Skip to search boxSkip to navigationSkip to main content

Comparing rating scales and preference judgements in language evaluation

  • Anja Belz
    ,
  • Eric Kow
Research output:
Chapter in Book/Conference proceeding with ISSN or ISBN
Conference contribution with ISSN or ISBN
Research
Peer-review

About the research

Rating-scale evaluations are common in NLP, but are problematic for a range of reasons, e.g. they can be unintuitive for evaluators, inter-evaluator agreement and self-consistency tend to be low, and the parametric statistics commonly applied to the results are not generally considered appropriate for ordinal data. In this paper, we compare rating scales with an alternative evaluation paradigm, preference-strength judgement experiments (PJEs), where evaluators have the simpler task of deciding which of two texts is better in terms of a given quality criterion. We present three pairs of evaluation experiments assessing text fluency and clarity for different data sets, where one of each pair of experiments is a rating-scale experiment, and the other is a PJE. We find the PJE versions of the experiments have better evaluator self-consistency and inter-evaluator agreement, and a larger proportion of variation accounted for by system differences, resulting in a larger number of significant differences being found.

Documents and links

Research output information

Output type

Research output:
Chapter in Book/Conference proceeding with ISSN or ISBN
Conference contribution with ISSN or ISBN
Research
Peer-review

Original language

English

Pages from-to (Number of pages)

Pages 7-15 (9 pages)

Publication milestones

  • Published - 01/01/2010

Publication status

Published - 01/01/2010

Place of publication

Stroudsburg, PA, USA

Publisher

Association for Computational Linguistics

Host publication title

Proceedings of the 6th International Language Generation Congerence (INLG'10)