versión On-line ISSN 2076-3433
S. Afr. j. educ. vol.28 no.4 Pretoria nov. 2008
The development of an enabling self-administered questionnaire for enhancing reading teachers' professional pedagogical insights
Senior Lecturer in the Education and Social Science Faculty at the Cape Peninsula University of Technology, Mowbray Campus. Her research and teaching interests are literacy pedagogy and inclusive education in teacher education. E-mail: email@example.com
I explore the development and refinement, validation and implementation of a questionnaire to define teachers' perceived professional competencies in teaching reading skills and strategies in Grades 1-7 classes in developing countries. Using the Concentrated Language Encounter (CLE) programme, implemented and expanded annually between 2001 and 2005 in 4 900 new classrooms in schools in South Africa, I gathered and analysed theoretically coherent feedback data from more than 1 000 qualified, active reading teachers to establish a set of competencies describing teachers' professional understandings of their pedagogical reading tasks. The study was grounded in the social constructivist, sociolinguistic and psycholinguistic theories originating from the works of Piaget, Vygotsky, Cambourne, and Goodman. Their foundational principles, together with the South African Revised National Curriculum Statement were defined and applied to the derivation of all items in the questionnaire. The questionnaire evolved through three phases of validation. Throughout phases two and three, several cautious varimax normalized factor analyses and scree plots were engaged to refine and develop the questionnaire, within the context of teaching reading in South African schools. The emerging teaching reading themes can be fed back to teachers to improve aspects of their teaching reading.
Keywords: cautious varimax normalized factor analyses; competencies; Concentrated Language Encounter; multilingual; pedagogical; scree plot; social constructivist; sociolinguistic and psycho-linguistic; validation
In this article I describe the development, validation and implementation of a profile of professional competencies, designated the "Core Indicators of an Effective Reading Teacher" (CIERTQ), from 1999-2003. During this period there have been numerous national and provincial studies of children's literacy levels but none has determined the teacher's own understandings of the core indicators of an effective reading teacher.
In December 2005 the South African national Grade 6 Intermediate Phase systemic evaluation report stated that 42.1% of all Grade 6 learners were working and achieving on grade literacy level. Ten years into democratic governance, the Grade 6 learner assessment confirmed that the legacy of apartheid has impacted deeply on the educational achievements of the children in our country. The Progress in International Reading Literacy Study (PIRLS 2007) report summarised the findings from South Africa's participation which took place in South Africa between 2004 and 2007. The research assessed the reading skills of Grade 4 learners, where a total of 40 countries and 45 education departments participated. South Africa achieved the lowest score of all the 45 education departments.
Van der Berg (2004) reports that inputs such as more money, more teachers, more teaching materials and learning aids and smaller class sizes are not crucial factors in school performance. The most powerful factors determining how well a school functions are the quality of the principal, the level of teacher commitment and the extent of community involvement. He mentions that, "The only solution must be an African one." Walker, Rattanavich & Oller (1992:2) argue that the reason why so many children fail to read has more to do with what goes on in school than with what the children bring to school.
It was precisely this expressed viewpoint that stimulated the origin of this work the acknowledgement or assumption that literacy achievement depends crucially on the nature and quality of instruction that learners receive in the classroom. In other words, why is it that some teachers achieve high literacy rates with their learners while others, working in the same school and same environment, do not?
While involved with a percussive study from 1998 to 2001, I was surprised to discover how many under-qualified teachers there were in the Foundation and Intermediate Phases of primary schools (Grades R to 7) who clearly had a limited understanding of current primary school reading instructions, approaches and practices. It was for this reason that in 1999, in a separate parallel study (Donald, Condy & Forrester, 2003:484) I decided to focus on using a well-researched literacy programme with a sound theoretical background, called the Concentrated Language Encounter (CLE), to improve the teaching of pedagogical skills in teachers working particularly in developing schools in South Africa.
