The Authenticity Gap in Human Evaluation
This addresses a critical methodological flaw in NLG evaluation, impacting researchers and practitioners who rely on human ratings for system comparisons.
The paper identifies that standard human evaluation protocols in NLG often fail to capture true annotator preferences due to violated assumptions, particularly with Likert scales, and proposes improvements and a new protocol (SPA) for open-ended tasks, showing that SPA recovers GPT-3 size ordering with statistical significance while the standard protocol recovers less than half of expected preferences.
Human ratings are the gold standard in NLG evaluation. The standard protocol is to collect ratings of generated text, average across annotators, and rank NLG systems by their average scores. However, little consideration has been given as to whether this approach faithfully captures human preferences. Analyzing this standard protocol through the lens of utility theory in economics, we identify the implicit assumptions it makes about annotators. These assumptions are often violated in practice, in which case annotator ratings cease to reflect their preferences. The most egregious violations come from using Likert scales, which provably reverse the direction of the true preference in certain cases. We suggest improvements to the standard protocol to make it more theoretically sound, but even in its improved form, it cannot be used to evaluate open-ended tasks like story generation. For the latter, we propose a new human evaluation protocol called $\textit{system-level probabilistic assessment}$ (SPA). When human evaluation of stories is done with SPA, we can recover the ordering of GPT-3 models by size, with statistically significant results. However, when human evaluation is done with the standard protocol, less than half of the expected preferences can be recovered (e.g., there is no significant difference between $\texttt{curie}$ and $\texttt{davinci}$, despite using a highly powered test).