Closing the gap betweenresearch and practice:

Foundations for theacquisition of literacy

Marion de Lemos

Closing the gapbetween research

and practice: Foundations forthe acquisition

of literacy

Marion de Lemos

This publication is the result of research that forms partof a program supported by a grant to the AustralianCouncil for Educational Research by State, Territory andCommonwealth governments. The support provided bythese governments is gratefully acknowledged.

The views expressed in this publication are those of theauthor and not necessarily those of the State, Territoryand Commonwealth governments.

First published 2002 by the Australian Council for Educational Research Ltd19 Prospect Hill Road, Camberwell, Victoria 3124 Australia.

Copyright © 2002 Australian Council for Educational Research

All rights reserved. Except as provided for by Australian copyright law, nopart of this book may be reproduced without written permission from thepublisher.

ISBN 0-86431-584-8

Printed in Australia by RossCo Print

Page

1

Introduction 2

Definition of literacy 3

A model of reading and writing 5

Learning to read: the self-teaching hypothesis 7

Focus of the review 8

The Australian context 9

Focus of Australian research intolanguage and literacy 10

The international context 15

Current approaches to the teachingof reading in Australia 24

Australian research on effects of phonicsversus whole language instruction 27

Relevance of research findings forteaching practice 34

The way forward 36

References 37

TABLE OF CONTENTS Page

The purpose of this review is to presentan overview of the research literaturerelating to the acquisition of literacy. Its focus is on empirical studies thatidentify the processes underlying theacquisition of literacy, and theinstructional strategies that are mosteffective in developing effective literacy.Its specific focus is on the acquisition of reading literacy.

The Context of the ReviewTeaching children how to read and writehas always been the primary objective ofeducation or schooling. However, recentconcerns that this major objective hasnot been achieved, or has not beenachieved at a satisfactory level, by manystudents by the end of the compulsoryyears of schooling has led to a renewedfocus on literacy at both Commonwealthand state level, and the introduction ofnew policies and practices which areaimed at improving literacy outcomes.These policies and practices paralleldevelopments that have occurred in a number of other countries, and haveincluded setting standards orbenchmarks to make explicit thestandards of achievement expected atdifferent levels of schooling, introducingprograms of national or state-widetesting to monitor the extent to whichthese standards are being met, andexamining the effectiveness of differentinstructional and interventionapproaches designed to improve literacy outcomes.

At the same time there have beensignificant advances over the past twodecades in the research on reading andon the processes underlying theacquisition of reading. This research has

led to the questioning of some of theassumptions on which current teachingpractices have been based, and haveidentified some of the critical factorsassociated with the acquisition ofreading skills.

This review clearly cannot hope to coverin any depth the vast and growingliterature on the development of readingliteracy. Nor would it seem useful toattempt to duplicate work that hasalready been done in terms of reviewingthe literature and drawing from such areview the implications for teachingpractice. Rather the review will draw onthe work already undertaken by expertsand expert committees, with the aim ofpresenting as clearly and succinctly aspossible the main issues that have beencovered in these reviews, and theimplications that are of particularrelevance to the Australian scene. In this sense, the review will be anattempt to pick out from the vastliterature on reading literacy the ‘plums’that might inform educators andeducational administrators of theessential findings to emerge from theresearch literature, and the implicationsof these findings for teaching practiceand educational policy.

Page

2

INTRODUCTION

Literacy has been defined in manydifferent ways, each of which reflects adifferent theoretical orientation.

A broader definition of literacy is usuallyadopted by those who see literacydevelopment as a social process, whichdevelops through exposure to literacypractices within a particular environmentand which cannot be separated from its social and cultural context. This viewrejects the notion that literacy can bedefined in terms of a set of narrowpsychological skills, and places emphasison literacy as a process of derivingmeaning from text. This definition ofliteracy usually covers other languageskills such as listening and speaking, aswell as a range of other skills includingthe interpretation of visual material, theuse and understanding of mathematicalconcepts and notation, computer‘literacy’, and critical thinking.

A narrower definition of literacy, usuallyreferred to as the conventional orcommonsense view of literacy, definesliteracy as the ability to read and write;that is, to convert the written text to thespoken word and vice versa. Under thisview the acquisition of literacy is definedin terms of acquiring the ability to bothcomprehend and produce written text.

