Abstract: | Most current popular methods of reading-material analysis use "cluster" sampling—sampling by sentences. In this study, such sampling is compared with simple random sampling of individual words. No differences were found in the proportions of the various parts of speech, but the clustered sample "significantly over-estimated the percentage of 'short' words, 'structural' words, and 'easy' words. It is suggested that the structure of the sentence… has imposed an orderliness that has biased the clustered sample." (PsycINFO Database Record (c) 2010 APA, all rights reserved) |