Making connections: the nature and occurrence of links in literacy teaching and learning.

Parr, Judy M. ; McNaughton, Stuart

Introduction

The importance of making connections in the context of learning, for example from known concepts to those to be acquired, from the literacy of the home to that of the school, is well documented. As key literacy processes, reading and writing are seen to have sizeable complementary elements and the potential to be mutually supportive. But, research provides little insight into the nature and extent of teacher actions in making links within the literacy classroom, specifically those between reading and writing, and between and among texts. The paper explores this issue in primary classrooms. The idea of connectedness in reading and writing can be viewed in different ways: writing related to reading; writer related to reader and text related to text (Nelson, 1998) or the nature and extent of the relationship between reading and writing can be examined empirically, theoretically and pedagogically. The latter is the approach taken to review existing work to explore the nature of the connection and explanations for why strengthening connections might be advantageous.

Research that systematically collects data to answer questions has examined the relationship between the comprehension and the production of written texts. Reading and writing achievement have been found consistently to relate to one another statistically to about the same extent that brothers and sisters share characteristics (e.g. Juel, 1988; Tierney & Shanahan, 1991). Such studies have been criticised for the lack of validity of the measures in actually measuring comprehension or composition but, more importantly, because the measures being compared do not focus on the same underlying processes and skills; there is no guidance from a unified theory of language processing (Parodi, 2007). An effort to match the kinds of questions in the comprehension tests and the parameters on which the written products were evaluated, using theory proposed by Van Dijk and Kintsch (1983), produced results that showed a similar difference between each structural level of language (micro, macro and super-structural) in both comprehension and production. There were positive and significant correlations between comprehension and production at all levels (Parodi, 2007). More specifically, a five year longitudinal investigation of reading-writing relationships across multiple levels of language found reciprocal word-level spelling and reading relationships early in schooling, while text level comprehension helped to predict text-level composition in grades 2 to 6 (Abbott, Berninger & Fayol, 2010). Over time, as students developed, the only significant relationships were between word reading and text composing and spelling and text comprehension. These data support the proposition that reading and writing may be related differently as they change with development (Fitzgerald & Shanahan, 2000).

These relational studies, shortcomings aside, suggest that writing and reading are related, but often mask the fact that they are interactive at some level (Shanahan & Lomax, 1986), while a bidirectional hypothesis assumes that reading and writing may also be independent (Eisterhold, 1991). At the level of the individual, one can be a good writer without being a good reader and vice versa, according to Langer (1986) who sees the different cognitive starting points as precluding a closer alignment.

The explanation for these relationships is hindered, as noted, by a lack of a unified theory of language processing and their significance in terms of learning is largely unexplained. Examining theory allows some inferences as to how and why connections might be important. From a theoretical viewpoint, reading and writing both involve constructive processes. Depending on the theoretical perspective (and it is very likely that teachers draw on both in their actions), these could be seen as happening through a process of socially shared cognition (for example, where, in sharing writing, peers point out the places where they have difficulty understanding and why) that results in the individual internalising this knowledge or through a process of becoming a member of a sustained community of practice (where students, by engaging in the craft of writing within a community like a classroom learn by various means including apprenticeship, modelling and feedback, the 'discourse' of schooled writing) (Wenger, 1998).

Constructive processes for the individual are similar in reading and writing so performance in one area may strengthen performance in the other, particularly as it may involve operating in a similar but still somewhat different context. Readers plan reading around a purpose and activate prior background knowledge. Writers go through similar processes; they have a purpose and think about what they know or need to know to accomplish this purpose. Readers construct their own meaning from interpreting cues in text; writers construct meaning while composing so as to convey their desired understandings to a reader. In composing writers develop their ideas in much the same way that readers do when they reread something and, in doing so, amend the meaning they have made of the text. Notably, in specifying cognitive processes in his 1996 model of the writing process, Hayes included text interpretation and reflection as well as production. In text interpretation, internal representations are created from linguistic and graphic input while reflection involves creating new internal representations from existing ones. Reading is seen by Hayes as a central process in writing and includes reading to evaluate but also reading of sources to inform writing. Knowledge acquired in one may support performance in the other. Linguistic and genre knowledge and schema, stored in long term memory, are available for both writing and reading. The notion that such discourse knowledge, thought to be largely acquired from reading, is important in writing has been supported by research (Olinghouse & Graham, 2009). So, writing involves reading and both reading and writing involve similar processes like monitoring one's own meaning making and knowledge. Logically, articulating and strengthening links between them should enhance both learners' awareness of how language works to achieve its communicative purpose and of the processes involved in reading and writing.