These two opposing definitions ofliteracy are associated with differentapproaches to the study of literacydevelopment. Those who define literacyin a broader sense and who view literacy

as a social process (the socio-culturalapproach) have focused on studiesdesigned to observe literacy practices in different contexts, and to identify theways in which literacy is used fordifferent social purposes. Those whodefine literacy in a narrower sense andview literacy as essentially the ability to read and write (the cognitive-psychological approach) have focused onstudies which have sought to identify theprocesses underlying the ability to readand write, and how these are developed.

Inevitably these two views of literacyhave resulted in different types ofresearch study. The socio-cultural view of literacy has led to descriptive studiesusing ethnographic and case studyapproaches, which document inconsiderable detail the interactionsbetween the literacy learner and theirenvironment in a range of differentcontexts, including the home, thecommunity and the school. Thecognitive-psychological approach has led to experimental studies designed to identify the specific processes thatunderlie the acquisition of reading andwriting, and the ways in which theseprocesses can be enhanced by specificteaching1.

For the purposes of this review, thenarrow definition of literacy will beadopted. This will allow the review tofocus on those aspects of literacy thatare seen as of critical importance in an

Page

3

DEFINITION OF LITERACY

1 A useful presentation of these two opposing views of literacy development is provided in the two specialissues of the Journal of Research in Reading (Vol. 16, 2, September 1993, and Vol. 18, 2, September 1995) whichpresent the positions of both the ‘new literacy group’, represented by Street, Bloome, and their colleagues (inthe 1993 issue), and the response of the reading research group, represented by Oakhill and Beard, Gough,Stanovitch, Perfetti, Ehri, Goswami, Juel, and others (in the 1995 issue); the paper by Gough in the 1995 issueis particularly useful in terms of clarifying the distinction between the positions held by these two groups.

educational context. The adoption of thisdefinition recognises that the school hasa special responsibility in terms ofteaching children how to read and write.While speaking and listening skills areacquired at an early age in the homeenvironment, relatively few childrenlearn to read and write before they cometo school, and it has traditionally beenthe role of the school to teach childrenthe skills of reading and writing, asdistinct from the skills of listening andspeaking. The teaching of moreadvanced skills and knowledge leadingto the development of critical thinkingskills in other areas of the schoolcurriculum is also dependent, at least toa large extent, on the ability to read andwrite. It can therefore be argued that,from an educational perspective, theability to read and write provides thefoundation for the development of thefurther skills that are associated with the definition of literacy in its broadercontext. That is, the definition of aliterate person as an educated person,rather than as simply a person who canread and write.

Page

4

Both reading and writing depend on theability to relate print to speech. Boththerefore require knowledge of thelanguage that underlies the printed andspoken forms of a specific language(such as English), and both requireknowledge of the language’sorthography (that is, the rules that relatethe printed form of the language to thespoken form).

In so far as reading and writing are bothdependent on the ability to convert printto speech and speech to print, they areoften regarded as differentmanifestations or mirror images of asingle common skill, and a measure ofone is often used as an index ofproficiency in the other. However, theskills that underlie the recognition andcomprehension of written text aresomewhat different to the skills thatunderlie the ability to produce well-constructed text, and from this point ofview reading and writing may beregarded as composed of different butrelated sets of skills.

The basic model of reading and writingthat underlies much of the currentscientific research on the acquisition of literacy is most easily understood interms of the simple model described by Juel, Griffith and Gough (1986).According to this model reading andwriting are each composed of twodistinct abilities; decoding (or wordrecognition) and comprehension in the

case of reading, and spelling andideation (or the generation andorganisation of ideas) in the case ofwriting. Thus word recognitioncombined with the skills involved inlistening comprehension provides thebasis for reading comprehension, whilespelling combined with the generationof ideas provides the basis for writing.

While the specific skills underlying theacquisition of reading and writing aredifferent, both share a commondenominator, in that both are dependenton the set of spelling-soundcorrespondence rules of the language, or what is termed in the literature theorthographic cipher.

Knowledge of the cipher is thereforeseen as critical to the acquisition ofliteracy, since it is a basic component ofboth decoding, which underlies theacquisition of reading, and spelling,which underlies the acquisition ofwriting. Knowledge of the cipher is inturn dependent on two main factors;phonemic awareness, or the knowledgethat the spoken word can be brokendown into a series of specific sounds,and exposure to print, which providesmodels of written text and specificletters and words, which can then beconnected to specific sound sequences.Phonemic awareness and exposure toprint are therefore the two factors thatare most critical to the acquisition of literacy.2

Page

5

A MODEL OF READING AND WRITING

2 The three phonological processes generally recognised as related to reading are phonemic orphonological awareness, phonological coding in working memory, and rapid access to phonologicalinformation in long term memory. Of these three processes, phonological awareness has been found tohave the strongest causal relationship to word reading skill, and is also the most amenable to instruction,which is why it is usually noted in the literature as being critical to the acquisition of literacy.

