Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 ISSN: 2548-5865 83 ASSESSMENT TO IMPROVE READING COMPREHENSION Tri Setianingsih (1) (Triset66@gmail.com (1) ) Faculty of Education for Language and Art (FPBS) Mataram Institute of Teacher Training and Education (IKIP) ABSTRACT Reading comprehension is important skill to acquire information and knowledge from any kind sources. Difficulties in reading comprehension for students in learning English as EFL/ESL has forced teacher to apply appropriate principles and strategies. One strategy that believed can help to solve the problem in reading comprehension is assessment. Four Roles Model can be adapted for assessment and used as a framework to address the complex nature of reading comprehension difficulties by considering the reader as: (a) a code-breaker, (b) a text-participator, (c) a text-user, and (d) a text-analyser. Breaking the code emphasised decoding of the words, and encoding of information, understanding the conventions of written, spoken, and visual multimodal texts by recognising and using the surface features of print. There are two kinds table of assessment. Table 1 presents a flexible reading comprehension assessment matrix showing the four roles of the reader with the before, during and after reading phases. Table 2 shows the elements within the matrix can be adjusted to reflect appropriate items suited for different stages of reading development or with a particular focus in mind. Keywords: Assessment, Reading Comprehension INTRODUCTION There are many environmental and personal factors that contribute to reading success. Reading comprehension is a complex interaction of language, sensory perception, memory, and motivational aspects. However, most existing assessment tools have not adequately reflected the complex nature of reading comprehension. Good assessment requires a multifaceted approach to reading diagnosis and flexible interventions in order to cater for individual learning needs. Hence, reading comprehension is a such important skill that students need to acquire some informations even though knowledge from any kind sources such as text, newspapers, and magazines. This discussion investigates the educational issues for the assessment of students with reading comprehension difficulties and suggests appropriate principles and strategies that teachers can apply to inform assessment and teaching practice that is suitable for Indonesian students in learning English as EFL/ESL. Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 84 DISCUSSION Assessment Educators generally agree that assessment is fundamental to the effective teaching of reading and to the design of individualised reading intervention programs (Clay, 1992; Coccamise & Snyder, 2005). However, Joshi and Aaron (2000) claimed that most assessment procedures currently being used in the schools today are based on limited theoretical models of reading and tend to give the impression that all comprehension difficulties are merely found within the reader. Generally speaking, these assessment procedures have sprung from specific theories that are concerned either with a single aspect of reading, such as word decoding, or are focused on global aspects, such as overall cognitive ability (Freebody & Frieburg, 2001; Joshi & Aaron, 2000). The reality is that reading successfully requires a complex interaction of language, sensory perception, memory, and motivation (Pikulski & Chard, 2005). Thus, a number of researchers have called for better assessment models of comprehension and appropriate intervention programs to reflect this complex process (Pressley, 2002a; Schunk, 2004). Furthermore, it is claimed that students with learning difficulties can make greater progress when instructional interventions are multifaceted by combining a number of approaches (Hay, Elias, & Booker, 2005; National Reading Panel, 2000). In contrast, Paris and Oka (1989) claimed that most existing standardised comprehension tests are inappropriate to assess the possible comprehension benefits of teaching students to use multiple reading strategies. To overcome these limitations educators may need a range of assessment strategies and instruments that robustly reflect the dynamic, developmental nature of comprehension within the reader and with his or her interactions with other external dimensions such as activity, text, and context (Duke & Pearson, 2002; Gillet & Temple, 1994; Snow, 2003). Changing Views Of Reading The Simple View of Reading held that word reading ability and listening comprehension account for nearly all of the variance in reading comprehension (Gough & Tunmer, 1986). Normally word reading and reading comprehension are highly correlated and one reason for the less- skilled comprehenders' initial failure may be that they focus more on word reading accuracy rather than comprehension monitoring (Cain &Oakhill, 1999). Furthermore, when teachers focus on word-level processing skills as a single indicator of reading performance the focus may be too restricted and may lead to an Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 85 inadequate assessment of reading comprehension difficulties (Bishop & Snowling, 2004). It is thought that unless a student is able to read words fluently, heavy