Knowledge
Whether the answer demonstrates accurate concepts and enough relevant depth for the demand.
Answer analysis
Feedback tied to the answer on the page
INSiqT separates the signals that earn marks from the issues that suppress them, then turns both into concrete editing instructions.
Evidence and facts
Use of constitutional provisions, cases, data, examples, thinkers, reports, or current context.
Structure
A direct introduction, logical body, useful subheadings, balance, and a conclusion that answers the question.
Visuals
Relevance, correctness, labels, and whether diagrams or tables make the answer clearer.
Presentation
Legibility, density, keyword visibility, spacing, and how quickly an examiner can find value.
Risk flags
Unsupported claims, generic material, missed directives, weak conclusions, and uncertain transcription.
What you get for each question
- • Estimated marks and a defensible range
- • What worked, quoted or referenced from your response
- • What lost marks and the likely reason
- • The highest-impact change for the next attempt
- • A model answer where your selected tier includes one
Built for comparison, not false precision
A single AI mark should never be treated as ground truth. INSiqT keeps the rubric stable and exposes warnings so the useful question is: are your answer skills improving across papers?