How Do Raters Judge Spoken Vocabulary?

The aim of the study was to investigate how raters come to their decisions when judging spoken vocabulary. Segmental rating was introduced to quantify raters' decision-making process. It is hoped that this simulated study brings fresh insight to future methodological considerations with spoken data. Twenty trainee raters assessed five Chinese students' monologic texts on vocabulary in this study. Both segmental rating and overall rating were retrieved from the raters. Rasch analysis suggested variation between raters in their judgment of vocabulary, although consistency was found in general. Besides, there was a mismatch between candidates' vocabulary scores and their lexical statistics. The raters' decision-making process was generally cumulative.