While word recognition and spelling areessential to the ability to read and write,these abilities do not in themselvesensure comprehension of complex textor production of coherent and wellorganised writing. These higher levelskills are dependent on a range offactors, including vocabulary knowledge,familiarity with particular areas ofknowledge, knowledge and valuesassociated with membership of aparticular social or cultural group, andcritical thinking skills. However, thesehigher level skills apply equally toeffective use of spoken language. Whatdistinguishes reading and writing skillsfrom listening comprehension andspeaking skills is the fact that these skillsare expressed though the medium ofwritten text rather than through themedium of the spoken language.

Research evidence that has beenaccumulated over the past two to threedecades has supported this model of thebasic processes underlying theacquisition of literacy, and particularlythe important role played by phonemicawareness in the development of readingand writing skills. While there may bedifferences in the specific modelsproposed by different researchers toexplain exactly how phonemicawareness, word recognition andspelling skills are acquired, and howthese skills interact in the process oflearning to read and write, there isgeneral agreement about the overallmodel and the crucial role of phonemicawareness and recognition of spelling-sound correspondences in thedevelopment of reading and writing.

Page

6

Once children have acquired anunderstanding of the alphabeticprinciple, and are able to translate printto sound through the process ofphonological recoding, this provides abasis for self-teaching based on theindependent generation of targetpronunciations for novel orthographicstrings. That is to say, as childrenencounter new words or letter clusterstrings they are able to applyphonological recoding to generate thesound equivalents of the unfamiliarwords or strings, and in this way toacquire the detailed orthographicrepresentations that are necessary forrapid, autonomous visual wordrecognition (see Share, 1995). Whilephonological recoding remains essentialto this process, other factors such asvisual processing skills and short termand long term phonological memoryplay a role and may lead to individualdifferences in the speed and efficiencywith which the child is able to increasethe number of words which arerecognised visually with a minimum ofphonological processing. This processdepends on the frequency of exposureto new words. Thus the more a childreads, the greater the number of wordsthat they will be able to recognisevisually, thus enabling more fluentreading and the freeing up of thecognitive demands of the task to allowfor more cognitive focus oncomprehension as opposed to decoding.This leads to what Stanovitch (1986) hastermed the Matthew effect, with thebetter readers reading more andtherefore increasing their exposure toprint, and consequently their word

recognition skills and their fluency andspeed of reading, while poor readers,who read more slowly, will have lessexposure to print and therefore lessopportunity to build up a store ofvisually recognised words, thus spendingmore of their time and cognitive energyon decoding unfamiliar words, andtherefore falling further behind in theirreading achievement.

This self-teaching mechanism is basedon two fundamental prerequisites –symbol-sound knowledge and phonemicawareness. Neither of these skillsdevelop spontaneously throughexposure to print. This has obviousimplications for reading instruction. As Coltheart (1980) has argued, gettingchildren ready to read means teachingthem the skills they will need in order toread. That is to say, explicit teaching ofsymbol-sound relationships andphonemic awareness.

Page

7

LEARNING TO READ:THE SELF-TEACHING HYPOTHESIS

Research on the acquisition of literacyhas tended to focus more on theacquisition of reading skills than on theacquisition of writing skills, although anumber of studies have looked at thedevelopment of spelling and the role ofinvented spelling in the development ofwriting (see, for example, Nicholson,2000, Chapter 9). This emphasis onreading can probably be attributed totwo main factors. First, the fact that thereis much greater emphasis in the schoolcurriculum on the teaching of readingthan on the teaching of writing, andsecond, the fact that standardisedassessments are more easily applied toreading than to writing, which makes theacquisition of reading skills moreamenable to scientific study than theacquisition of writing skills (see, forexample, Nelson and Calfee, 1998).

This review will therefore focus morespecifically on research into theacquisition of reading skills, which formthe basis for the development of readingliteracy. Following Tunmer (1999), readingliteracy may be defined as comprisingthe following abilities:

• the ability to read at a level necessaryfor self-sustained growth in literacy

• the ability to understand in print whatwould be expected to be understoodin the corresponding spoken languageby native speakers of the same age; and

• the ability to understand, use andreflect on written texts in order toachieve one’s goals, to develop one’sknowledge and potential and toparticipate in society.

