Comparing Human and AI Rater Effects Using the Many-Facet Rasch Model
It addresses the need for reliable automated scoring in low-stakes educational assessments, though it is incremental as it evaluates existing LLMs on new data.
This study compared ten large language models (LLMs) with human expert raters in scoring writing tasks, finding that ChatGPT 4o, Gemini 1.5 Pro, and Claude 3.5 Sonnet achieved high scoring accuracy, better reliability, and fewer rater effects.
Large language models (LLMs) have been widely explored for automated scoring in low-stakes assessment to facilitate learning and instruction. Empirical evidence related to which LLM produces the most reliable scores and induces least rater effects needs to be collected before the use of LLMs for automated scoring in practice. This study compared ten LLMs (ChatGPT 3.5, ChatGPT 4, ChatGPT 4o, OpenAI o1, Claude 3.5 Sonnet, Gemini 1.5, Gemini 1.5 Pro, Gemini 2.0, as well as DeepSeek V3, and DeepSeek R1) with human expert raters in scoring two types of writing tasks. The accuracy of the holistic and analytic scores from LLMs compared with human raters was evaluated in terms of Quadratic Weighted Kappa. Intra-rater consistency across prompts was compared in terms of Cronbach Alpha. Rater effects of LLMs were evaluated and compared with human raters using the Many-Facet Rasch model. The results in general supported the use of ChatGPT 4o, Gemini 1.5 Pro, and Claude 3.5 Sonnet with high scoring accuracy, better rater reliability, and less rater effects.