In the present study I adopted and modified several methodological techniques from an existing empirical approach to research data gathering which was used by Gilder, Irwin-Carruthers & Kent (1985). Through the training of physiotherapists they developed professional skills by identifying a profile or matrix of their perceived core tasks and key competencies. They designed and re-evaluated a comprehensive list of competencies derived by consensus for professional physical therapy practice.
The Concentrated Language Encounter Programme (CLE)
CLE is a reading programme where children learn language mainly through encounters with others in which the children concentrate intensely on making themselves understood (Walker & Rattanavich, 1992:12). Walker & Rattanavich (1992:12) elaborate by stating that:
The CLE approach was developed to deal with the educational needs of Aboriginal children. Its general principles are applicable for all students, whether advantaged or disadvantaged, whether first or second language learners, or whether living in urban or rural situations.
Between the 1970s and 1980s, Walker (1992) and Gray (in Rattanavich & Christie, 1993) initiated a new literacy programme for Australian aboriginal children in Australia where they were experiencing gross literacy failure. The theoretical basis for the literacy programme, which became known as Concentrated Language Encounter (CLE), was grounded in the assumptions that language and learning were conceived in the wrong order.
Earlier approaches identified and taught literacy content and skills that children needed to learn in isolated and disconnected ways. Christie (2000) comments that, during the nineteenth century the subject English had been concerned with grammar, spelling, composition and literary studies. In such traditional approaches, the first focus was on teaching content and skills; context was used only later if either approach failed, or a check was needed for word identification.
In South Africa, traditional approaches to teaching literacy in both the mother tongue and as a second language had always viewed language in terms of many small discreet units which comprised syntax and parsing, phonics, spelling, grammar and punctuation. These skills were taught and often committed to memory. The theoretical basis of such an approach involved the idea that students would learn various discreet units of language in isolated ways, often by repetition and drill, and that language learning developed by progressing from smaller to larger units. Only later in their development would students understand how these discreet units could be put together to create a coherent language structure. The problem with these theories was that they tended to divorce language from considerations of purpose and meaning. The role of language in learning and language development had never been considered. Figure 1 portrays Gray's interpretation of a traditional approach to language teaching.
When working with the Aboriginal children, Gray and his colleagues reversed the accustomed order for curriculum planning and concentrated on the context for learning and upon the development of language abilities within that context. Gray (in Rattanavich & Christie, 1993) argues that proficient readers use syntactic and semantic cues to interrogate texts and, through these skills, language learning develops to ultimately produce effective, literate learners. Gray (in Rattanavich & Christie, 1993), Christie (1989) and Cambourne (2002) believe that language should be learnt by using text, through coherent sentences in which meaning is salient and negotiated in many ways. This process is described in Figure 2.
Working within this general model of a language curriculum, Gray and Walker planned for the development of teaching and learning contexts in which teachers and students negotiated the nature of the learning activity as well as the nature of the language needed for that activity. Teachers would scaffold appropriate models of language, while students would be fully engaged in developing and using the appropriate language. The students required "real life" opportunities to practice "real life" language in a supportive environment. They would be encouraged to take risks in order to become better learners.
Christie (2000) discusses a model of language development taken from Halliday (1989) which states that language development involves: learning language (where learning language is a basic resource in listening, speaking, reading and writing), and learning through language (which refers to the ability to use language to learn, build relationships and express information about one's world). In 1977, borrowing a term from Cazden (1996), Gray and Walker initiated the term "concentrated language encounter" in which a number of role plays and language games and activities would be generated, including talking, reading and writing about things together, in which both teachers and children would take an active part.
The CIERTQ instrument evolved progressively through three phases of validation with phases two and three forming the major part of this study.
