Answers are grounded in this paper and cite their sources.
On two different occasions twelve judges rated each of 30 short filmed speeches on each of 10 scales of speaking performance. The retest reliability of each judge on each scale was calculated. Median retest reliability of the twelve judges was highly correlated with years of training and experience in the teaching of speech. There was a tendency for global characteristics to be rated more reliably than specific observable characteristics.
No takes yet. Share an insight, caveat, or question.
Theodore Clevenger (1963) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: