Good Student, Bad Student, and the Kids Who Are Both

Most of us grew up in a system that sorted kids into simple categories. Good student, bad student. Good at reading, bad at math. If the system got more specific than that, it usually meant you were either in the gifted program or you were getting extra help. Not both. The categories were mutually exclusive.

Scott Barry Kaufman didn’t fit either one. He had an auditory processing disorder and crippling anxiety.¹ The system saw the disability and labeled him as disabled. What the system didn’t see was that Kaufman was also gifted. He was curious, creative and intellectually ravenous. His IEP said he needed to stay in special education because he was anxious.² It didn’t ask why he was anxious, or what might happen if someone showed him what he could do instead of reminding him of what he couldn’t.

One teacher did. She saw past the label. She encouraged him to sign up for harder classes, join orchestra, try out for musicals. “She saw beyond my label,” Kaufman later wrote, “to some hidden strengths that were bursting to come out from deep within me.”³

Kaufman eventually earned a PhD in cognitive psychology from Yale.⁴ He now studies intelligence at Columbia. He edited the book that named the field: Twice Exceptional.⁵

His story is a sample size of one, but the system that nearly lost him hasn’t changed much since.

What “twice-exceptional” means

A twice-exceptional (2e) student is someone who is both gifted and has a learning or developmental challenge, at the same time. High cognitive ability plus dyslexia. Exceptional creativity plus ADHD. Intellectual giftedness plus autism. The combinations vary.

What they share is Kaufman’s paradox: the gifted side can compensate for the disability, making the disability invisible. Or, the disability can suppress performance, making the giftedness invisible. So either a kid is pigeon-holed in the wrong place, or their strengths and weaknesses sort of average each other out in the eyes of the education system, hiding the truth. A kid who reads three grade levels ahead but can’t sit still looks like a behavior problem, not a gifted one. Or, teachers see a kid who does “fine” and move on.

Think of it this way. Nobody would confuse deafness with math ability. A person who is deaf and also a math genius obviously has two separate things going on. The deafness doesn’t make them worse at math. Getting better at math won’t fix their hearing. Nobody would make them sit through a lecture without an interpreter and then wonder why they failed the exam. The accommodation (the interpreter) isn’t a special advantage. It’s what lets their actual ability show up in the room.

For 2e students, the separation is just as real but harder to see. A kid with severe executive function challenges and exceptional analytical ability has two neurologically independent things going on. One doesn’t cancel the other out. Taking away the accommodation (extended time, a reader, assistive technology) doesn’t reveal the “real” student. It hides them. It’s the equivalent of taking away the interpreter. The difference is that executive function challenges don’t look like deafness. They look like laziness, or carelessness, or not trying hard enough. That’s why the accommodation sounds like a perk to people who don’t understand the neurology. It isn’t. It’s what lets the student show up as themselves.

No one knows how many 2e students there are. In fact, some of you reading this right now might be thinking “hey, that sounds like me”, or “that sounds like my kid!” You might be right.

The systems that screen for giftedness and the systems that diagnose disability operate separately in most schools. Most students are never evaluated for both.⁶ Twice-exceptionality is a spectrum, not a binary, and by most expert estimates the population is far larger than commonly assumed.

Representing what these kids achieve is 2.5 times harder than identifying them in the first place

I wanted to understand why the system keeps losing students like Kaufman. Not the individual stories (there are thousands) but the structural picture. What, specifically, is broken? So we took the system apart. We computationally mapped 309 components across two overlapping value chains: one for identifying 2e students, one for grading, testing and credentialing them (accurately communicating what they achieve, such as in the college admissions process).⁷

Most conversations about 2e students focus on identification. Can the system find these kids? That’s important. What we found is that the system designed to represent what these students achieve is 2.5 times more bottlenecked than the system designed to find them.⁷

About one in ten parts of the identification system is meaningfully stuck. In the grading and credentialing system, it’s closer to one in four.⁷ In other words, once a 2e kid is found, the harder problem is accurately representing what they achieve. That includes designing fair assessments, building accommodations into testing, creating credentials that carry context, and then communicating all of that from one institution to the next. When that chain breaks, the student pays for it. A kid does A-level work with accommodations, but the transcript can’t show how it was achieved. A family moves districts and the new school has no record of the 2e profile the old school spent months building. A college application arrives looking unremarkable because the system flattened a remarkable student into a GPA.

Why does the chain break? Identifying a 2e student mostly happens inside one school. A psychologist administers tests, a team meets, a plan gets written. It isn’t easy, but it’s doable. Grading and credentialing are a different story. They require many institutions to agree on shared data formats, shared standards and shared workflows. It’s not enough to understand you’re looking at a 2e student. You have to communicate that understanding from one bureaucracy to the next. That coordination problem is structurally harder.

The tools exist. The training works. The system still fails.

Here’s what surprised us most. Every training tool in the system is available. The people can be trained. The system still can’t function.⁷ This isn’t a knowledge problem or a skills problem. It’s an architecture problem.

The bottlenecks cluster where institutional coordination is required: moving data between systems, designing new credentials, adopting new standards. At roughly 80% of bottlenecks, the barrier isn’t a lack of technology.⁷ It’s the institutional environment that surrounds it. Until December 2024, the national data standard that schools use to exchange student information didn’t have a slot for disability accommodation plans.⁸ It literally couldn’t carry the data. It does now, but only about 26% of states have implemented the standard at all.⁷ The standard for carrying richer learner records between institutions was only finalized in March 2025.⁹ The specifications are arriving. The institutional adoption is years behind.