First, the preliminary improvements in the questionnaire developed through seven formative versions as it passed through successive pilot trials with different small groups of self-selected reading teachers (teachers from Grades 2-7, principals, subject advisors, learning support teachers, final-year teacher training students and lecturers) from 1999 to early 2002 while they attended the CLE workshops. To improve the content validity of the instrument, versions 3, 4, 5, and 6 were given to professional teachers and they were asked to verify whether the wording and general layout of the CIERTQ was in accordance with the Benchmarks for Literacy for Grades R-4 (1999).
During this phase the questionnaire involved further development and refinement of the CIERTQ. In 2001 and 2002, I trained 60 Western Cape Education Department (WCED) officials selected from seven Education Management Development Centres (EMDC), and they in turn selected the schools to be trained in CLE and then trained those teachers.
Teachers in Phase 2 attended a one-day CLE workshop. Version 8 was administered to 533 reading teachers in early 2002. It was re-administered to 360 teachers six months later. One hundred and seventy-three of the pre- tested teachers were present at both the pre-and repeat workshops. After qualitative and quantitative analysis of the generated data, version 8 of the CIERTQ was improved and version 9 was reformulated in readiness for another large-scale trial.
Phase 3 was the final administration of the CIERTQ, version 9, to a new relevant self-selected study group of 144 reading teachers who were attending the 2003 National Professional Diploma in Education course in teaching Literacy in the primary schools, offered in the Education Faculty of the Cape Technikon, Western Cape.
Throughout Phases 2 and 3 several cautious varimax normalised factor analyses were engaged to refine and develop the questionnaire, within the context of teaching reading in economically developing schools in South Africa.
Version 10 was developed but not trialled.
The results of this study will be discussed according to the three phases that it progressed through, although more emphasis will be given to the last phase.
As a result of the workshops, four literacy issues recurred both qualitatively and quantitatively throughout the seven trials up to the end of 2001. Three educational issues in literacy emerged and these focused on clarifying the structural parameters in the CIERTQ itself as a research instrument for producing data. These surfacing trends were classified as technical, organizational, and scoring issues. A fourth issue emerged during interviews which raised a concern regarding the personal support needed for teachers of literacy.
Some of the technical issues referred to all items on the CIERTQ indicated typical reading skills generic to Grades 1-7. The wording of all items was appropriate, their expression was unambiguous, unbiased, simple and clear in meaning, with no redundancies, no spelling errors, and written in the correct tense.
The organizational suggestions implemented the following critical recommended requirements: that the instruction page of the CIERTQ had to be clearer, have simple language, and include an example of how to complete the form. Seven categories of descriptors were recommended, developed and classified by teacher agreement, and items were grouped accordingly by consensus within each category.
The scoring recommendations made by some respondents suggested that it should be clear, effective, unambiguous and simple. The scoring scale was developed from a two-point scale [yes/no] to a recommended five-point scale for the final draft with scores ranging from 0 [no idea] to 4 [essential] (Fraser, 1989).
During interviews it became evident that teachers were adamant that they needed support when implementing or attempting a new methodology for the teaching of reading.
By November 2001, values of Cronbach alpha reliability coefficients for the seven desired features of version 7 of the CIERTQ, comprising 42 items, had been derived. To obtain these coefficients, the scores of a convenient sub-participant group of 45 trained teachers were used (although only 15 supplied intact data). The seven Cronbach alpha coefficients obtained were: = 0.67 for reading for meaning (6 items), = 0.85 for reading to communicate (six items), = 0.86 for reading to learn (six items), = 0.79 for reading to develop critical thinkers (six items), = 0.70 for reading to be sociable (six items), = 0.75 for reading for pleasure (six items), and = 0.80 for reading to develop skills (six items).
Therefore, for this particular convenient participant group, version 7 of the comprehensive questionnaire had satisfactory reliabilities for the seven inclusive categories.
By the end of Phase 1 of the pilot study, the sectors of re-grouped items began to emerge with a better systemic framework for the questionnaire in a re-arranged form of the CIERTQ one which was more logical, defensible and clear. The qualitative and quantitative data complemented and supported each other in ways that gave thrust to the unfolding of the next phase of development of the CIERTQ.
