THE IMPORTANCE OF STRUCTURED TEXT TALKS FOR STUDENTS' READING COMPREHENSION: AN INTERVENTION STUDY IN SPECIAL SCHOOLS

Abstract

The present intervention study reveals that students diagnosed with an intellectual disability (ID) are able to construct meaning from written expository text through guided social interaction. There were 31 students recruited from four special schools participating in this intervention study. The study involves a pre-test phase and a post-test phase. The students were divided into two intervention conditions: (a) reciprocal teaching (RT), which involved practice in four reading strategies-prediction, question generating, clarification, and summarisation-and (b) inference training (IT), which involved practice in answering inference questions, i.e., where you have to read between the lines to find the answer. The training included 16 sessions over 8 weeks. Pre- testing and post-testing included seven tests. Improvement of test results was obtained in both conditions to about the same extent, indicating that both conditions were beneficial.

(ProQuest: ... denotes non-US-ASCII text omitted.)

...

Abstract

The present intervention study reveals that students diagnosed with an intellectual disability (ID) are able to construct meaning from written expository text through guided social interaction. There were 31 students recruited from four special schools participating in this intervention study. The study involves a pre-test phase and a post-test phase. The students were divided into two intervention conditions: (a) reciprocal teaching (RT), which involved practice in four reading strategies-prediction, question generating, clarification, and summarisation-and (b) inference training (IT), which involved practice in answering inference questions, i.e., where you have to read between the lines to find the answer. The training included 16 sessions over 8 weeks. Pre- testing and post-testing included seven tests. Improvement of test results was obtained in both conditions to about the same extent, indicating that both conditions were beneficial.

Keywords: structured text talk, reciprocal teach- ing, inference training, reading comprehension, intellectual disability

Introduction

Great efforts have been made to improve the life quality and social inclusion of students with intellectual disability (ID). One important aspect of the social inclusion and quality of life is literacy (1, 2).Nevertheless, literacy has not been taken into account to any great extent to empower students with ID in Sweden. As demonstrated in a Swedish report, efforts to promote reading comprehension instruction have been limited. Instead,the social, emotional, and esthetical dimensions of education have had the highest priority (3). The report also critiqued that many teachers in special schools lacked appropriate qualifications such as formal training in special education.

Research has demonstrated that reading in general and text comprehension in particular have been identified as challenging skills for students with ID, who usually do not achieve the level of their chronological and educational peers(4, 5). However, there is reason to believe that their cognitive potential is often seriously underestimated and that students with an ID may have the unexpected capacity to understand written text if they are given proper stimulation and instruction (2, 6, 7). Some of these researchers have used different instructional programmes in reading comprehension built within a sociocultural paradigm (8-10).

Literature review

Two instructional programmes in reading comprehension have proven to be successful for readers with ID. The two programmes are reciprocal teaching (11) and inference training (12, 13).

Reciprocal teaching(11) is a widely used technique for systematic text talks for promoting reading comprehension. Reciprocal teaching (RT) refers to an instructional activity that takes place in the form of a dialogue between teacher and students regarding segments of text. The teacher and students take turns assuming the role of teacher in leading this dialogue. RT is thus a group activity in which students read a passage of an expository text paragraph by paragraph. During the reading, the students acquire and practice four reading comprehension strategies:

1. Prediction occurs when students hypothesise what the author will bring up next in the text. In order to do this successfully, students must activate the relevant background knowledge that they already possess regarding the topic. Furthermore, the opportunity has been created for the students to link the new knowledge they will encounter in the text with the knowledge they already possess.

2. Generating questions. When students initiate questions, they first identify the kind of information that is significant enough to provide the substance for a question. Then, they pose this information in question form and self-test to make sure that they can indeed answer their own questions. This is an important strategy for active reading. Many students with comprehension difficulties read texts in a passive way; they are not aware that they, as readers, are expected to question the author of the text they are reading.

3. Clarifying. When the students are asked to clarify, their attention is called to the fact that there may be many reasons why a text is difficult to understand (e.g. new vocabulary, unclear reference words, and unfamiliar and perhaps difficult concepts). This strategy fosters monitoring of comprehension breakdowns and the use of strategies such as selective search for relevant content and taking the necessary measures to restore meaning (e.g., reread, ask for help).