Pedagogically, articulating and building connections provides students with knowledge to use strategically in both composing and comprehending. The theoretical traditions of reading and writing pedagogy have different roots. Writing initially drew on the Aristotelian rhetorical tradition while reading drew on physiological and psychological paradigms. Reading and writing have received overlapping attention, pedagogically, only since the 1970's (Langer & Allington, 1992; Langer & Flihan, 2000). Historically, some developments drew reading and writing together (like the attention to cognition), while others pushed them apart, the notion of centripetal and centrifugal forces operating (Nelson & Calfee, 1998). With the building of multi-disciplinary knowledge and through integrative movements such as comprehension as construction, reader response or whole language, interconnectedness has been fostered (Nelson & Calfee, 1998).

A number of instructional approaches and programs in one area, have been shown (incidentally and by design) to promote improved performance in the other. For example, Reading Next (Biancarosa & Snow, 2004) identified intensive writing as a key element of an effective adolescent literacy program as it improves reading, reinforcing an earlier assertion that when reading and writing are taught together, they improve achievement (Tierney & Shanahan, 1991). More specifically, a limited literature, largely aimed at practitioners, discusses notions like enabling the borrowing of ideas by students from literature as they develop their craft (Lancia, 1997) and the positive results from such (Corden, 2007). This professional literature also discusses the modelling, demonstrating and using of texts by teachers to highlight features of the craft of writing (e.g. Harwayne, 1992; Olness, 2005) or the fact that literature can provide a model whereby through listening, reading and conversation students learn about text at both a structural level and at a language use level (Rosemary & Roskos, 2002). Evidence that instruction in one supports the other comes from a meta-analytic review by Graham and Hebert (2010). They present evidence of three major instructional ways in which writing has been shown to improve reading, namely, (i) having students write about the texts they read through response, summaries, note-making and answering questions; (ii) teaching students the writing skills and processes that go into creating text like the process of writing, text structures, paragraph or sentence construction skills or teaching spelling, and (iii) increasing how much students write.

However, a consideration of the few empirical studies that investigate the extent to which explicit or recognisable connections are made between reading and writing, and between or among texts, suggests that, in normal classroom conditions, teachers do not readily articulate these links. Duncan (1999) observed a sample of New Zealand new entrant teachers (teachers of students starting school) over a six week period. These teachers, while espousing the importance of teaching reading and writing together, made virtually no explicit connections for the students in their classrooms. Similarly, Soter and colleagues (2008) concluded, from their analysis of discourse features across small group discussion approaches in US elementary classrooms, that the incidence of 'extra-textual' connections, as they termed the affective, inter-textual and shared knowledge connection, was 'barely apparent' (p. 383). In the pedagogical literature there is little evidence of the nature and extent of connections made between reading and writing, between readers and writers or between texts. Although there is consensus that reading and writing are linked and that they have been shown to be mutually facilitative in the development of literacy abilities and the learning of content, there are no developed models in the literature for using this interconnectedness in ways that demonstrably foster teaching and learning (Close, Hull & Langer, 2005). We do not have pedagogical models for using the relationship in productive ways; we lack frameworks to see how to translate the relationship into practice. The current study presents, as a starting point, a framework for considering the pedagogical moves of teachers in making links in literacy learning settings, in particular those made between reading and writing.

Method

The initial work was conceptual and focussed on developing a framework as a lens through which to view the nature and site of links between reading and writing. This framework was tested through use in classroom observations, then a revised framework applied to a corpus comprising transcriptions of both writing and reading lessons to examine utility and explore normal classroom practice.

The framework

Through a review of literature, the draft framework was generated. It was tested by a group of literacy leaders in a cluster of schools interested in promoting reading-writing links. They trialled the framework in 10 classrooms, using it to categorise and describe the nature of any links they observed teachers making, and provided us with feedback that resulted in some minor changes.

