannot be perfectly sliced into 100 or 1 ,000 perfectly neat little numerical bins. Yeah. What would a ruler designed to measure true intellectual growth and curiosity actually look like? Would it be a number at all? Or do we need to invent an entirely new language for measuring the mind?
Nobody is credited on this episode yet.
An AI-created episode based on a human-composed blog post: Changing Vision of Technology Literacy
An AI-generated episode based on an AI-generated blog posts based on an old microblogging post: The Inquiring Technologist: Why Great Solution Design Begins with "Too Many" Questions
A brief AI-generated episode based on a human-composed blog post: Gathering Organizational Intelligence for Network Security
An AI-generated episode based on a human-composed blog post: Models of Technology-Rich Teaching
An AI-generated episode based on a blog post designed for students in a community college course. Concepts of Operating Systems