| This study was designed to investigate linguistic variables which greatly affect the reading difficulty of non-textbook reading materials for Arab alien adults and to develop a method by which the reading difficulty of these reading materials written in English by Arab adults who have English as a second language can be identified.;A total of 21 passages were selected from Iraqi non-textbook reading materials that were written in English and published in 1980, 1981, and 1982. These passages were assumed to cover different levels of reading difficulty. Rankings of the 21 passages according to their relative reading difficulty were obtained from 37 Arab adults selected from the State of Michigan. The passages were analyzed by using 16 linguistic variables which can be categorized into quantitative and qualitative variables. Product-Moment Correlation, Spearman's Coefficient of Rank Correlation, the Correlation of Concordance, Factor Analysis, and Multiple Regression were employed for presenting and analyzing obtained data.;Intercorrelations among the 16 linguistic variables were submitted to Factor Analysis to reduce data to a smaller set of linguistic variables. After rotation by Varimax method, these factors which were described as word, sentence structure, and grammatical complexities, were found to account for approximately 54.57 percent of total variance. The variable having the highest factor loading was selected to represent each of the three factors resulting from factor analysis. The selected variables were dependent clause length in words, word length in letters, and number of personal pronouns. These variables made significant contributions to determine the difficulty of reading materials when they were submitted to stepwise regression analysis which yielded a readability formula with the three variables. The formula was as follows: Y = .88222X(,13) + 7.11842X(,7) + .29391X(,6) - 35.34012.;The investigator concluded that the reading difficulty of non-textbook reading materials is a function of dependent clause length in words, word length in letters, and the number of personal pronouns. As the three predictor linguistic variables increase, the reading difficulty level increases. |