4. Summarising provides the opportunity to identify and integrate the most important information in the text. A text can be summarised across sentences, paragraphs, and the passage as a whole.

The rationale for RT was developed within the sociocultural paradigm. The four strategies are examples of the kinds of cognitive activity in which successful learners engage while interacting with texts (14). Furthermore, RT is based on three theoretical principles that were prominent in the work of Vygotsky (15). The first principle is that the origins of all higher cognitive processes are first social; that is, that mental functioning occurs first between people in social interactions. The second one is the zone of proximal development, i.e., the distance between the actual developmental level and the level of potential development under adult guidance or in collaboration with more capable peers. The third one is that psychological processes are acquired in contextualised, holistic activity; that is, the strategies are not broken into component skills, nor are they practiced in isolation (14).

Since the introduction of the RT programme (11), numerous studies have been conducted to examine its efficacy. These studies have revealed improvement in students' abilities to summarise, generate questions, clarify, and predict (16-18). However, surprisingly little research has been conducted to extend RT to persons with ID. Exceptions are van den Bos, Nakken, Nicolay, and van Houten (8) and Alfassi, Weiss, and Lifshitz (9).

Inference training. Raphael (12, 13) was the main source of inspiration for the programme of inference teaching.When students have difficulties answering questions it is often assumed that it is because they have not read the text carefully. However, it may also depend on the fact that they need to be taught how to analyse a question in order to find the correct answers. In Inference training (IT), the students are asked to answer four different types of questions related to brief texts. The Question- Answer Relationship (QAR) method presents a three-way relationship between questions, text content, and reader's knowledge.

The QAR method divides questions into two broad categories; "In the Book" (text-explicit) questions and "In My Head" (text-implicit) questions. "In the Book" questions fall into two subcategories: "Right There"questions and "Think and Search" questions. In "Right There Questions" the answer is easily found in the text since the exact words for the questions and answers are located in the same sentence. In the "Think and Search" questions, the answer is in the text but requires gathering information from different sections in the selection. Thus, the task is to integrate different sentences in the passage using text-connecting inferences. "In My Head" questions fall into two subcategories: "Author and You" questions and "On My Own Questions".In "Author and You" questions, the answer is just implicitly stated in the text. Thus, the reader has to combine previous knowledge with text information to create a response. These types of questions, mostly "why- questions" and "how-questions", require more complex inferences. For "On My Own Questions", the answer is not in the text. The reader uses previous knowledge to answer the question. Guided practice is important in IT.Most of the students don't even know what it means to think in their own head. Consequently, the teacher's role is to model how to do by using thinking aloud so students can "see" the teacher's thinking process (19). In IT, the text is not segmented.

In summary, the aims of the present study were to compare and evaluate two models of structured text talks, Reciprocal teaching (RT) and Inference training (IT), as possible interventions for students with intellectual disability.

RT was selected as "the treatment condition" and IT as the "control condition". RT and IT were chosen because the method of reading instruction (i.e., the focus on strategies) is the only aspect that distinguishes the two programmes and is thus manipulated. There, IT was used as a "placebo". Thus, the design ensured that the teacher could not know which condition was the treatment and which was the control.Accordingly, the study avoids the "Hawthorne effect", namely that teachers modify their behavior simple because they are observed. By introducing a placebo the idea is that the control condition may constitute an equivalent to the sugar pill in medical research. Furthermore, the idea is that the present study adds to previous research by confirming previous results about reading interventions. Although making discoveries by exploratory research is more exciting, confirmatory research is needed to produce cumulative knowledge, i.e., scientific progression (20).

Method

Participants

There were 31 participants recruited from four special schools in a region in western Sweden. Two of the schools were located in immigrant- dense suburbs of a big city. The third school was located in a suburb where the population had predominantly white-collar jobs, and the fourth school was located in the inner city, where the population also had white-collar jobs (21).