demands are made on working memory during a slow and tedious decoding process that requires the reader to use focused attention to identify each succeeding word (Spencer & Hay, 1998). A number of other researchers have also shown that there is strong association between speed of word reading and text comprehension (Hay, Elias, & Booker, 2005; Jenkins, Fuchs, van den Broek, Espin, & Deno, 2003). According to Joshi and Aaron (2000) a revised model referred to as the Component Model was later proposed to account for the significant effect of fluency on reading comprehension. It was shown that adding speed of processing to the Simple View of Reading significantly improved prediction of reading comprehension. Fluency not only involves efficient decoding of words, but in order for reading comprehension to progress effectively, the reader must focus attention on making meaning while using automatic processes for word recognition. To a large degree, fluency will be affected by the quality of prior experiences and knowledge structures that children apply to read text information (Reutzel, Camberwell, & Smith, 2002). Moreover, faster rates of word recognition would directly affect comprehension and enhance the chunking of information into meaningful information units in working memory by enabling the expansion and elaboration of existing knowledge structures (Jenkins et al., 2003; Pikulski & Chard, 2005). Therefore, a comprehensive assessment of fluency must not only include measures of oral reading accuracy and rate of oral reading but also the quality of oral reading. This is particularly important for older children, as there is evidence to suggest that fluency contributes relatively more to comprehension at higher levels of reading development (Jenkins et al., 2003; Pikulski & Chard, 2005). It is vital that fluency is assessed in relation to reading for understanding but there are a range of other factors that may need to be considered when selecting suitable assessment tools. The Use of A Single Direct Measure Neale Analysis of Reading Ability (NARA: Neale, 1988) is an example of a decontextualised or direct measure of reading accuracy, comprehension and reading rate. Westwood (2003) maintained that the test is generally highly regarded and used in most Australian schools by regular teachers and special education teachers to assess and identify students with reading and comprehension difficulties. The test is a measure of reading accuracy, reading rate and reading comprehension and is comprised of a battery of short narratives with graded readability levels. While undertaking Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 86 this test, a student would be required to read a series of stories and orally answer eight comprehension questions for each passage. There have been a number of comparisons conducted to verify the adequacy of this test, for example, Graves, Fitzgerrald, Miller, and Pillay (2002) found that the reading ages derived from the NARA, in most cases, were almost identical to the spelling ages derived from the South Australian Spelling Test (Westwood, 2005). Hatcher and Hume, (1999) found that Verbal IQ (which is often dependent on vocabulary subtests) is also correlated highly with NARA reading comprehension. No direct assessment tool is perfect, and awareness of the strengths and limitations of each instrument will guide the educator's selection of the most appropriate testing tool and interpretation of the scores (Cain & Oakhill, 2006a). For example, Spooner, Baddeley, and Gathercole suggested that the comprehension component of the NARA was less reliable than the reading accuracy measure. One reason for this was that the researchers maintained that reading comprehension and word accuracy were strongly interrelated and could not be easily separated. However, one of the obvious strengths of the NARA is that this is not as problematic as other tests because misread words are corrected during the reading. Cain and Oakhill (2006a) suggested that a more reliable measure of reading ability would be to use the NARA accuracy scores in conjunction with a separate test for reading comprehension such as the TORCH (Mossenson, Hill, & Masters, 1987). Other researchers claimed that the NARA comprehension score was doubtful because the passages were read orally rather than silently (Graves et al., 2002). Ehri and McCormick (1998) maintained that progress in reading beyond the early stages is dependent on oral language development. This is because text comprehension draws on a broad range of different language skills-these include lower-level lexical skills, such as word reading efficiency and vocabulary knowledge, sentence-level skills, such as knowledge of grammatical structure, and higher-level text processing skills, such as inference generation and comprehension monitoring (Cain & Oakhill, 2006b). There is considerable evidence that difficulties in reading comprehension are often accompanied by inadequate oral language (Leach, Scarborough, & Rescorla, 2003). For example, a number of researchers have identified that word reading, vocabulary, and grammatical awareness are related to reading comprehension (Cain & Oakhill, 2006b; Nation & Norbury, 2005). Hay, Elias, and Booker (2005) also