Educational justice. Reliability and consistency of large language models for automated essay scoring and its implications. JALT. 2025;8(1):67-77. doi:10.37074/jalt.2025.8.1.21