Using version 8 of the CIERTQ (with seven sectors), 173 teachers completed both the pre-questionnaire (early 2002) and the repeat questionnaire (late 2002). A cautious factor analysis was then conducted on the repeat programme intervention responses. The varimax normalized rotation indicated that there were possibly eight factors in the responses of the 173 respondents, instead of the seven previously provisionally found with the 533 respondents' pre-questionnaire responses, prior to the CLE workshop. The 42 items were then regrouped more logically and consistently, guided by the provisional factor analysis. They were partially reclassified to formulate version 9, with improved wording of all items. To clarify the meaning and interpretation of the main headings generated for each sector, three colleagues who lecture with me were asked to give suggestions for a reformulated version of the questionnaire.
The reformulated version of the questionnaire (version 9) was then promulgated for further trial and refinement on two separate occasions during workshops, namely, in May 2003 with 105 respondents and on August 2003 with 39 qualified teachers (Grades 1-7). These teachers had chosen voluntarily to extend their professional development in Education by participating in further informal studies after normal class hours. All participant teachers had also attended at least 10-12 hours of additional voluntary workshop training in the CLE approach to reading and writing. The following seven research questions were asked throughout the study:
1. What were the sources of items?
The main source of the items in the CIERTQ came from the current government policy documents, interviews and workshops with expert teachers and recent reading theory.
2. How many sectors and items in each version of the CIERTQ?
The answer to this question is provided in Table 1 which shows the progressive development and refinement of sectors and items from version 8 (in Phase 2) to version 9 and 10 (in Phase 3).
Only one item was discarded as a consequence of the refinement of version 8. In the developments from version 8 to version 10, the total number of items remained almost stable at 41 or 42, despite the continuing trials and an internal reshuffling of items occurring after each pilot study.
3. How important were the items?
A means and standard deviation analysis of the respondents' responses was completed. Items which scored a mean of at least 2.0 ("relevant"), up to a maximum possible score of 4.0 ("essential") by the teachers of reading have been designated as "core" indicators, in agreement with the criteria embedded in current reading theories of reading pedagogy (Walker, 1992; Halliday, 1996; Cambourne, 2004; Goodman, 2005).
With regard to the responses to version 9 of the CIERTQ, the highest scoring items (the maximum possible score being 4.00) were those numbered 3, 32, and 41:
• The best teachers of reading are those that: "Allow time for their learners to retell the story in their own words" (item 32; mean = 3.74).
• The best teachers of reading are those that: "Use correct grammatical structures when speaking to their learners" (item 3; mean = 3.72).
• The best teachers of reading are those that: "Allow their learners time to respond to what has been said" (item 41; mean = 3.72).
The lowest scoring items were items 14 and 22:
• The best teachers of reading are those that: "Make time to talk about the author's styles of writing" (item 14; mean = 2.72).
• The best teachers of reading are those that: "Use graphic aids such as graphs and tables to interpret information in a text" (item 22; mean = 2.81).
4. Were the numbers of items revised?
Continuous qualitative feedback on the content and expression of items in version 8 was obtained from 126 professional colleagues who supplied written and verbal critiques of the original wording of each item in the questionnaire.
In version 9, four items did not change their wording at all, six items' wording remained very similar and 31 items were reworded. One item was eliminated. Quantitative feedback was obtained through cautious factor analyses of the responses of 144 teachers to version 9 of the presented scale items. A cautious factor analysis identified that eight items expressed with compounded wording were individually identified. Each of these over-elaborated items was therefore subsequently divided into simpler, shorter, revised items with more consistent wording.
For example: In version 9, item 24 read: "Encourage their learners to read with appropriate rate and expression".
However, for version 10, the wording of this item was altered to read: "Encourage their learners to read with expression" which became item 33. Any further reference to the word "rate" was omitted.
5. Do the emerging clusters or themes of items become more scattered or more cohesive as these subsequent versions are clarified and improved?