Page

8

FOCUS OF THE REVIEW

There have been a number of Australianreports and publications relating toliteracy and literacy development overthe past ten years. These have includedpolicy documents, such as the 1991policy statement Australian Languageand Literacy Policy (DEET, 1991), and the1998 publication Literacy for All: TheChallenge for Australian Schools (DETYA,1998), outlining the National Literacy andNumeracy Plan agreed to by allCommonwealth, State and TerritoryEducation Ministers in 1997. There havealso been reports of review committeessuch as the 1992 report The LiteracyChallenge, by the House ofRepresentatives Standing Committee onEmployment (1992), and results ofnational and state surveys including thereport on the National School EnglishLiteracy Survey (Masters and Forster,1997), as well as publications relating tothe development of national standards,profiles and benchmarks.

Two summary reviews of the researchliterature on factors related to thedevelopment of literacy and research onthe teaching and learning of literacy havebeen undertaken at ACER over the pastfive years; the paper on Factors Relatedto the Development of Literacy (deLemos and Harvey-Beavis, 1995),prepared as one of the backgroundpapers for the National School EnglishLiteracy Survey, and the review ofliterature prepared for the LiteracyAdvance Project commissioned by theCatholic Education Commission ofVictoria (Ainley, Fullarton, Frigo andOwen, 1999). Other notable Australianreviews of the research literature onreading and literacy developmentinclude the summary of the research

literature on learning to read, publishedby the Tasmanian Department ofEducation and Arts (1994), the review ofresearch into language and literacy byFreebody and Gilbert (1999), and thereview of recent developments inlanguage and literacy development byBowey (2000). The Tasmanian publicationwas directed to teachers, and intendedto provide them with a knowledge basethat would inform their teaching practiceand lead to more effective teachingstrategies based on established evidencerelating to the effectiveness of differentapproaches to the teaching of reading.The Freebody and Gilbert reviewprovides a broad overview of Australianresearch in the area of literacy andlanguage over the past 30 years, whilethe Bowey review provides an overviewof current theoretical research on theacquisition of reading and theimplications of this research inunderstanding the underlying causes ofreading difficulties.

This review is designed to supplementrather than to duplicate these variousreports and reviews, focusing specificallyon the findings that are of mostsignificance in terms of identifying thecritical factors that are related to literacydevelopment, and the teaching practicesand strategies that are most effective inimproving literacy outcomes.

Page

9

THE AUSTRALIAN CONTEXT

The review on research into languageand literacy by Freebody and Gilbert(1999) provides a useful starting point forthis review in that it traces the majorfactors that have influenced Australianresearch in this area over the past 30years, and as such provides a frameworkfor placing current developments inperspective.

The authors of this review point to theenormity of their task in terms of thescope, complexity and time framecovered by the review, and acknowledgethat their review is necessarily selective,and influenced by their own view ofwhat is important and consequentialabout the last 30 years of work in thisarea.

Australian research into language andliteracy is considered under two mainheadings: developments in theorisingand research methodologies applied to language and literacy, and distinctivedomains of research.

Under the heading of theoretical andmethodological developments, Freebodyand Gilbert note the shift in emphasisfrom the paradigms derived fromeducational assessment and testing,which dominated research on languageand literacy in the 1950s and 1960s, tonew paradigms that have emerged,derived from sociological and linguistic

perspectives, and which define languageand literacy in terms of sociallyconstructed practices.

Some reference is made in this sectionto research deriving from the cognitivepsychological model, which hasemanated mainly from universitydepartments of psychology or specialeducation units in faculties of education.Reference is also made to Chall’s 1967survey of the research literature onreading acquisition, and the subsequent‘great debate’ between holistic andanalytic methods of teaching reading.The work of Australian researchersfollowing this tradition is noted(Andrews, 1989, 1992; Byrne and Fielding-Barnsley, 1989). While recognising thatresearch efforts in these directions havebroad applicability, it is noted that theirinfluence has been restricted to debatesin special education and psychology, andthat the diminished visibility of this lineof research in the Australian contextrelates to influences coming from othersources. Specific reference is made tothe 1966 international seminar atDartmouth College in the United States3,convened to consider critical problemsin the conceptualisation of language andliteracy education, and the extent towhich the impact of this seminar and theresearch traditions it authoriseddominated Australian language and

Page

10

FOCUS OF AUSTRALIAN RESEARCH INTOLANGUAGE AND LITERACY

3 It is of some interest to note that Freebody and Gilbert describe this seminar as attended by ‘languageresearch teams’ from North America and the United Kingdom, while Dixon’s (1967) report on thisconference describe it as attended by a group of 50 people ‘all concerned in one way or another with theteaching of English’. The topics covered by the conference related to the teaching of English as a subject atboth primary and secondary level rather than the acquisition of literacy or reading skills, and was attendedmainly by professors and lecturers in English in University departments of Education or in teacher traininginstitutions, rather than by researchers in reading, linguistics or language acquisition.

literacy discourses and their attendantresearch enterprises for the ten years tofollow. The influence of Britishresearchers such as Britten, Burgess,Martin and Rosen4, based at the Instituteof Education at the University of London,is also noted, and the extent to whichtheir views dominated theorisations oflanguage and literacy research inAustralia through the early 1980s.