Applying the framework to classroom practice

Participants and context

Two corpora consisting of transcripts of literacy lessons of teachers of students in Years 1 to 8 of schooling were analysed. This was essentially a convenience sample where sessions that were selected were those that were predominantly teacher-led. One group of teachers (N = 16) was taking guided reading (Smith & Elley, 1994) with a small group of students and the second group (N = 12) were conducting a writing lesson where the teacher was modelling writing (including using text as a model) or leading either a whole class or a small group session in shared writing or similar. Note that many of the texts used by teachers in both settings are locally produced, designed to connect to students' social and cultural backgrounds. The guided reading lesson teachers were part of a literacy research-practice collaboration where the particular focus was to enhance the reading comprehension of students (McNaughton, MacDonald, Amituanai-Toloa, Lai & Farry, 2006) and all of the seven participating schools from which the sample of teachers came were in the lowest category of schools nationally in terms of socio-economic classification of the area from which they draw students. The writing lesson teachers were from the five schools, part of a national Literacy Professional Development project where they had identified writing as their specific focus (as opposed to a reading-focus); these schools were from the sample of schools selected as cases for the research accompanying the professional development project (Parr, Timperley, Reddish, Jesson & Adams, 2006). These latter schools ranged in socio-economic terms but tended to be mid-range schools in terms of national categorisations. Both groups were engaged in literacy professional learning projects which were school based, job embedded and designed around the demonstrated learning needs of teachers in relation to the needs of their students. Both were about a year into the project (a planned three year project in the case of the reading sample teachers and a two year project for the writing sample teachers).

Procedure

In both groups teachers were observed directly: one group wore radio microphones and one group was video recorded. Both sets of recordings were subsequently transcribed. Teachers were being observed by literacy experts both from within and outside their school in order to provide feedback about key aspects of their practice. In each case, teachers were not aware at the time of observation that a subsequent specific focus of analysis would be reading and writing links.

Analysis

Analysis of the transcripts using the framework (explained and illustrated below) allowed a test of whether it was adequate to describe instances, specifically the nature and context, of links. Then, applying the revised framework to transcripts provided an indication of typical practice and also a comparison of that practice in the two literacy instruction settings, guided reading and writing.

The lessons selected, while predominantly teacher-led, varied in both length and amount of teacher input. To take account of this variability, as we were interested in teacher focused interactions, the total word count for each transcript was obtained and the instances of links were calculated as a rate per 1000 words (the transcripts ranged in length from just under 1000 to over 7,000 words). The site of connection was problematic to identify from transcripts. Without a shared history of classroom events, it was difficult to know if a reference to a text was to a previously produced student text or a published text we did not recognise. Similarly, a reference to planned writing may not be readily perceived if the planned writing had been introduced previously and was not identified as such in the transcript. Consequently, reliability between raters for these categorisations was initially low. Providing exemplars raised the inter-rater percentage agreement to around 60 per cent but this is still far from ideal so the results should be regarded as tentative. Making the decision in real time rather than from transcripts would allow checking with the teacher, while observing over an extended period of time is likely to permit more certainty in identifying the sites. The inter-rater reliability with respect to the coding of the nature of the link made was relatively robust. Where both raters identified the occurrence of a link, agreement as to its nature was high (>95%). However, in a small percentage of cases (8%), only one of the raters identified a link being made.

Results and comments

The framework

The framework has two dimensions: sites or context where links are made, and types of link.

Sites

The three major 'textual' sites where links are made are: (i) published text to writing, for example from a shared book the class is reading to a piece of writing planned or in process; (ii) student writing to text whereby a piece of writing or the knowledge acquired from it is linked to a current reading text, and (iii) text to text which covers links between and among texts (produced by others). This latter site is not specifically a site where links are made between reading-writing but may involve an examination of what the writers of the texts did and why and may relate to current or future writing. In all cases, text is viewed broadly as print, electronic, oral, visual or multi-media.

Type of link

The second dimension of the framework concerns the type of link; the nature of the link. Four different types of links are proposed: content or context link, structural or how language works link, process or strategy link, and real world knowledge or prior experience link. These are illustrated with examples from the classroom data. As always with transcriptions, the examples lose a little of the sense they had in the real time, contextualised classroom setting.

Content or context link

Here the connection focuses on comparing or contrasting content (e.g. themes, ideas, concepts, arguments, moral/message, setting, characters).