For a refined version 9 with 41 items a cautious factor analysis was performed on the response data. Using the more linguistically refined 41 items of version 9, the three prominent factors with factor loadings > 0.50 that appeared to be emerging were as follows:
• Factor 1 now includes "reading for understanding and application", "serendipitous items" and "reading for strategy development" as the most prominent dimensions; and the "serendipitous items" appear to be items of support. In short, all nine items (variables) strongly associated with factor 1 could be interpreted and described as pedagogically strategic items.
• Factor 2 (four items) appears more concerned with "reading for meaning and interpretation"; and
• Factor 3 (five items) appears to be concerned with "reading for socializing", but is still heterogeneous in its composition, and may be considered as representing "research and surveillance".
Also, whereas the unrefined versions 8 of the CIERTQ yielded only three or four items with factor loadings > 0.70, the more refined version 9 has now produced nine items with factor loadings > 0.70. This result implies that the construct validity of the CIERTQ is improving, as the wording of its items becomes progressively refined.
These 41 items' measured associations are depicted on a scree graph in Figure 3, illustrating the extent to which they have now tended to form small groups of identifiable and significant clusters along a straight line. What has emerged is a discernable sense of structure in the teachers' responses that has a natural coherence in harmony with current theory.
6. What were the qualitative reflections made by two reading expert education officials of versions 9 and 10 of the CIERTQ?
After the final factor analyses, I interviewed two literacy experts (education officials) together for their evaluation of the proposed composition and wording of version 10 of the CIERTQ. At that time, they both worked in developing and developed schools in the Western Cape and were both leaders of literacy in their regions. They agreed that the sectors in version 10 of the CIERTQ fairly reflected the learning outcomes advocated in the National Curriculum Statement (2002).
In addition they suggested that the designated subheading, Reading for research and surveillance, be reworded as Reading for scanning and research. They felt that a teacher's scanning and researching are similar and equally important tasks, but with one being a simpler task and the other being a more complex task.
As a result of the renaming, item 41, "Allow time for their learners to skim and scan the text to find relevant information", should be included in the sector called Reading for scanning and research.
The experts made favourable comments about the sector named Reading for reflection and analysis. One stated: "I like it; I like it because, at the end of a book, you have to reflect on what you've read and make your own analysis."
One interviewer wondered whether inserting an open-ended prompt question at the beginning of the survey would elicit an initial, helpful, instinctive response to the teaching of reading. This suggestion was taken up and is now offered on the cover page of the version 10 of the questionnaire in the appendix.
The suggested changes to the items and sectors were accepted as offered, and they were all incorporated into the wording and structure of final version 10 of the CIERTQ.
7. What was the structure and composition of the final version 10 of the CIERTQ?
After a series of factor analyses and interviews with literacy experts, revisions, and re-arrangements the CIERTQ has evolved and developed to become the final proposed version 10 (see Appendix). This version has emerged with seven modified or identified sectors with more appropriate items in each sector, but has not yet been trialled.
This study is a unique contribution to issues around the teaching of reading. It has shown that it is possible to identify, clarify and verify a profile of core indicators of an effective reading teacher that is comprehensive, and consistent with the current policy documents and in terms of current theoretical studies on the teaching of reading.
The CIERTQ has wide reliability and validity for primary schools that operate within multilingual reading cultures and diverse reading approaches, particularly in economically developing regions of South Africa and possibly beyond.
It is interesting, from a methodological point of view, that it has taken nine versions of the CIERTQ to produce a stable and coherent provisionally final set of professional competencies. However, the most interesting aspect of this methodology shows that the use of large repeat samples has led to a better defined and well-integrated description of the component sectors in the emerging profile. Therefore a strong methodological feature of this research has been the clear beneficial outcomes of an increasingly well-defined product.
