Pronunciation learners can benefit from peer feedback in a
Computer-mediated Communication (CMC) environment that allows them to notice
segmentals and suprasegmentals. This paper explores the intelligibility
judgments of same-L1 peers using P-Check (Version2, https://ver2.jp), a
Learning Management System (LMS) plug-in that aggregates peer feedback on
local intelligibility (Munro & Derwing, 2015). P-Check randomly delivers
written prompts for learners to record. Recordings are randomly delivered to
peers who choose from a drop-down menu which utterance was perceived.
Aggregated judgments from peers and from the instructor are displayed to
learners as feedback on intelligibility. This study used eight segmental
contrasts: /b-v/, /s-θ/, /l-ɹ/, /l-ɹ/-clusters, /æ-ʌ/, /ɑ-ʌ/, /ɑ-oʊ/, and
/i-ɪ/. Participants (N=38) made 3,451 intelligibility judgments on 1,203
recordings. The effects of rater listening discrimination proficiency and of
utterance intelligibility were examined in six contrasts using Generalized
Estimating Equations (GEE). Results showed that intelligibility was
generally a significant predictor of judgment accuracy, but rater listening
discrimination proficiency was not.