Example: The teacher is working with the lowest reading group (NZ primary students are mostly grouped flexibly according to the level of book they are currently able to read independently; most books used in the early years of NZ primary schools are categorised by teachers and placed into a level) in a guided reading session. She has seized a teaching moment and taken an opportunity that presented itself when the young student skipped ahead looking at the pictures in the book and noticed the snake. The student asked whether 'Rags' (in the title) was the dog or the snake. Realising that s/he is thinking about the previous book where the snake was the main character, the teacher takes this confusion and turns it into a teaching point. The point comes from the fact (presumably) that the books have some of the same characters (as books often do at early reading levels) but in different stories, different characters are fore-grounded and narrate the story from their viewpoint.

T: All right let's have a look at the contents page. Chapter 1--Meeting Rags. Now there's some clues in here. Makes me want to ask who is Rags? What does it make you want to ask? Does it make you want to ask anything Clarissa?

S: Is Rags the dog or the snake?

T: Is Rags the dog or the snake? What makes you talk about the snake? Why did you say the snake?

S: He's in the wood. He's in the pile of wood.

T: Oh, you've jumped to Chapter 4. Oh, ok. You always do that to me. You always jump, that's ok.

So, obviously the story's like the one we read yesterday. It has a narrator. The narrator tells it from their point of view. Who told the story yesterday?

S: The snake.

T: The snake told the story. Who's going to tell the story today?

S: The puppy.

T: It might be the puppy. Well do you think we should start reading and find out?

Structural or how language works link

Connections of this type concern identifying how texts work, largely organisationally and linguistically, to achieve their purpose. They involve features of text like the elements of structure at text or local level (main point, topic sentence or elements of story schema like complication and resolution) or language resources like choice of particular words, the use of figurative language or the use of grammatical structures like passive constructions or tense.

Example: Here the teacher is making a link between the text they are reading (the second chapter of 'No safe harbour' by David Hill) and a movie text that has a similar setting/ context. The link, however, is to do with the schema structure of the narrative and the notion of build-up to a climax. It could also be argued that the link was helping them to see how to activate a schema they already had (for future use in both reading and writing) which is a link to do with process (schemas help by making the load on memory lighter so there is more capacity for higher level thinking).

T: Just quickly, tell the person next to you what you think is going to happen next.

S: [talking]

T: OK thank you. What do you think B ... could happen next?

S1: Um on the way the ferry starts leaving they might hit a sand bar or something and they might have to turn at the right place and smash into the sand bar and tip over or else it could just crash beside some rocks on the wharf.

T: Gee that's a lot to happen on one page isn't it [Teacher laughs] Well done.

S2: Or they could both drown.

T: Yes only chapter two but yes good thinking. J ...?

S3: Well they might start rescuing each other.

T: Yeah I know that you all know that the boat's going to sink and as someone just said it's on the front. But I don't think it's going to start sinking in chapter two. Why do I think it's not ...?

S4: And swim all the way back to the wharf.

T: Why do you think that L ...?

S4: Cos um if they sink they'll probably die and there won't be anything to say at the end.

T: Could be ... and do you think the author's going to want to keep us in suspense a little bit longer just yet?

S: [A few students say ...] Yeah.

T: Okay if you went to the movies ... who has seen Titanic?

S: [Some students raise their hands]

T: Like from the minute Jack got on the boat, did you just want it to sink next?

S: [A few students say No]

T: Did a whole three hour movie have to keep going first?

S: [A few students say Yeah]

T: OK it's the same, probably the same with this book.

Process/strategy link

In this category, connections are made between the use of processes of reasoning or problem-solving strategies used to make meaning in reading and writing. These strategies and processes may involve predicting, questioning or searching for evidence to confirm or disconfirm. Examples include showing that the act of summarizing to capture the gist is used both to aid comprehension and to make decisions about adding to or revising written text. Another example would be indicating that the activation and use of schemas from long term memory similarly aids text construction as well as comprehension.

Example: In this lesson the students are using a familiar and loved text 'Greedy Cat' (Cowley, 1997) to generate some content for an argument that they are going to write. They are learning to pose questions in order to produce argument points to write in the same way that a reader might generate questions in order to locate material in a text to test predictions or to establish content to help comprehension.

T: What are we thinking about when we write arguments? The main things we're thinking about. S ...?

S: Make sure you give your opinion

T: Yes, you write your opinion. What else, N ...?

S: Reasons?