The participants consisted of 21 boys and 10 girls aged from 12 to 16 years (m = 13.7). All of them were diagnosed as intellectually disabled, and all had additional diagnoses, such as cerebral palsy (CP), autism, autism spectrum disorders (ASDs), attention deficit hyperactivity disorder (ADHD), and CATCH-22 syndrome. Four of the students were regarded as moderately intellectually disabled and were enrolled in a special programme, whereas 27 were regarded as mildly intellectually disabled. For all of the students, the medical specialists had presumed that the intellectual and socio- linguistic disability would make it too hard for them to follow ordinary school instruction in the regular school system. Hence, special school placement was indicated.In this article, ID is defined in accordance with the American Association on Intellectual and Developmental Disabilities:

Intellectual disability is a disability characterised by significant limitations, both in intellectual functioning and in adaptive behaviour, which covers many everyday social and practical skills. This disability originates before the age of 18 (22).

In the present study, there were 13 students who had language backgrounds other than Swedish. The selection of participants was primarily based on the interest of teachers who informally announced their willingness to be included in the study. The reason was that recruiting teachers to intervention studies is difficult because teachers argue that they simply do not have the time. There were seven teachers, two male and five female. All were educated teachers and three of them had also a university degree in special educational needs (23). In the special schools, the students from grade six and up are not divided into separate classes but they are all taught together in the same group. Before the study, the teaching of reading in the selected classes primarily included independent reading of self-selected books and materials with no explicit instructional ambition above stimulating the students to retell what they have read, sometimes accompanied by informal discussions. Some teachers also read stories aloud to the students.

Randomisation. The 31 students were divided into two almost equally-sized groups. For instructional purposes, the two main groups were divided into subgroups with two to four students in each subgroup. In total, there were 10 subgroups. The assignment of conditions (RT and IT) to the main groups was randomised.

IT was selected as a control condition for a proper evaluation of RT since the IT condition included many of the same elements as the RT condition. However, IT lacked emphasis on explicit comprehension strategies as discussed above.

Procedure

The study is a quantitative one and involves a pre-test phase, an intervention phase, and a post-test phase. In the next paragraph, the three phases of the intervention will be outlined.

Phase 1

Pre-testing.Reading ability was assessed before and after the intervention period for both groups of students.All tests were done in one-to-one condition in the participantsschools, involving one researcher and one student at a time. A Norwegian computer-based test battery, Logos (24), which also has a well-standardised Swedish version, was used. Logos is a diagnostic instrument in the sense that the basic underlying functional impairments in dyslexia can be revealed. In this study, the following seven tests from Logos were used:

1. Conceptual understanding. Here, the students were orally introduced to 22 different concepts, one at a time. The students were expected to explain the meaning of the concepts. The maximum score was 22. The results of correct answers were reported in percentages.

2. Fluency was assessed by oral reading of five texts. The average number of words read correctly per minute was used as the measure of reading fluency.

3. Listening comprehension. Five parallel texts of equal character, length, and level of difficulty were used for assessing listening comprehension. Tape-recorded texts were read to the students. After each section, there were three questions. The maximum score here was also 15 points. The results of correct answers were reported in percentages.

4. Oral reaction time. A dice with one or two dots was shown on the computer screen. Students were to respond as quickly as possible as to whether the dice had one dot or two dots. There were 15 assignments to complete, and the duration for each assignment was clocked.

5. RAN. Rapid automatized naming (RAN) of different objects. Faster times in RAN trials have been found to be a good indicator of reading competence. Different well-known categories of objects were used (25). In total, 48 objects were shown on the computer screen. The students were asked to name the objectsrapidly. The total time in seconds for naming objects was used as a measure of RAN.

6. Reading comprehension was assessed by three questions related to each passage used to assess fluency. One of the questions required inference. The maximum score on comprehension was 15 points. The results of correct answers were reported in percentages.

7. Word recognition. The ability of a reader to recognise, in all, 40 written words correctly.

The total testing time for the seven tests was 45 minutes. Although all variables are of interest, the focus dependent variables were fluency and reading comprehension. Reading compre- hension, conceptual understanding, and fluency are all indicators of the effect of the inter- vention on students' ability to understand writ- en texts. Furthermore, listening comprehension is theoretically interesting as an indicator of the effect of the intervention on the poor readers. Moreover, rapid automatized naming and word recognition are indicators of the effect of the intervention on phonological and decoding abilities. Although these measures are important in the context of reading research, the focus of the study was comprehension and understanding and not "the mechanical side" of reading.

