Concepts
Ideas that keep showing up in the findings, each defined once.
The scissors
A credential rises while the outcome it promises falls.
A credential and the outcome it stands for moving in opposite directions, so the two lines cross like open scissors. In the Texas data: the state's "college, career, or military ready" label for Black graduates roughly doubled (37% for the class of 2017, 75% for 2024) while actual enrollment in a Texas college fell (54% for the class of 2016, 45% for 2022), and every other group shows the same shape.
The access gap
Same test scores, different odds of getting the advanced class.
The distance between how students score and how often they are placed in the advanced track, measured inside the same building. In 429 Texas high schools, White students complete advanced courses about 6 points more than Black students who score the same on the state Algebra test, and at 14 schools Black students score equal or higher yet still complete advanced courses less. Preparation explains only about half of placement (the correlation between the score gap and the course gap is 0.50); the rest is whatever sits between scoring well and being in the room.
The frozen gap
One measure says the gap closed. The harder measure says it never moved.
A gap that closes on one measure while staying still on another, which usually means the measure moved, not the students. Texas' readiness-checklist gap between Black and White graduates closed from 28.5 points to 11.4; over the same era the SAT/ACT college-ready score gap went from 33.6 to 31.3, essentially unchanged since 2012.
The moving ruler
The line moved because the definition moved, not the students.
A measure whose definition changes underneath a series, so the line moves when no student did. Texas school data is full of ruler moves: adopting the federal dropout definition nearly tripled the reported rate in one year (0.9% to 2.6%, 2004-05 to 2005-06), with the state itself calling the years not comparable; the 2023 accountability refresh raised the score needed for an A rating from 60 to 88 against a class that had already graduated; and a 2018-19 masking rule change cut the share of masked cells from 38% to 20% with no change in any student's data. The defense is to mark every break and refuse to draw a line across one.
The folklore check
Trace the number everyone repeats to its primary source.
Taking a number people repeat and tracing it to a primary source. The repeated number is often wrong: "about 3 percent of high school football players play in college" matches an old Division I-only figure, while the rate across all NCAA divisions is 8.1 percent, and "two in ten thousand make the NFL" is too grim, since NCAA's own chained rates put drafted alone near eleven in ten thousand.