Data for: Comparing Raw AI, Context-Enhanced AI, and Human-in-the-Loop Translation in Multimodal Energy and Power Engineering Communication

This replication dataset supports the study examining raw AI translation, context-enhanced AI translation, and human-in-the-loop translation in multimodal energy and power engineering communication. The dataset includes anonymized evaluation scores for 27 translation units assessed by six evaluators, translation-unit metadata, the scoring rubric and codebook, AI configuration and prompt documentation, randomization and adjudication records, and statistical analysis code. Materials subject to corporate non-disclosure agreements or third-party copyright restrictions are not reproduced. Such materials are represented through anonymized identifiers and non-identifying metadata. No personally identifiable information is included.