The unintended consequences of nondisclosure

Federal civil-rights guidance says transcripts generally may not indicate a student received accommodations, because noting accommodations can identify disability status. Colleges generally may not ask applicants whether they have a disability before admission.¹⁰

This protection was designed to prevent discrimination. For a 2e student, the protection becomes the problem. The student uses accommodations that let them produce work reflecting their actual ability. The transcript records the grade. It cannot record how it was achieved. The student’s story arrives at the next institution incomplete. Not through neglect. By design. The college that admits her doesn’t know she needed a reader, or extended time, or assistive technology. The support structure that produced the A doesn’t follow her. She starts over.

Some schools already build richer records: mastery transcripts, portfolio assessments, narrative evaluations. The problem isn’t creating them. It’s getting the next institution to read them. The Common App has no field for it. Admissions offices don’t have rubrics for it.⁷

The legibility problem

The deeper I got into the data, the clearer it became that the 2e problem isn’t really about tools or training or even funding. It’s about “legibility.” Can the system actually see these students clearly enough to represent who they are and what they can do? Right now, it can’t, and the distortion happens in layers. Gifts hide disabilities. Disabilities hide gifts. Then the credentialing system adds a third layer: it strips the context that would make the student’s story whole.

A school psychologist spends 12 hours assembling a 2e profile by hand because no system fuses cognitive, achievement and behavioral data automatically. That profile exists on paper in one district. It can’t follow the student to the next one.⁷

A state wants to know whether its accommodations are actually helping 2e students. The data sits in three systems that don’t talk to each other. The question goes unanswered.⁷

The system can’t improve what it can’t measure. Right now, it can’t measure.

Here’s the thing. The system was designed for students whose strengths and challenges sort neatly into categories. Gifted or struggling. Good at this, bad at that. That design made sense when the goal was efficiency. It stops making sense when the goal is accuracy. Fewer kids fit those categories than we’d like to believe. The infrastructure that would make 2e students visible (portable learner profiles, linked records, credentials that carry context) would serve any student whose story doesn’t compress into a GPA and a test score. That’s a lot of students.

Kaufman didn’t fit the categories. One teacher saw him anyway. He became a cognitive scientist at an Ivy League university. Most kids who don’t fit the categories aren’t that lucky. They don’t get the teacher. They get the system. The data showed us something we didn’t expect: seeing the student isn’t the hard part. The hard part is getting the adults to carry what they see from one institution to the next. The teacher sees the kid. The transcript doesn’t. Right now, the transcript speaks for them and doesn’t know their story.

Endnotes

¹ Kaufman has described his auditory processing disorder and anxiety in multiple published accounts. See: Cava, Anya. “Why So Many Gifted Yet Struggling Students Are Hidden in Plain Sight.” NPR, May 9, 2018.

² Kaufman described his IEP in an interview: “It was put in my IEP… he needs to forever be in special education because he’s an anxious person.” See: Reber, Debbie. “Dr. Scott Barry Kaufman on Redefining Giftedness and Intelligence.” Tilt Parenting, Episode 139, January 8, 2019.

³ Kaufman, Scott Barry. “The Twice Exceptional Movement: Supporting Bright and Creative Students with Learning Difficulties.” scottbarrykaufman.com, January 25, 2018.

⁴ Kaufman received his PhD in cognitive psychology from Yale University. He also holds a B.S. from Carnegie Mellon University and an M.Phil. from the University of Cambridge. See: scottbarrykaufman.com/bio.

⁵ Kaufman, Scott Barry, ed. Twice Exceptional: Supporting and Educating Bright and Creative Students with Learning Difficulties. New York: Oxford University Press, 2018.

⁶ The systems for identifying giftedness and diagnosing disability operate independently in most U.S. school districts. See: Foley-Nicpon, M. and Assouline, S.G. “The Paradox of Twice-Exceptionality.” Belin-Blank Center, University of Iowa. See also: Reis, S.M., Baum, S.M., and Burke, E. “An Operational Definition of Twice-Exceptional Learners.” Gifted Child Quarterly 58, no. 3 (2014): 217–230.

⁷ Growth Science Ventures. “Identifying Twice-Exceptional K-12 Students: A Value Chain Analysis.” April 2026. Analysis of 309 components across two value chains covering identification and grading/credentialing of twice-exceptional students in U.S. K-12 education.

⁸ Ed-Fi Alliance. “December 2024 Releases: Data Standard 5.2, ODS/API 7.3, and MetaEd 4.5.” December 18, 2024. Data Standard v5.2 added Section 504 program support to the Ed-Fi Unifying Data Model.

⁹ 1EdTech Consortium. “Latest Comprehensive Learner Record Standard Improves Security, Portability and Verification of Digital Credentials.” Press release, March 4, 2025.

¹⁰ U.S. Department of Education, Office for Civil Rights. Federal guidance on disability disclosure in educational records and preadmission inquiries. See also: Section 504 of the Rehabilitation Act of 1973; Americans with Disabilities Act of 1990.