When introducing a new literacy approach to a province there will always be teething problems, and the expert education officials were very aware of the issues that were impeding the implementation of CLE into schools. At the same time they are very positive about the programme and its benefits to both the teachers and the learners.
Cambourne B 2002. The conditions of learning: is learning natural? The Reading Teacher, 55:758-762. [ Links ]
Cambourne B 2004. Holistic, integrated approaches to reading and language arts instruction; the constructivist framework of an instructional theory. In: Farstrup AE & Samuels SJ (eds). What research has to say about reading instruction. USA: The International Reading Association. [ Links ]
Cazden CB 1996. A report on reports: Two dilemmas of genre teaching. In: Christie F & Foley J (eds). Some contemporary themes in literacy research. Münster: Waxman Publishing. [ Links ]
Christie F 1989. Language education. Oxford: Oxford University Press. [ Links ]
Christie F (ed.) 2000. Pedagogy and the shaping of consciousness: linguistic and social processes. London: Continuum. [ Links ]
Donald D, Condy J & Forrester J 2003. Literacy development: lessons learned from a pilot project in South Africa. The Reading Teacher, 56:484-492. [ Links ]
Fraser BJ 1989. Assessing and improving classroom environment. What Research says to the Science and Mathematics Teacher, 2:1-7. [ Links ]
Gilder N, Irwin-Carruthers S & Kent A 1985. Competency profile of the professional physiotherapist. Unpublished mimeograph, Cape Town: University of Cape Town. [ Links ]
Goodman K 2005. What is whole in whole language. Canada: RDR Books. [ Links ]
Halliday MAK & Hasan R 1989. Language, context, and text: aspects of language in a social-semiotic perspective. Hong Kong: Oxford University Press. [ Links ]
Halliday MAK 1996. Literacy and linguistics: a functional perspective. In: Hasan R & Williams G (eds). Literacy in society. New York: Addison Wesley Longman. [ Links ]
Howie S, Venter E, Van Staden S, Zimmerman L, Long C, Scherman V & Archer E 2007. PIRLS 2006 Summary report: South African children's reading achievement. Pretoria: University of Pretoria. [ Links ]
Piaget J 1969. Judgment and reasoning in the child. London: Routledge & Kegan Paul. [ Links ]
Rattanavich S & Christie F 1993. Investigating text-based approaches to the teaching of literacy in Thailand. In: Goyne G & Purves AC (eds). International Association of Research in Mother Tongue Education: Mother Tongue Education Research Series. New York: Waxman Verlag GmbH. [ Links ]
South Africa National Department of Education 1999. Benchmarks: Literacy for Grades R - 4. Internally circulated unpublished mimeograph. Pretoria: Government Printer. [ Links ]
South Africa National Department of Education 2002. Revised National Curriculum Statement Grades R to 9 (Schools) Policy Document: Home Language. Pretoria: Government Printer. [ Links ]
Van der Berg S 2004. Throwing money at education problem isn't enough. Available at http://www.sundaytimes.co.za/articles/TarkArticle.aspx?ID=946527. Accessed July 4 2004. [ Links ]
Walker RF 1992. Developing a CLE primary school literacy programme. In: Walker R, Rattanavich S & Oller W (eds). Teaching all the children to read. Buckingham: Open University Press. [ Links ]
Walker RF 1992. General introduction. In: Walker R, Rattanavich S & Oller W (eds). Teaching all the children to read. Buckingham: Open University Press. [ Links ]
Walker RF 1992. Reflecting on the contexts of reading. In: Walker R, Rattanavich S & Oller W. Teaching all the children to read. Buckingham: Open University Press. [ Links ]
Walker RF & Rattanavich S 1992. Improving literacy teaching in developing countries. In: Walker R, Rattanavich S & Oller W. Teaching all the children to read. Buckingham: Open University Press. [ Links ]
I appreciate the guidance, critical insights and unequivocal support of Prof Rochford (UCT), who was the supervisor for my PhD, and who was tragically killed earlier this year.