Options
Linguistic variables as predictors of Chinese text readability
Loading...
Type
Thesis
Author
Tham, Tuck Meng
Supervisor
Soh, Kay Cheng
Chew, Cheng Hai
Abstract
The objective of this study is to investigate the linguistic variables which influence readability of Chinese texts for Secondary school pupils in Singapore. Specifically, the study aims at answering the questions :
1. Which lexical factors - characters, words (content and function) will have greater predictive power of readability of the Chinese texts?
2. Which syntactical factors - sentence, sentence length, sentence complexity (simple sentence, compound sentence and complex sentence) will have greater predictive power of readability of the Chinese text?
The study was done through an analysis of twenty passages, ten selected from the currently used approved texts, the Chinese Instructional Materials for Secondary Schools (CLIMS) and ten from the de-listed texts which were used from 1979 to 1983. Each of the selected passages was content-analysed for indices of the linguistic variables which were subsequently used as predictors of text readability.
The same set of selected passages were given another treatment, that is, empirical evaluation. The passages were transformed into a series of twenty cloze tests which were then field-tested on 498 Secondary 1 and Secondary 2 pupils. Each pupil completed four cloze tests, two based on the CLIMS passages and two on the de-listed texts. The reliabilities of the twenty cloze tests were evaluated for internal consistency and the median of Kuder-Richardson reliabilities in .72. The validity of the cloze tests was ascertained by comparing passage rankings given by ten experienced Chinese Language teachers with mean cloze scores based on pupil performance. A product-moment of .94 indicates a high level of agreement, thus evidence the validity of the tests.
Finally, indices of linguistic variables obtained through content-analysis and data obtained from empirical evaluation (i.e. pupil performance in the cloze procedures) were correlated to ascertain the relationships between the linguistic variables and passage readability (as indicated by cloze procedures results).
The findings of the study is that the most predictive factor is syntactical or sentence factors. Sentences written in a more complex form and passages containing more sentences are more predictive of ease in reading, contrary to what has been found with English texts in the literature. Lexical or word factors turned out to be less predictive than were expected. This might be due to the fact that the selected texts are all from the textbooks which are either in use or had been once in use and hence the semantic factors have been taken care of when the texts were produced.
The findings of the present study suggest that writers of Chinese Language secondary school textbook need not be overly concerned with lexical factors. However, they need to pay more attention to syntactical factors. The findings further imply that in the teaching of Chinese Language at the secondary school level, the teacher needs to accord greater emphasis on the relationship between characters and words through ample use of clauses and sentences rather than paying too much attention to the teaching of individual characters or words.
1. Which lexical factors - characters, words (content and function) will have greater predictive power of readability of the Chinese texts?
2. Which syntactical factors - sentence, sentence length, sentence complexity (simple sentence, compound sentence and complex sentence) will have greater predictive power of readability of the Chinese text?
The study was done through an analysis of twenty passages, ten selected from the currently used approved texts, the Chinese Instructional Materials for Secondary Schools (CLIMS) and ten from the de-listed texts which were used from 1979 to 1983. Each of the selected passages was content-analysed for indices of the linguistic variables which were subsequently used as predictors of text readability.
The same set of selected passages were given another treatment, that is, empirical evaluation. The passages were transformed into a series of twenty cloze tests which were then field-tested on 498 Secondary 1 and Secondary 2 pupils. Each pupil completed four cloze tests, two based on the CLIMS passages and two on the de-listed texts. The reliabilities of the twenty cloze tests were evaluated for internal consistency and the median of Kuder-Richardson reliabilities in .72. The validity of the cloze tests was ascertained by comparing passage rankings given by ten experienced Chinese Language teachers with mean cloze scores based on pupil performance. A product-moment of .94 indicates a high level of agreement, thus evidence the validity of the tests.
Finally, indices of linguistic variables obtained through content-analysis and data obtained from empirical evaluation (i.e. pupil performance in the cloze procedures) were correlated to ascertain the relationships between the linguistic variables and passage readability (as indicated by cloze procedures results).
The findings of the study is that the most predictive factor is syntactical or sentence factors. Sentences written in a more complex form and passages containing more sentences are more predictive of ease in reading, contrary to what has been found with English texts in the literature. Lexical or word factors turned out to be less predictive than were expected. This might be due to the fact that the selected texts are all from the textbooks which are either in use or had been once in use and hence the semantic factors have been taken care of when the texts were produced.
The findings of the present study suggest that writers of Chinese Language secondary school textbook need not be overly concerned with lexical factors. However, they need to pay more attention to syntactical factors. The findings further imply that in the teaching of Chinese Language at the secondary school level, the teacher needs to accord greater emphasis on the relationship between characters and words through ample use of clauses and sentences rather than paying too much attention to the teaching of individual characters or words.
Date Issued
1987
Call Number
LB3045.8 Tha
Date Submitted
1987