There were no norms for students with ID. However, the norm set in the test is Grade 4 of compulsory school.

Phase 2

Intervention. Pre-testing was followed by 8 weeks of RT or IT followed by post-testing. There were two weekly 30-minute sessions of reading-comprehension instructions delivered to the subgroups one at a time. A total of 16 sessions of structured text talks were included in the intervention programme. The intervention was performed during autumn 2013.

Intervention texts.The rationale for selecting texts was that (a) they reflected students' interests and (b) they included a picture for both conditions. Consequently, the students were interviewed by the researcher prior to the intervention about their interests. The outcome of these interviews resulted in 16 authentic expository texts, each of them including a picture, selected from newspapers, and mostly from an easy-to-read newspaper titled 8 sidor ('8 pages'). 8 pages (26) address people with ID.

The role of the picturesin the text may be critical to students with limited experience in reading and whose second language is Swedish (27) since pictures can be used as tools to build background knowledge needed to understand the text (28). The same texts were used in the RT and IT conditions.

The intervention texts included a wide range of topics, including animals, food, sports (e.g., the Paralympics and Wimbledon), the Internet, and news (e.g., the catastrophe in India in which 40 students died after eating food served at the local school).As it can be seen from Table 1, the texts were rather brief, varying in length between 25 and 123 words and had short sentences. LIX value is also included. LIX (29) is a readability formula that was developed by Björnson (30). LIX = word length + sentence length where, word length = percentage of words of more than six letters; and sentence length = average number of words per sentence. According to LIX, values under 40 indicate that the text is easy to read.

Text talk procedure

At each session, a new text was read. As mentioned above, all 31 students read the same texts. In both conditions, the teachers started by asking the students to look at the picture and tell what it represented. Then, they were asked "What do think the text will be about?"

RT condition (Treatment). Only in RT was the text read segment by segment. Each student received one segment of text at a time. If the students had received all the text at once, we would not have been able to practice the strategy of prediction. Each segment was read aloud by the teacher. Then each student was asked to read one or two sentences. If a student could not decode the words, he or she read together with the teacher. One strategy was introduced at a time. At the beginning of each session, the strategies learned thus far were repeated. In order to help the students come up with their own questions, they were instructed to use question words like who, what, why, when, how. These words were also written on the whiteboard. After each segment, the students were instructed to take turns and pose one or more questions each to a peer in the group. When the question had been answered by the student, he or she posed a new question, and so it went until each student in the group had posed one or two questions. In order to facilitate summarisation, the students were instructed to use the adverbs first, then, and lastly. In the current study a strategy was useddeveloped in Lundberg and Reichenberg called 'new words'-to elaborate upon RT (10). Rather than asking, "Did you find any difficult words" the teachers were thus encouraged to ask, "Did you find any new words?" The difference between the two ways of posing questions is that in the first type of question there is an implicit performance expectation for the students. The result may be increased performance anxiety; students generally do not want to be evaluated in public during whole-class instruction. The second question does not raise any performance expectations but rather encourages students to be curious while reading.

IT condition (Control). In IT, the text was not segmented, thus, the teacher started by reading the whole text aloud. Then, he or she asked the students to read some sentences each. If a student could not decode the words, he or she read together with the teacher.All four question types were introduced during the first session.The teacher was modelling the strategies involved in answering the questions, and explained how he or she had found out the correct answer.This is an important part of the learning process since many students can be confused when they learn that a question could have two or sometimes three possible answers (31).The teaching of IT is built up around the instructional idea of a gradual release of responsibility to the students. In the beginning, the teacher had to model the strategies much more, but less and less later on (12).

