Checked against the state's own scorers
We grade officially scored student writing and compare EnlightenAI with each state's trained scorers. Every study shows its method, its limits and every number.
STAAREssays: Ideas
Florida B.E.S.T.Purpose/Structure
Smarter BalancedOrganization/Purpose
DREAM field studyGrade 8 essays
State test studies
How to read these results
How we test, and what the numbers mean
The way a teacher uses EnlightenAI. For each prompt, five officially scored papers act as the teacher's graded examples, and EnlightenAI grades every other paper once. Papers used as examples are left out of the results, and any calibration note is written on half of the prompts and checked on the other half.
It is quadratic weighted kappa (QWK), the standard way state testing programs measure how closely two graders agree. It runs from 0 to 1, gives more credit for close misses than far ones, and gives 0 to a grader that hands every essay the same score. Above 0.8 is excellent.
Two trained scorers reading the same essay don't always agree, and states publish how often they do. If EnlightenAI agrees with the official score as often as two of the state's own scorers agree with each other, it is grading to the state's standard.
Released state papers are online, so AI models may have seen them during training. The DREAM essays have never been public, so they show how EnlightenAI does on new student writing. On those essays its agreement with DREAM's official scores was 0.77, and it was never more than a point off. Our guide to reading state scoring results explains how to compare any grader's numbers fairly.
Yes. Each study has a verification kit with every graded response, the examples used, our scores and feedback, and a script that recomputes the tables. Get in touch to request one.
See how EnlightenAI scores your students' writing
Talk with our team about your state test and your rubrics, or start free and grade your next stack of essays today.