This review by Freebody and Gilbertdocuments the influences underlying theshift in emphasis from studies based onexperimental designs and quantitativeassessment in the 1960s and 1970s, tostudies, from the early 1980s onwards, thatare based on more descriptive methodsderived from sociological models. Thesemethods include those based on textlinguistics, which use documentarymethods to analyse the conventions ofspoken and written language, as well asethnographic methods, deriving fromboth anthropological and sociologicaltraditions, which utilise observationalmethods of documenting material andinteractional patterns (as inanthropological studies), as well as theanalysis of texts and interactions in orderto apply critical theories of socialorganisation, patterns of dominance andcontrol, and avenues for changemotivated by cultural and social equity (as in sociological studies).

Research Studies on Language andLiteracy funded by DETYAThe major source of funding forAustralian research into language andliteracy comes from the Department ofEducation, Training and Youth Affairs(DETYA), and particularly from theChildren’s Literacy National Projects

Programme (CLP). This programcommenced in 1992 and continued to1996, after which it was incorporated intothe Commonwealth’s broadbandedLiteracy Programme, Grants for NationalLiteracy Strategies and Projects. Over theperiod 1992 to 1996 sixteen nationalprojects were funded under theChildren’s Literacy National ProjectsProgramme. These projects wereexpected to improve understanding ofissues relevant to the national delivery ofhigh quality children’s literacy programswithin government and non-governmentschools.

A summary of the findings of thesesixteen projects and their implicationsfor future literacy research andprofessional development strategies toimprove literacy teaching practices inclassrooms was put together followingthe 1998 Researchers’ Conferenceorganised by the Literacy Section of theLiteracy and Special Programmes Branchof DETYA. This conference was designedto bring representatives from each of theprojects together to focus on directionsfor future national literacy research andappropriate professional developmentbased on the CLP research. Theoutcomes of this conference werereported in two volumes: an ExecutiveSummary, which describes the CLP andprovides an overview of the main themesof these projects and a summary versionof the position papers and presentationson each project, and a ConferenceReport, which includes in addition to theabove a programme evaluation, togetherwith detailed appendices containing asummary of key findings andrecommendations from all sixteen CLP projects (Gunn, 1999).

Page

114 See, for example, Britton, 1970; Britton et al, 1975; Burgess et al, 1973; Rosen and Rosen, 1973.

While these sixteen projects covered arange of topics, they were in most casesdescriptive studies of literacy practices indifferent contexts. Of the sixteen studies,six focused on literacy practices in boththe home or community context and theschool context, while five focused onaspects of literacy in the school orclassroom context. Three of the projectsfocused on assessment and reportingprocedures, one on the relationshipbetween first language development andsecond language acquisition, and one onthe development of a classroomresource to support the use of orallanguage as a tool for learning. In mostcases the methodology was based ondescriptive methods, includingobservation and analysis of relevantdocuments, and in many of the studiesintensive case studies of individualchildren, families, teachers, classes orschools constituted a major part of thestudy. The studies covered both primaryand secondary level, with only one studyfocusing on the acquisition of earlyliteracy skills. Relatively few provided anyhard data in terms of measures of theliteracy achievement of students, or datathat would allow for any analysis of therelationships between specific school orbackground variables and literacyachievement. No studies focused on anexamination of the effectiveness ofdifferent approaches to the teaching ofreading, or on the effectiveness ofdifferent types of intervention strategiesfor students at risk. From thisperspective, these studies fall outside thefocus of this particular review.

