ETS uses both human raters and automated scoring methods to offer a complete and accurate picture of a test taker's ability. While automated scoring models have advantages, they do not measure the effectiveness of the language response and the appropriateness of its content. Human raters are needed to attend to a wider variety of features, such as the quality of ideas and content as well as form.
Additionally, studies have shown that prompts designed for fully automated scoring have been more vulnerable to prompt-specific preparation and memorized responses.
The TOEFL test uses automated scoring to complement human scoring for the two tasks in the Writing section. Combining human judgment for content and meaning, and automated scoring for linguistic features, ensures consistent, quality scores