Programme validation. The first three sessions in RT and IT were led by the first author of this article. She modelledthe activities, and through a gradual learning process, the students were quite capable of using most of the strategies (11). The teacher sat next to the researcher. Afterwards, the researcher discussed the session with the teacher, and the teacher was free to ask questions about what the researcher had done. When the teacher led the sessions, the researcher attended two lessons and monitored the teacher's practice. The teacher was then instructed to video-record lessons 6, 9, 12, and 15. This was done in order to check that the teachers used the method as they had been instructed.

Phase 3

Post-tests.After the intervention, there was a post-test. The same tests as on the pre-test were used.

Statistical procedure.The data from the seven tests (Logos) was analysed using a repeated measures ANOVA for differences over time (Time) and group differences over time (Time x Group). The analysis was conducted in IBM SPSS V.20.

An assumption of ANOVA is that the variables are fairly normally distributed. However, several dependent variables were positively skewed. Accordingly, the variables were transformed when needed using the log +1 transformation.

Another assumption of mixed-design ANOVA is sphericity, which tests for equal variance among groups. Mauchly's sphericity test revealed that the assumption of sphericity was violated. Consequently, the Greenhouse-Geisser correction was applied to correct for violations of sphericity. The reason is that the Greenhouse-Geisser correction is the most conservative form of correction.

To control for differences in the baseline between groups (IT and RT) at the beginning of the study, an independent samples test was conducted on the pre-test scores(32).

Table 2 summarises descriptive statistics for the used variables. Scores are both in totals and differentiated by condition. Reading compre- hension, listening comprehension and conceptual understanding have been converted into percentage scores.

Ethical considerations

The ethical considerations were addressed in the following ways. The researchers provided all students with both verbal and written information outlining the details of the research project. The participants were also informed that they were free to decline as well as interrupt their participation if they did not like it. Since the participants were under 18 years old, their parents were also informed. The parents' consent was secured via consent forms.

Results

In the results section, the pre- and post-test scores on the seven tests for the two instructional conditions will be presented.

Initial group differences

To test if the groups were comparable before intervention, an independent samples t- testwas conducted. The t-test revealed no significant group differences in the mean for oral reaction time (t=1.31, N.S.), conceptual understanding (t=0.33, N.S.), listening comprehension (t=1.93, N.S.), and reading comprehension (t=1.52, N.S.) before the intervention. However, the t-test revealed significant group differences in the mean for RAN (t=4.0, p<0.001), wordrecognition (t=3.05, p=0.005), and fluency (t=3.2, p=0.003) before the intervention.

Time. Table 3 presents means and standard errors on reading-related tests (Logos) for the two treatment groups, RT and IT, before and after intervention. The analysis of the results revealed a significant difference over the phases of the intervention on fluency (F=1,64.17, p<0.001, h= 0.69). The analysis of the results revealed a significant difference over the phases of the intervention on reading comprehension (F=1, 16.07, p<0.001,h= 0.36). The analysis of the results revealed a significant difference over the phases of the intervention on listening comprehension (F=1, 39.40, p<0.001,h= 0.65). The analysis of the results revealed a significant difference over the phases of the intervention on word recognition (F=1, 8.56, p<0.05,h = 0.23). The analysis of the results revealed no significant difference over the phases of the intervention on RAN (F=1, 2.91, N.S.,h= 0.09). The analysis of the results revealed a significant difference over the phases of the intervention on conceptual understanding (F=1, 26.00, p<0.001,h = 0.47). The analysis of the results revealed a significant difference over the phases of the intervention on oral reaction time (F=1, 6.04, p<0.05,h=0.17 ).

Group × Time. Turning to the Group × Time effects of the intervention, the analysis of the results revealed significant group difference of the intervention on word recognition (F=1, 4.50, p<0.05,h=0.13 ). However, the effect size was low. All other tests revealed no significant group differences. Furthermore, a re-run of the model was conducted to control for effects of student gender. However, no significant effects of student gender were found.

In summary, in both conditions, the test results were improved after the intervention but not very differently between the conditions. The analysis of variance only revealed one significant difference between groups in terms of treatment effects. However, the groups did not differ according to any of the variables related to the understanding of written texts. Nevertheless, all groups improved after the intervention on the focal dependent variables.