Other Australian Research intoReading and Literacy DevelopmentOther sources of funding for Australianresearch into literacy and readingdevelopment include research grantsfrom the Australian Research Council,funding from state education systems,and postgraduate research studiesfunded through the university system.Funding from the Australian ResearchCouncil, as compared with fundingthrough the DETYA CLP program, hastended to cover a more diverse range ofstudies reflecting different theoreticaland methodological viewpoints, and it isfrom this source that most of the studiesof reading development based on acognitive-psychological approach andemploying an experimental methodologyhave been funded. These studies havebeen undertaken mainly bypsychologists in university departmentsof psychology or special education,confirming the pattern noted byFreebody and Gilbert (1999), and havefocused on basic research into theprocesses underlying readingdevelopment. The work of theseresearchers is perhaps best representedin the collection of papers published inthe 1996 Special Issue of the AustralianJournal of Psychology5, and morerecently the collection of papers onlanguage processes and problemspublished in the latest issue of theAustralian Educational andDevelopmental Psychologist.6 The papersin the 1996 Special Issue of the AustralianJournal of Psychology were presented ata symposium on Reading andDevelopmental Dyslexia, which was heldat the University of Tasmania in February

Page

12

5 Australian Journal of Psychology, Volume 48, Number 3, 1996.

6 Australian Educational and Developmental Psychologist, Vol. 17, Number 1, 2000).

1996. This symposium, funded mainly bythe ARC under their Special ResearchInitiative Program, brought together 12Australian researchers, all of whom hadreceived their research funding from theARC, to discuss and review their work inthe reading area. As noted by Pratt andColtheart7, the issues covered and theemphases of the different contributorsvaried, but all researchers at thissymposium were firmly of the view thatan understanding of the alphabeticprinciple and a knowledge of soundletter correspondences underlie thedevelopment of proficient reading.They also noted the concerns of theparticipants that many educationprograms currently used in Australia andelsewhere do not give sufficientrecognition to the development of thefundamental skills involved in worddecoding. While the research reportedin these papers clearly has implicationsfor teaching practice, these studies havegenerally had little impact on system-wide approaches to the teaching ofreading or teacher understandings of theprocesses underlying the acquisition ofreading.

Systematic evaluations of specificteaching approaches or interventions onstudent outcomes are relatively rare inAustralia, but an evaluation of theVictorian Early Years Literacy Program iscurrently being undertaken by Hill andCrévola at the University of Melbourne(Crévola and Hill, 1998) while theCatholic Education Commission ofVictoria is funding an evaluation of theimpact of different approaches to theteaching of literacy in Victorian Catholic

schools (Ainley and Fleming, 2000). A study of the effectiveness of ReadingRecovery was undertaken in New SouthWales in 1991 (Center, Wheldall,Freeman, Outhred and McNaught, 1995),and more recently an evaluation of theMacquarie University Schoolwide EarlyLanguage and Literacy Program (SWELL),(Center, Freeman and Robertson, 1998;2001a; 2001b).

Summary of Australian Research onReading and Literacy Developmentin Recent YearsFrom the Freebody and Gilbert review ofAustralian research into language andliteracy and the summary of projectsfunded by the DETYA Children’s LiteracyNational Projects Programme, it is clearthat Australian research into languageand literacy over the past two decadeshas been dominated by the view thatliteracy is a socio-cultural phenomenonthat cannot be separated from its socialcontext. As a consequence, the bulk ofthe research into literacy, and particularlythe research funded through the DETYACLP program, has been research intoliteracy practices in a variety of socialcontexts, in which the dominant researchmethods are descriptive andethnographic, with an emphasis onobservational and case study techniques.

At the same time, there is an activegroup of researchers, based mainly inuniversity departments of psychologyand special education, and more recentlyincluding people with a background inspeech pathology8, who have undertakenresearch into the development of the

Page

13

7 Pratt and Coltheart, Guest Editorial, Australian Journal of Psychology, Volume 48, Number 3, 1996.

8 See, for example, comment by Janet Fletcher, in her Guest Editorial for the Special Issue of the AustralianEducational and Developmental Psychologist (Vol 17, No 1, 2000).

processes underlying the acquisition ofreading. This research is in general basedon a cognitive psychological model, andapplies scientific and experimentalapproaches to the study of readingdevelopment.

There is relatively little Australianresearch which has involved thesystematic evaluation of educationalprograms designed to enhance literacyskills. This is despite the widespreadadoption of programs such as theWestern Australian First Steps Programand the Victorian Early Years LiteracyProgram, as well as interventionprograms such as Reading Recovery.There have also been various stateassessment programs that have beenadopted at primary and school entrylevel, presumably on the assumption thatsuch assessments will have a positiveeffect on student outcomes. Howeverthere is as yet no cumulative body ofresearch which can be used as a basis forevaluating the impact of these programsor initiatives on students’ literacyachievement.

