Relaciones entre decodificación, conocimiento léxico-semántico e inferencias en niños de escolaridad primaria

Abstract

Reading comprehension is a complex process which lies upon three interrelated basic skills: decoding, lexical semantic knowledge and wordto- text integration processes. Decoding entails the reader's ability to recognize grapheme-phoneme correspondence. Lexical semantic knowledge refers to the knowledge of words in a given language, and it includes both word identification, i.e., visual processing of the graphemes' shapes and lexical activation; and semantic awareness, i.e., knowledge of meaning. Decoding and lexical semantic knowledge are deemed as pertaining to a lower level in relation to integration, in which category we include such skills as inferencegeneration. Previous research, notably in England, the United States of America and Argentina, has shown that difficulties in reading comprehension may be due to faulty decoding, poor lexical semantic knowledge and inferencemaking failures. Although each of these skills is necessary to achieve comprehension, none of them alone can guarantee it. It is considered that decoding and lexical knowledge are associated skills, and influence each other. The effects of decoding in text comprehension are mediated by knowledge about the meaning of words and the effects of compre hension in decoding depend on the achievement of sufficient word meanings. Thus, the aims of this paper were to study the relations among reading comprehension skills and to study how decoding skills and lexical semantic knowledge contribute to inference-making generation. In broad terms, the hypotheses were reading comprehension skills are interrelated and decoding skills and lexical knowledge are significant predictors of inference-making generation. This survey was conducted among 94 primary school children of an average age of 8 to 10 years old. They were given the Letter Recognition and Lexical Processes Tests from the PROLEC Test Battery, the Vocabulary Subtest from Wechsler Intelligence Scale for Children (WISC-III) and the Inference-making Ability Assessments from Cain and Oakhill (Canet Juric & Burin, 2006). Results have shown that in the decoding assessment, the highest number of errors and correct answers are found in pseudo word reading. Regarding lexical knowledge, the highest average of right answers is found in word reading. As regards wordto-text integration, the average number of in ferences incorrectly drawn is 14.62. Since the tests we used measured right and wrong answers, indices of decoding and lexical knowledge were created. Moreover, our findings suggest significant and moderate-to-strong relations among reading compre hension skills. Decoding skills and lexical semantic knowledge have accounted for 51% of inferences. However, lexical knowledge alone has made the most significant contribution. The set of variables has accounted for 48% of coherence inferences and 45% of elaborative inferences. The only significant predictor of both types of inference-making was lexical knowledge. To conclude, as posited by international research and our first hypothesis, comprehension skills are interrelated. Taking into consideration our second hypothesis, our findings suggest that compre hension skills account for inference-making generation and that lexical knowledge reveals itself as a predictor of the children's potential ability for inference-making. In other words, vocabulary is a basic component of more complex skills such as inference. This is consistent with what international research suggests. As regards the performance of the general sample in the assessed comprehension skills, it resembles that of the child population of the same age, surveyed in Tucumán and in. Taking into consideration the way in which comprehension skills relate to and affect each other could allow us to develop intervention programs which would help improve the performance of children who demonstrate text comprehension difficulties.