Discussion

The aims of the present study were to compare and evaluate two models of structured text talks, RT and IT, as possible interventions for students with ID. In both conditions, the students' reading comprehension, listening comprehension, and conceptual understanding increased significantly.

Thus, the findings indicate that even students with mild and moderate intellectual disabilities can be cognitively active and creative, as well as show commitment, and that they can enjoy reading, provided that they are introduced to texts that grab their interest and which are accompanied by structured teaching.

However, the pre-testing and post-testing of various reading-related skills did not yield significant differences between the conditions, except on one test. The students who practiced IT performed significantly better on the word recognition test. However, the effect size was small, and the groups differed significantly on the word recognition test before the interven- tion started. Accordingly, the only group difference in the ANOVA test is likely due to differences unrelated to the intervention. Consequently, the present study makes a contribution to previous research (10) by confirming previous findings. Thus, the study builds cumulative knowledge about reading interventions for students with ID.

One possible reason for the lack of differential effects of the two intervention conditions might be that the instructional procedures were too similar. The ambition was, of course, to create two conditions highly similar in all respects, including selection of participants, group size, number of sessions, session time, and texts, except for the critical variables of strategies and interaction pattern. Rather, the results can be interpreted as that the key factor seems to be active text talk. Accordingly, if students with ID are given regular opportunities to practice reasoning about either content or strategies in reading comprehension, then the results will be the same, independent of the reading programme. One limitation of the present study was the lack of a second comparison group with no intervention except their daily routines in the special school (i.e., no placebo). Such a base condition might have facilitated the comparison between pre- and post-test scores. However, the design reduces the likelihood of a Hawthorne effect. Another limitation is that the study did not statistically control for the theoretically important covariates such as intelligence, additional diagnosis, or motivation. Future research could benefit from adding such covariates to the design of intervention studies.

Moreover, future research can also benefit from controlling the effect of pictures. In the current study, each text talk started with showing the picture included in the text.Accordingly, pictures were treated as a constant for both conditions and not as a variable. As such, the pictures did not compromise the results. However, the role of picturesin the text may thus be critical to students with limited experience reading and Swedish as a first language, as pictures can help students to become more proficient creators of internal visual imagery that supports comprehension (28). Thus, mani- pulating pictures in conjunction with reading interventions could be one way to progress the research. Another way to progress research would be control for the effects of students with an immigrant background.

Conclusion

The findings indicate that structure in text talks is more important than the specific reading instruction programme-i.e., RT or IT-for reading comprehension.

Furthermore, the findings also suggest that reading comprehension instruction is efficient even in students with ID.Consequently, there are reasons to believe that the cognitive potential of students with ID is often seriously underestimated and that the students may have a capacity to understand written texts if they are given proper stimulation and instruction.

Conflict of interests

Authors declare that have no conflict of interests

.../References

1. Katims DS. Literacy instruction for people with mental retardation: Historical highlights and contemporary analysis. Education and Training in Mental Retardation and Developmental Disabilities 2000;35(1):3≠15.

2. Browder DM, Wakeman SY, Spooner F, Ahlgrim- Delzell L, Algozzine B. Research on reading instruction for individuals with significant cognitive disabilities. Exceptional Children 2006;72(4):392≠408.

3. Swedish School Inspectorate T. Undervisningen i svenska i grundsärskolan. 2010.

4. Ratz C, Lenhard W. Reading skills among students with intellectual disabilities. Research in Developmental Disabilities. 2013;34(5):1740≠1748.

5. Conners FA. Reading skills and cognitive abilities of individuals with mental retardation. International Review of Research in Mental Retardation 2003;27:191≠229.

6. Hudson ME, Browder D, Wakeman S. Helping students with moderate and severe intellectual disability access grade-level text. Teaching Exceptional Children 2013;45(3):14≠23.

7. Spooner F, Knight V, Browder D, Jimenez B, Di- Biase W. Evaluating evidence-based practice in teac- hing science content to students with severe deve- lopmental disabilities. Research and Practice for Per- sons with Severe Disabilities 2011;36(1-2):62≠75.

