Posted at Nov 21/2005 12:54 AM by Peter Y:
The lone difficulty in working through the study was in maintaining consistency and logic in the application of parts of speech. Far too often, words walked the razor edge between two different parts of speech, and subtle shadings of context were all that separated them from one pile or the other. Deciding when a word is a verb or an adjective, an adverb or a conjunction, is often too subjective.
To better deal with this subjectivity, larger sample sizes (some 8400 words all together) were used in the hopes that the tricky instances would even out over time.
Posted at Nov 21/2005 12:57 AM by Peter Y:
Indeed, the stable nature of the study--due to its fixed and unvarying subject matter--allowed the test protocal to be implemented with few other major difficulties. The biggest challenge was finding the time to assess and process the vast number of words selected for the study. The amount of time necessary to examine well over 8,000 words one-at-a-time, by hand was greater than imagined.