{
  "id": 6591,
  "url": "https://arxiv.org/abs/2603.28553v2",
  "title": "Multimodal Analytics of Cybersecurity Crisis Preparation Exercises: What Predicts Success?",
  "summary": "Instructional alignment, the match between intended cognition and enacted activity, is central to effective instruction but hard to operationalize at scale. We examine alignment in cybersecurity simulations using multimodal traces from 23 teams (76 students) across five exercise sessions. Study 1 codes objectives and team emails with Bloom's taxonomy and models the completion of key exercise tasks with generalized linear mixed models. Alignment, defined as the discrepancy between required and en",
  "authors": "Conrad Borchers, Valdemar Švábenský, Sandesh K. Kafle, Kevin K. Tang, Jan Vykopal",
  "category": "research",
  "topics": "safety-alignment,children-education",
  "orgs": null,
  "regions": null,
  "published_at": "2026-03-30T15:14:56.000Z",
  "fetched_at": "2026-07-14T16:32:37.308Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/6591",
  "original_url": "https://arxiv.org/abs/2603.28553v2",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}