found that many students in the middle and upper school with reading difficulties had problems comprehending text passages because they couldn't identify and process the information contained in phrases, sentences and relationships between Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 87 sentences. Furthermore, Cain and Oakhill (2006a) maintained the assessment of readers with expressive language difficulties would present greater problems on the NARA because they would have greater difficulty in producing oral responses for answers to questions at the end of each passage. To overcome this confusion, some researchers recommend the assessment of both reading and listening comprehension using parallel measures to distinguish children who experience language- related comprehension difficulties from children whose problems are caused primarily by word decoding difficulties (Gunning, 2006). CONCLUSION In the past, reading comprehension assessment was reliant upon tools that were designed around simple literacy models that focused on a narrow set of skills. However, there needs to be a broader understanding of comprehension that goes beyond viewing reading problems as being solely within the learner. Reading comprehension is complex and multifaceted, no single instrument, by itself, will provide the necessary information to guide the design of appropriate individualised teaching interventions for struggling readers. The problem is that formal, decontextualised instruments tend to be limited in focus and don't give enough direction for suitable teaching practices. Teacher-designed instruments are more informative but less reliable because they vary with content, test conditions, and assessor variables. However, there is a broad consensus that teachers can ensure quality practice by incorporating a range of contextual and direct assessment instruments and observations. What is certain is that strategies and instruments should robustly reflect the dynamic, developmental nature of comprehension to include other external dimensions such as activity, text, and context. Teacher-designed informal reading inventories can supplement standardised tests to broaden the focus and to provide more relevant information. This requires teacher judgement. However, there is a danger that reliability may suffer without a consistent assessment framework. The Four Roles model of literacy is an example of a literacy framework that can provide some structure to give teachers direction for assessment choices. Such an organisational arrangement should provide a theoretical framework to give consistency without restricting the assessor's ability to make informed decisions related to the various reader roles and strategies. To be effective, this framework will need to be ongoing and have a clear purpose. It should also be sensitive to the reader's stage of reading development and consider the before, during, and after reading phases. Where possible, assessment should be dynamic and on going and should actively involve the reader in making Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 88 choices and allow for metacognitive decisions to be articulated while reading. Feedback from such activity should inform teachers as to the motivational and self-regulatory reading behaviours of the children they are attempting to assist. Whereas, different countries have different strategies to assist their students even though to improve their reading comprehension. In Indonesia which is English is not their mother tongue or English is still as EFL/ESL, students is still poor with their vocabularies and found many strange words, they need to translate the difficult words into bahasa Indonesia by themselves by opening the dictionaries or the teachers use both languages Indonesia and English in giving the instruction and helping children to get the meaning of the text. If meaning is not the problem anymore, to comprehend the text is easier and students will be faster to improve their reading comprehension. Journal of English Language Teaching Volume 3 Nomor 2, Agustus 2016 http://ojs.ikipmataram.ac.id/index.php/joelt ISSN: 2548-5865 89 REFERENCES Brown, D. (2003). Language Assessment Principles and Classroom Practices. USA: Longman. CIERA. (2001). Improving the Reading Comprehension of America’s Children. Michigan: University of Michigan. Duke, N.K., & Pearson, P.D. (2007). Effective Practices for Developing Reading Comprehension. London: Blackwell. Gunning, T.G. (1992). Creating Reading for All Children. Boston: Allyn and Bacon. Harris, A.J., & Sipay, E.R. (1980). How to Increase Reading Ability: A Guide to Developmental and remedial Methods. New York: Longman. Harris, L., & Smith, C.B. (1986). Reading Instruction, Diagnostic teaching in the Classroom. New York: Macmillan Publishing Company. Rasinky, T.V. (2005). Assessing Reading Fluency. Pasific Resources for Education and Learning. Honolulu, Hawai’i. U.S.: Department of Education. Slavin, R.E., & Cheung, A. (2005). Effective Reading Programs for English Language Learners and Other Language-Minority Students. Bilingual Research Journal. Wooley, G. (2008). The Assessment of Reading Comprehension Difficulties for Reading Intervention. Australian Journal of Learning Difficulties, Vol.13.