Page

14

While the socio-cultural approach has itssupporters in other countries, it is lessdominant and has not been as influentialin terms of its impact on educationalpolicy and research as has been the casein Australia. This is particularly evident inthe United States, where a series ofreports published over the last ten yearshas been influential in drawing attentionto the findings of research in the readingarea, and the implications of thisresearch for teaching practice andeducational policy.

There is now an international body ofresearch on the processes underlyingthe acquisition of reading. Researchers in the United States, Canada, the UnitedKingdom, New Zealand and Australiahave contributed to this research, as wellas researchers in a number of Europeancountries. While recognising thecontribution of researchers from allthese countries to this accumulated bodyof research, this review draws mainly onrecent United States reports which havesought to synthesise the availableresearch evidence relating to theacquisition of reading and theeffectiveness of different approaches tothe teaching of reading, and theimplications of this research for teachingpractice.

The most recent of these reports is theReport of the National Reading Panelconvened by the Director of the NationalInstitute of Child Health and HumanDevelopment, at the request of theUnited States Congress, in order toassess the status of research-basedknowledge on reading, including theeffectiveness of various approaches toteaching children to read (NationalReading Panel, 2000). This report is

particularly significant in that the Paneldecided to adopt a set of rigorousstandards to assess the efficacy ofmaterials and methodologies used in theteaching of reading and in theprevention or treatment of readingdisabilities. The standards adopted werethe same as those applied to researchinto the efficacy of interventions inpsychological and medical research, onthe basis that the standards applied todetermining the efficacy of educationalinterventions should be no less rigorousthan those applied to determining theefficacy of behaviourally basedinterventions, medications or medicalprocedures proposed for use in theprevention or treatment of medical orpsychological conditions affecting theperson’s physical or psychological health.

If this approach were to be applied toother areas of education and educationalresearch, it could well mark a turningpoint in the history of education. Up tonow innovative educational practices andinterventions have been adoptedwithout any requirement for research-based evidence as to their effectivenessor their impact on children’s learning orother aspects of their social orpsychological development (includingthe possibility of unintended negativeeffects). The application of a new set ofstandards requiring evidence of theefficacy of any proposed newintervention or initiative, includingevidence relating to any possibleunintended ‘side-effects’, would havesignificant implications for education.While it may be argued that educationalinterventions do not carry the samelevels of risk as medical interventions,and the same standards are not therefore

Page

15

THE INTERNATIONAL CONTEXT

applicable, there is nevertheless a goodcase to argue that students and theirteachers should not be subjected tochanges in teaching methodology oreducational practices unless there isevidence to support the supposedbeneficial effects of the change inpractice. In an era of continuouslychanging policies and practices ineducation, a more measured approachwhich involves an investigation of theimpact of any proposed change prior toits widespread adoption could well havebeneficial effects all round.

In order to understand the context inwhich this panel was requested, at thehighest level of government, to assessthe research-based knowledge relatingto the effectiveness of differentapproaches to the teaching of reading, it is necessary to understand thebackground to this request andparticularly the heated and at timesacrimonious educational debate thatpreceded it.

Predecessors to the Report of theNational Reading PanelPrior to the release of the report of theNational Reading Panel there were fourlandmark reports on the status ofreading instruction in the United States,each regarded as providing anauthoritative view of the researchevidence relating to the effectiveness ofdifferent approaches to the teaching ofreading at the time of their publication.

The first of these reports was Jean Chall’sinfluentional book Learning to Read, theGreat Debate, published in 1967. This wasfollowed in 1985 by the report Becominga Nation of Readers, by Anderson,

Hiebert, Scott and Wilkinson. MarilynAdams’ book Beginning to Read:Thinking and Learning About Print, waspublished in 1990, and the reportPreventing Reading Difficulties in YoungChildren, edited by Snow, Burns andGriffin, was published in 19989.

Each of these reports was commissionedor funded by a major national body.Chall’s study, undertaken over the period1962 to 1965, was funded by a grant fromthe Carnegie Corporation, while theAdams’ study, undertaken over theperiod 1987 to 1989 at the ReadingResearch and Education Centre at theCentre for the Study of Reading at theUniversity of Illinois, was funded, in part,by the Office of Educational Researchand Improvement. In the case of thesetwo reports there was a single authorwho worked with a group of advisors.The Anderson et al report, published in1985, was funded by the Commission onReading of the National Academy ofEducation, and while conducted underthe aegis of the funding body, thecontent of the report clearly reflectedthe views of the joint authors. The Snowet al report was undertaken under theauspices of the National ResearchCouncil of the National Academy ofSciences, and was funded and jointlysponsored by three federal agencies; theOffice of Special Education Programs inthe Department of Education, the Officeof Educational Research andImprovement – Early Childhood Institutein the Department of Education, and theNational Institute on Child Health andHuman Development – Human Learningand Behaviour Branch. The report itselfwas prepared over a period of three