(There are a series of interchanges around why reasons are given and around the notion of ordering reasons by importance)

T: OK. Now we know Greedy Cat is tough on cat doors. We know he doesn't have a great diet- he eats out of garbage bins! But does he really need to go on a diet? What clues are there in the text about why he might need a diet? Ask the same sort of questions that we do sometimes when we are testing out our predictions about a story and what is going to happen. What question are you going to ask, Q ...?

S: In the picture he is like, fat so maybe this means he--not getting through the cat door and breaking it?

T: So what might you ask yourself to help you look in the text for some reasons why Greedy Cat needs to diet?

S: Ummm. If he's being fat, what problems does he have?

T: Good one. See how we ask questions to look for reasons to support the argument about why he needs to go on a diet? If he's got problems being fat, he needs a diet. Just like you think up questions to test out when you are reading.

Real world knowledge link (prior experience)

This type of connection to experience or real world knowledge is not specifically one between reading and writing but, commonly, one to reading or to writing. The connection between such 'text' and either reading or writing may be emotional, experiential or autobiographical. Activating prior knowledge may aid comprehension or aid retrieval of relevant content in writing.

Example: This extract is from a Year 7 and 8 classroom where the teacher aims to help understanding of the text through an analogy to something the students are likely to understand through their experience.

T: It could get stuck to the branches and the leaves. That's a bit like the leaves. Have you ever put a um, something like a carnation into dye and seen what happens? Did you guys do that last year?

T: Yeah you could. So what's, how does the flower turn green or yellow or purple?

S1: Because the dye soaks up all, wait the leaves soak up all the dye.

T: Yeah but there's the flower eh? So how does the flower end up like that?

S2: Oh, it soaks it into the petals.

T: Yeah it does. So why I'm saying is because

sometimes the poisons and stuff get sucked up through the roots, up through the stalk of it, and into the tree. So that's one way too that poisons can get soaked up. [Teacher attends to student with hand up] Yeah?

S3: Sister, about the mangroves, can they die when they collect the poisonous?

T: Oh yes if they got poisons into them they would die just like anything else. So you can kill them. So it's probably partly they absorb some of them into them and partly they do what you said with the leaves. What else do mangroves do to protect the land?

Instances of links in typical classroom practice

The first analysis presented considers the sites of the connections made in both guided reading and a largely teacher-led segment of a writing lesson. The second analysis concerns the nature of the connections made in reading and writing. These data are presented by focus (writing or reading), then a comparison is made.

Sites connected

The sites of connection in guided reading were patterned as might be expected. A significant number (almost 90%) were made between texts (TT). This is shown in Figure 1 which presents the occurrence of linkages between sites per thousand words. This text to text reference appears to show a much higher occurrence than that noted by Soter and colleagues (2008) in all forums, save Literature Circles, although direct comparison is difficult. A very small number of links in guided reading sessions were made explicitly between the current, or another, reading text and writing (TW) (around 10 per cent). Even if some references may have been wrongly classified, given the difficulty of inference from transcript data, clearly teachers were not utilising the potential to make connections to planned writing.

[FIGURE 1 OMITTED]

Interestingly, in the writing lessons, the site of most linkages (71%) was categorised as text to text, with the connection of these texts to writing implicit, rather than stated. This may be because, in teacher-led aspects of the writing lessons, teachers were often preparing for writing by generating ideas and content from various texts, including oral texts. The unstated assumption was that these ideas were to be 'used' in the writing that was to follow but to be classified as text to writing, the link had to be made explicitly. Given transcripts are located at one point in time and such references could be made prior to or following the lesson segment, this may over-estimate text to text references in writing and underestimate text to writing references. However, specifically selected texts were also used as exemplars that illustrated the purpose and functional form of the writing being undertaken. Sometimes such texts were student pieces; sometimes short 'cameos', specifically selected or written by the teacher to exemplify a particular aspect of writing and, at other times, they were texts the students had read and shared- short stories, novels, biographies and pieces of transactional writing. Given such practice, it might be expected that more links might be expected to be noted from text to the writing in hand. The remainder of the linkages in transcripts were from the writing being undertaken, including the modelling by the teacher of constructing a text, to existing texts.