8. Van den Bos K, Nakken H, Nicolay P, Van Houten E. Adults with mild intellectual disabilities: can their reading comprehension ability be improved? Journal of Intellectual Disability Research 2007;51(11):835≠849.

9. Alfassi M, Weiss I, Lifshitz H. The efficacy of reciprocal teaching in fostering the reading literacy of students with intellectual disabilities. European Journal of Special Needs Education 2009;24(3):291≠305.

10. Lundberg I, Reichenberg M. Developing reading comprehension among students with mild intellectual disabilities: An intervention study. Scandinavian Journal of Educational Research 2013;57(1):89≠100.

11. Palincsar AS, Brown AL. Reciprocal teaching of comprehension-fostering and comprehension- monitoring activities. Cognition and Instruction 1984;1(2):117≠175.

12. Raphael TE, Au KH. QAR: Enhancing comprehen- sion and test taking across grades and content areas. The Reading Teacher 2005;59(3):206≠221.

13. Raphael TE. Question-answering strategies for children. The Reading Teacher 1982:186≠190.

14. Palincsar AS, Klenk L. Fostering literacy learning in supportive contexts. Journal of Learning Disabilities 1992;25(4):211≠225.

15. Vygotsky LS. Mind in society: The development of higher psychological processes: Harvard univer- sity press; 1980.

16. Rosenshine B, Meister C. Reciprocal teaching: A review of the research. Review of Educational Research 1994;64(4):479≠530.

17. Lederer JM. Reciprocal teaching of social studies in inclusive elementary classrooms. Journal of Learning Disabilities 2000;33(1):91≠106.

18. Alfassi M. Reading for meaning: The efficacy of reciprocal teaching in fostering reading compre- hension in high school students in remedial reading classes. American Educational Research Journal 1998;35(2):309≠332.

19. Mims PJ, Hudson ME, Browder DM. Using read-alouds of grade-level biographies and systematic prompting to promote comprehension for students with moderate and severe developmental disabilities. Focus on Autism and Other Developmental Disabilities 2012;27(2):67≠80.

20. Tukey JW. We need both exploratory and confirmatory. The American Statistician 1980;34(1):23≠25.

21. Mills CW. White collar: The American middle classes: Oxford University Press; 2002.

22. Jackson R. Mental Retardation: Definition, Classification and Systems of Support American Association on Mental Retardation. British Journal of Developmental Disabilities 2003;49(96):53≠54.

23. Ratz C. Do students with Down syndrome have a specific learning profile for reading? Research in Developmental Disabilities 2013;34(12):4504≠4514.

24. Høien T. Logos. Handbok. Diagnostisering av dyslexi och andra lässvårigheter. Bryne, Norway: Logometrica; 2007.

25. Wolf M, Denckla MB. RAN/RAS: Rapid automatized naming and rapid alternating stimulus tests: Pro-ed; 2005.

26. 8 Pages Stockholm: Centrum för lättläst. Available from: http://www.8sidor.se

27. Cummins J. Bilingualism and special education: Issues in assessment and pedagogy: Multilingual Matters Clevedon; 1984.

28. Hibbing AN, Rankin-Erickson JL. A picture is worth a thousand words: Using visual images to improve comprehension for middle school struggling readers. The Reading Teacher 2003:758≠770.

29. Lix Readability Formula : The Lasbarhetsindex Swedish Readability Formula. Available from: http://www.readabilityformulas.com/the-LIX- readability-formula.php.

30. Björnsson CH. Läsbarhet. Stockholm: Liber; 1968.

31. Fajardo I, Tavares G, Ávila V, Ferrer A. Towards text simplification for poor readers with intellectual disability: When do connectives enhance text cohesion? Research in Developmental Disabilities 2013;34(4):1267≠1279

32. Field A. Discovering statistics using IBM SPSS statistics: Sage; 2013.

Monica REICHENBERG

Department of Education and Special Education, University of Gothenburg, Sweden

Recived: 21.06.2014

Accepted: 09.08.2014

Original article

Correspondending address:

Monica REICHENBERG

Department of Education and Special Education

Box 300

405 30 Gothenburg, Sweden

E-mail: [email protected]

Telephone number 046-031-7862451,

Mobile: [phone]