Articles

More articles

• Tool	AI training	Proof
  Brisk	Never used	SOC 2 Type II, state agreements
  Class Companion	No identifiable data to AI providers	State agreement by email
  CodeGrade	Not stated	ISO 27001
  CoGrader	Not used; providers barred	SOC 2 Type 1
  EssayGrader	Not stated	NIST-aligned, A4L member
  GPTZero	API text not used	SOC 2 Type II

  Comparison · 13 October 2026

  What ten AI grading tools say they do with student data

  Ten AI grading tools on AI training, FERPA and COPPA, audits and deletion of student work, from each vendor’s privacy pages on 13 October 2026.

• The storm in chapter three is more than weather. When the sea is “closing in like a fist”, the image mirrors Ana’s fear of the trial, and the short sentences that follow speed the reader up with her.

  This shows the author uses setting. Later the calm harbor does the opposite.

  Evidence 4 quote fits the claim

  Analysis 2 say how, not that

  Comparison · 9 October 2026

  Free AI grading tools for teachers, and what the free plan counts

  What ten AI grading tools give you for free, how each counts its limit, and what the paid plans cost, from vendor pages checked 9 October 2026.

• The storm in chapter three is more than weather. When the sea is “closing in like a fist”, the image mirrors Ana’s fear of the trial, and the short sentences that follow speed the reader up with her.

  This shows the author uses setting. Later the calm harbor does the opposite.

  Evidence 4 quote fits the claim

  Analysis 2 say how, not that

  Comparison · 8 October 2026

  AI essay graders compared, from Brisk and CoGrader to Turnitin

  Eight AI essay graders compared on what they grade, rubrics, AI detection, integrations, free plans and price, from vendor pages checked 8 October 2026.

• Student’s answer

  Points

  $\text{(a)}\;\; 4w + 8 = 36$

  1

  $4w = \htmlClass{lab-err}{24}$

  36 − 8 = 28

  $w = 6\ \text{cm}$

  0

  (a) 1 of 2

  Comparison · 8 October 2026

  AI grading tools for handwritten math and science, compared

  Eight AI grading tools for handwritten math and science compared on scans, partial credit, carried errors, free plans and price. Checked 8 October 2026.

• def average(scores):
      total = sum(scores)
      return total / len(scores)

  test_three_scorespassed

  test_one_scorepassed

  test_empty_listZeroDivisionError

  2 of 3 tests passed

  Comparison · 8 October 2026

  AI grading tools and autograders for programming assignments

  Ten tools for grading coding assignments compared on tests, AI feedback, languages, similarity checks, LMS, free plans and price. Checked 8 October 2026.

• Student’s answer, Question 1

  Points

  $\text{(a)}\;\; m = \dfrac{11 - 3}{5 - 1}$

  M1

  $m = \dfrac{8}{4} = \htmlClass{lab-err}{3}$

  A08 ÷ 4 = 2

  (a) 1 [2]

  Guide · 8 October 2026

  How AQA, Edexcel, OCR and the IB mark an error carried forward

  AQA, Edexcel, OCR and the IB all follow an error through. They differ on the working they need to see and on where it stops.

• Whole-class feedback

  9th grade, linear equations test

  Done well

  • Collecting like terms (Q1–Q2)
  • Checking answers by substitution (Q5)

  Common errors (number of students)

  • Bracket not multiplied out fully, e.g. $2(x + 3) \to 2x + 3$11
  • Sign lost when moving a term across7
  • No working shown on Q65

  Re-teach at the start of next lesson

  • Expanding brackets: two worked examples on the board, then three for students
  • Moving terms: “do the same to both sides”, not “move and change sign”

  Do now

  • Redo the question where you made the error, in green
  • Then Q8–Q10 on the sheet

  Written once for the class instead of on every paper.

  Template · 8 October 2026

  A whole-class feedback sheet for a graded test

  One page for the whole class: what went well, which errors came up and how many made each, and what the class works on first next lesson.