Abstract
A preliminary experiment was performed to discover what effect the topic of speech and repetition of recording session would have on the number of types in samples of spoken monologues. Ten normal subjects were recorded on three occasions, and on each occasion they spoke on three different topics for five minutes each, though two of the topics were constant across occasions. Despite the fact that the subjects felt they spoke more easily on one topic, that of their own choice, there was no evidence of a significant topic effect or an order effect upon measures closely related to the Type-Token Ratio. The conclusion was that sampling speech in this fashion is a suitable design to study changes in Type-Token Ratios where change over time due to specific causes (e.g. psychiatric illness, exhibition of drugs, etc.) is being investigated.
Get full access to this article
View all access options for this article.
