ERIC Number: ED022644
Record Type: RIE
Publication Date: 1968-Jun-1
Reference Count: N/A
A Taxonomy of Cloze Research, Part I: Readability and Reading Comprehension.
Potter, Thomas C.
General information about the cloze technique and summaries of experiments with adults and children in which it was used are presented. Results indicated that the most valid and reliable cloze test for measuring passage difficulty is one in which (1) an every nth mechanical mutilation system is used, (2) not more than 20 words in every 100 are deleted, (3) passage length is at least 250 words, (4) at least 50 words are deleted in order to insure adequate sampling of passages, and (5) the exact word deleted is indicated as the most useful and efficient scoring criteria. It was also indicated (1) that deletion ratios of 1.10 and 1.12 in longer passages may be valid for certain purposes, (2) that scoring systems other than the exact word (synonym, form class) provide less interscorer reliability and require more time, and (3) that the separate scoring of form classes or content and function words may provide specific information for specialized purposes. Further research is recommended. A bibliography is included. (BS)
Descriptors: Classification, Cloze Procedure, Communication (Thought Transfer), Measurement Techniques, Readability, Reading Comprehension, Reading Research, Test Reliability, Test Validity
Southwest Regional Laboratory for Educational Research and Development, 11300 La Cienega Blvd., Inglewood, Calif. 90304.
Publication Type: N/A
Education Level: N/A
Sponsor: Office of Education (DHEW), Washington, DC. Bureau of Research.
Authoring Institution: Southwest Regional Educational Lab., Inglewood, CA.