The process of reading involves not only word recognition but also text-level interpretation (higher-level processing) through inferencing in the active construction of meaning from text. Lower primary school students focus on developing their ability to decode words and to comprehend simple text, but students in the higher Year levels use their reading skills together with their cognitive skills and past experience to further their information. If students are to be motivated to comprehend text effectively, they must have developed, and must have control of, appropriate cognitive strategies. They must also be able to monitor and plan their reading comprehension progress (metacognitive skills).
Comprehension monitoring and comprehension are closely related: they are dependent on each other for optimal functioning. This relationship between the two skills reflects the active essential processes involved in constructing meaning. As one reads, one is formulating hypotheses that are continually being modified and elaborated by the integration of further information. This constructive, inferential process can provide information about how well a text is understood.
The readers' responses to the text become more inferential, more critical, and more constructive as their reading skills develop. If students are to be academically successful, they need to use these higher-level skills. Students' abilities to generate and evaluate inferences are very much involved in their academic performances, and as such must be carefully monitored.
The research presented will describe the development of an inferential comprehension test for students in their third year of school through to their seventh (ages approx. 8 to 12). The initial stage involved the construction of the materials. The decision to use a developmental approach meant that each Year level required its own test. Consequently, five subtests were prepared. Areas of consideration included the format, topic selection, item selection, and readability level.
The development of this test will be traced from the pilot study, through various changes and more in-depth studies to the final version, which was standardised on 2,522 primary school students. The results of the initial analyses, using classical test analyses procedures, had several major drawbacks. The main concern was that the results for these analyses might be true for this particular sample on those particular tests, but might vary according to the level of proficiency of any further samples. Also, the developmental progression of the five test levels could not be plotted using these analyses. Exploration of the literature led to the realisation that this problem could be addressed and overcome by Rasch measurement. The results of the analyses indicated that this Test of Reading for Inferential Comprehension meets the requirements of the Rasch model. Reliability and validity were also very strong. The inclusion of a diagnostic graph allows the results to be easily interpreted, providing a useful clinical tool.
The test was then used to evaluate the inferential skills of students with hearing impairment. The results indicated that this test was useful in determining the inferential difficulties of these students. The results of this study, then, indicate the potential of the test for clarifying the nature of inferential skills in special populations. The final chapter discusses the future research for using the test, as well as consideration of inferencing skills of students.