Types of links

The mean occurrence of the different types of links, calculated per 1000 words, together with the standard deviations as an indication of the variation across classrooms is shown in Table 1. In the guided reading sessions, with respect to the nature of the link, the majority were to content or context, and to prior experience in terms of real world knowledge. The category of content or context link accounted for 31 per cent of all links made and links to prior knowledge, 30 per cent. This latter link is something teachers in New Zealand, especially those with a diverse student body, do readily in such a situation to activate prior knowledge to assist with comprehension or perhaps when they are introducing new vocabulary. The other two types of links, to how language works and to processes and strategies, accounted for around 20 per cent each. An analysis (Friedman's non-parametric test for k-related samples) showed a non-significant difference between the types of links made in reading (F (3, 16) = 5.549, p > .05).

In writing, the greatest number of links, on average, made reference to aspects of how language works and the rate of reference to these was greater than to other types. About 44 per cent of links made were of this type. Such links that referred to structure or use of language may have been particularly salient to the teachers observed as they were, at this point in the professional development, working with detailed, diagnostic scoring rubrics for writing. The next largest category was links to prior, real world experience (32%). Fewer links were of the types concerning content or context (12%) or links concerning processes and strategies (11%). Unlike in reading, there was, in the writing lessons, a significant difference between types of links made (F (3, 12) = 16.210, p < .01).

It seems that guided reading and writing lessons, by their nature, each offer different opportunities to make links. Again, the small sample size needs to be borne in mind but a repeated measures analysis of variance showed a significant interaction between type of link and teaching focus, either guided reading or writing (F (3,78) = 3.385, p <. 05). The likely differences by type are shown in Figure 2. Bonferroni pair-wise comparisons indicated that this difference between reading and writing focused lessons occurred in terms of type 1 links, those referring to content or context. In Figure 2, there appears to be a difference between reading and writing for type 2, references to how texts work, with more references in writing. This difference approached significance (p = .071).

[FIGURE 2 OMITTED]

Discussion

It appears from the analyses presented that, overall, the extent to which teachers make links in literacy instruction settings where they are leading the learning as in guided reading and teacher-led writing is low. Considered in averages, this rate was about one link per thousand words of classroom discourse. In a similar way to Duncan's (1999) new entrant (kindergarten) teachers (who all professed the belief that reading and writing were linked and should be presented as complementary processes, but only one of whom on one occasion made such a link explicit in six weeks of observation), the teachers in the current study were not explicitly linking reading and writing in their instructional setting, as far as could be determined from the analysis of a transcript. While not suggesting that the frequency of links marks quality connections for learning, data from the classroom transcripts do suggest that optimal use is not being made of the opportunities to connect something new to something students already understand, a text read and discussed to a piece of writing that is planned, for example.

The data indicated that teachers mostly make links from text to the 'text' of real world experiences. This may be a phenomenon more associated with the New Zealand context where the texts used in guided reading, in particular, may be conducive to such. This is because many of the quality texts available for selection are locally constructed to be particularly relevant socially and culturally. This selection and construction of texts is designed to facilitate the use of experiential schemas to aid vocabulary acquisition, understanding of structural features of text and comprehension. Making these experiential links has been shown to enhance learning, particularly in cases where students come from different cultural backgrounds (Si'ilata, Dreaver, Parr, Timperley & Meissel, 2012). It requires knowledge about linguistic and cultural background, but little specific language content knowledge on the part of the teacher. Arguably, increasing the instances of links to organisational/ structural or other features of language and to processes and strategies could be efficacious in terms of student understanding and performance in reading and writing but perhaps teacher knowledge for this purpose is lacking (Wong Fillmore & Snow, 2002).

The details of how best to make these links are sparse. Empirical work of Parodi (2007) suggests that links be made at a macro-structural or text level. However, contrary to this, the work of Abbott and colleagues suggest that, particularly for younger students, the links be made at the same level of text. Distilling from a plethora of findings, they conclude that connections across writing and reading may be easier to create within the same level of language. Making connections across levels and across reading and writing may be challenging early in schooling, although older students may be able to do this.

The framework presented in this article is an attempt to provide some detail through the specification and exemplification of the type of link that might be made explicitly for students. There are several aspects that arise from the study regarding refinements or extension of both the framework and the testing of it. If the framework is to be used for professional learning, then perhaps the category of language links needs to be subdivided into language links at different levels of text- micro and macro- or structural/ organisational and the use of language resources. While sites of connection and types of connection were considered, a third dimension that needs to be considered concerns the degree of explicitness with which the link is made and the nature of support for the link such as through the use of artefacts. Applying the framework for research or professional learning in real-time is likely to make decisions of categorisation easier as there are multiple cues in the classroom situation (that are not in a transcript) and the possibility of checking with teachers and students.

The framework is designed to delineate the specific ways in which links might be made to foster the teaching of literacy. It is helpful in diagnosing individual teacher patterns of practice with regard to making links in order to support development of facility in making a range of types of links. Such an exercise could serve to sensitise teachers to what they actually do. It would allow them to reflect on how they connect the sites of reading and writing in a primary classroom where, in our context, ideally literacy overlaps into content areas. Ultimately, the aim of research should be to examine the relationship between the nature and degree of explicit reading and writing links in the classroom and student understanding. Checking whether students perceive the link being made and how they are able to utilise it, is an important area for further research. The goal would be to examine patterns of progress and achievement in reading and writing under conditions where explicit links are made and deliberately taught.

References

Abbott, R.D., Berninger, V.W. & Fayol, M. (2010). Longitudinal relationships of levels of language in writing and between writing and reading in Grades 1 to 7. Journal of Educational Psychology, 102, 281-298.

Biancarosa, G. & Snow, C. (2004J. Reading Next--A vision for action and research in middle and high school literacy: A report to Carnegie Corporation of New York. Washington, DC: Alliance for Excellence in Education.

Close, E.A., Hull, M. & Langer, J.A. (2005). Writing and reading relationships in literacy learning. In R. Indrisano & J.R. Paratore (Eds.), Learning to write, writing to learn: Theory and research in practice (pp. 176-193). Newark, DE: International Reading Association.

Corden, R. (2007). Developing reading-writing connections: The impact of explicit instruction of literary devices on the quality of children's narrative writing. Journal of Research in Childhood Education, 21, 269-289.

Cowley, J. (1997). Greedy cat is hungry. Wellington, New Zealand: Learning Media.

Duncan, S. (1999). Reading and writing relationships: Knowledge and processes. (Unpublished doctoral dissertation). University of Auckland, New Zealand.

Eisterhold, J. (1991). Reading-writing connections: Toward a description for second language learners. In B. Kroll (Ed.), Second language writing: Research insights for the classroom (pp. 88-101). Cambridge, England: Cambridge University Press.

Fitzgerald, J. & Shanahan, T. (2000). Reading and writing relations and their development. Educational Psychologist, 35, 39-50.

Graham, S. & Hebert, M. (2010). Writing to read: Evidence for how writing can improve reading: A report to Carnegie Corporation of New York. Washington, DC: Alliance for Excellence in Education.

Harwayne, S. (1992). Lasting impressions: Weaving literature into the writing workshop. Portsmouth, NH: Heinemann.

Hayes, J.R. (1996). A new framework for understanding cognition and affect in writing. In C.M. Levy & S. Randall (Eds.), The science of writing: Theories, methods, individual differences, and applications (pp. 1-27). Mahwah, NJ: Lawrence Erlbaum.

Hill, D. (2003). No safe harbour. Wellington, New Zealand: Mallinson Rendel.

Juel, C. (1988). Learning to read and write: A longitudinal study of 54 children from first through fourth grades. Journal of Educational Psychology, 80, 437-447.

Lancia, P. (1997). Literary borrowing: The effects of literature on children's writing. The Reading Teacher, 50, 470-475.

Langer, J.A. (1986). Reading and writing and understanding: An analysis of the construction of meaning. Written Communication, 3, 219-267.

Langer, J.A. & Allington, R. (1992). Curriculum research in reading and writing. In P.W. Jackson (Ed.), Handbook of research on curriculum (pp. 687-725). New York, NY: MacMillan.

Langer, J.A. & Flihan, S. (2000). Writing and reading relationships: Constructive tasks. In R. Indrisano & J.R. Squires (Eds.), Perspectives on writing: Research, theory and practice (pp. 112-139). Newark, DE: International Reading Association.

McNaughton, S., MacDonald, S., Amituanai-Toloa, M., Lai, M. & Farry, S. (2006). Enhanced teaching and learning of comprehension in Years 4-9 in Seven Mangere Schools. Final Report. Retrieved from http://www.tlri.org.nz/sites/ default/files/projects/9206finalreport.pdf

Nelson, N. (1998). Reading and writing contextualized. In N. Nelson & R.C. Calfee (Eds.), The reading-writing connection. Ninety-seventh Yearbook of the National Society for the Study of Education (pp. 266-285). Chicago, IL: University of Chicago Press.

Nelson, N. & Calfee, R.C. (1998). The reading-writing connection viewed historically. In N. Nelson & R.C. Calfee (Eds.), The reading-writing connection. Ninety-seventh Yearbook of the National Society for the Study of Education (pp. 1-52). Chicago, IL: University of Chicago Press.

Olinghouse, N.G. & Graham, S. (2009). The relationship between the discourse knowledge and the writing performance of elementary-grade writers. Journal of Educational Psychology, 101, 37-50.

Olness, R. (2005). Using literature to enhance writing instruction: A guide for K -5 teachers. Newark, DE: International Reading Association.

Parodi, G. (2007). Reading-writing connections: Discourse-oriented research. Reading and Writing, 20, 225-250.

Parr, J.M., Timperley, H.S., Reddish, P., Jesson, R. & Adams, R. (2006). Literacy professional development project: Identifying effective teaching and professional development practices for enhanced student learning. Auckland, New Zealand: UniServices.

Rosemary, C.A. & Roskos, K.A. (2002). Literacy conversations between adults and children at childcare. Journal of Research in Childhood Education, 16, 212-231.

Si'ilata, R., Dreaver, K., Parr, J.M., Timperley, H. & Meissel, K. (2012). Tula'i Mai! Making a difference to Pasifika student achievement in literacy. Literacy Professional Development Project 2009-2010. Research component Pasifika study. Auckland, New Zealand: UniServices.

Shanahan, T. & Lomax, R. (1986). An analysis and comparison of theoretical models of the reading-writing relationship. Journal of Educational Psychology, 78, 116-123.

Smith, J.W.A. & Elley, W.B. (1994). Learning to read in New Zealand. Auckland, New Zealand: Longman Paul.

Soter, A.O., Wilkinson, I.A., Murphy, K.P., Rudge, L., Reninger, K. & Edwards, M. (2008). What the discourse tells us: Talk and indicators of high-level comprehension. International Journal of Educational Research, 47, 372-391.

Tierney, R.J. & Shanahan, T. (1991). Research on the reading-writing relationship. Interactions, transactions and outcomes. In P.D. Pearson, R. Barr, M. Kamil & P. Mosenthal (Eds.), Handbook of reading research (Vol. 2, pp. 246-280). New York, NY: Longman.

Van Dijk, T. & Kintsch., W. (1983). Strategies of discourse comprehension. New York, NY: Academic.

Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. New York, NY: Cambridge University Press.

Wong Fillmore, L. & Snow, C.E. (2002). What teachers need to know about language. In C.T. Adger, C.E. Snow & D. Christian (Eds.), What teachers need to know about language (pp. 7-54). McHenry, IL: Delta Systems.

Judy M. Parr and Stuart McNaughton

University of Auckland

Judy Parr is Professor and Head of the School of Curriculum and Pedagogy in the Faculty of Education at the University of Auckland. Judy's research focuses on literacy, particularly writing, its development and assessment and on enhancing teacher practice in order to improve student literacy. Publications span literacy, assessment, technology, professional learning and school change.

Stuart McNaughton is a Professor and Director of the Woolf Fisher Research Centre at the University of Auckland. His research interests are in literacy and language development, the design of effective educational programs for culturally and linguistically diverse communities and cultural processes in development. Publications include research on development in family and school settings; instructional design; and intervention models in large scale interventions.

Table 1. Mean occurrence per 1000 words of different types of
links in guided reading and writing lessons

Type of link    Lesson focus   Mean   Std. Dev.   N

1. Content      Reading        1.42   1.26        16
                Writing        0.49   0.53        12
                Total          1.02   1.10        28
2. Language     Reading        0.92   1.29        16
                Writing        1.72   0.82        12
                Total          1.26   1.17        28
3. Process      Reading        0.86   1.00        16
                Writing        0.46   0.88        12
                Total          0.69   0.96        28
4. Experience   Reading        1.34   1.36        16
                Writing        1.21   0.51        12
                Total          1.29   1.07        28