Page

169 A historically based review of this report, which identifies the links between this report and itspredecessors, is provided by Pearson (1999).

years by a Committee made up ofexperts drawn from the fields ofcognitive psychology, languagedevelopment, special education,medicine and literacy education. In thiscase the authorship of the report is lessclear, although it can probably beassumed that different sections of thereport were prepared by differentmembers of the Committee, with a finaldraft prepared by the editors andapproved by the Committee. As Pearson(1999) points out, given a Committeerepresenting different philosophicalpositions on both methodological andcurricular questions, the process ofarriving at conclusions andrecommendations that were acceptableto all members of the Committee musthave involved a difficult process ofnegotiation and compromise to reachconsensus.

Taking each of these reports in order.

Jeanne Chall’s book ‘Learning to Read:the Great Debate’ was a response to thedebate regarding the role of phonics inlearning to read. While this topic hasbeen the subject of debate over manyyears, it reached a peak in the UnitedStates in the 1950s, with the publicationin 1955 of Rudolph Flesch's book WhyJohnny Can't Read. This book followedthe shift in emphasis in the teaching ofbeginning reading from a code-basedapproach (first teach the alphabet, andthe ability to read will follow) to ameaning-based approach (start byteaching children to recognisemeaningful words (look-say), and theywill gradually pick up the code). Fleschargued that written English is alphabetic,

and thus phonetic, and that phonicinstruction is the only natural system ofteaching children to read. Mastery of thealphabetic code is therefore the key tolearning to read, and the failure of manychildren to learn to read was attributedto failure to teach children the code.

Chall’s comprehensive review of existingmethods and research on beginningreading was designed to provide a basisfor evaluating the arguments in favour ofthese opposing approaches to theteaching of reading. The results of thisreview pointed to a consistent andsignificant effect of phonics instructionas a factor in reading achievement, andChall’s conclusions were supported by aseries of further reviews and studieswhich were undertaken in response tothis debate.10

Becoming a Nation of Readers, byAnderson et al (1985), was a response tothe report A Nation at Risk: TheImperative for Educational Reform, by theNational Commission on Excellence inEducation, which warned of the risk forAmerica of shortcomings in levels ofliteracy at the secondary level. TheAnderson et al report reviewed theevidence relating to reading, includingthe processes involved in reading, thefactors that affect reading development,and classroom practices relating to theteaching of reading. They came up with aset of recommendations with regard tothe conditions likely to produce citizenswho read with high levels of skill, and doso frequently and with evidentsatisfaction. These recommendationscovered a range of practices relating tothe home, the school, and the teacher,

Page

17

10 See, for example, Bond and Dykstra (1967), Pflaum, Walberg, Karegianes and Rasher (1980), and Adams(1990). A more detailed review of Chall’s study as well as the outcomes of the First Grade Reading Studies isprovided in an unpublished paper by de Lemos (1997).

but insofar as instructional practiceswere concerned, there was a clearemphasis on the importance of well-designed instruction in phonics and thecontinuing application of phonics toword identification and reading inmeaningful contexts.

The review by Marilyn Adams, publishedin the book Beginning to Read: Thinkingand Learning About Print, wasundertaken in response to the ongoingand increasingly acrimonious debatebetween proponents of phonics andproponents of the whole languageapproach to the teaching of reading. This report provided a comprehensivereview of the literature relating to theacquisition of reading, including anhistorical overview of the developmentfrom ideographic to alphabetic writingsystems, and the central role ofphonemic awareness and phonics inproviding a basis for mapping symbols to sound in order to convert the spokenwork to the written symbol, and viceversa. This review drew on thesubstantial body of research that hadaccumulated over the 1970s and 1980s,which demonstrated the link betweenphonemic awareness and subsequentachievement in reading, and indicatedthat successful phonics instruction isdependent on the child’s ability torecognise the individual sounds withinwords and to break words down intotheir separate sounds. This work waswidely quoted, and became recognisedin the 1990s as the most authoritativereview of the research literature relatingto reading to date.

However, despite its wide acceptancewithin the research community, thisreport failed to resolve the readingdebate. As a consequence a newsynthesis of the research evidencerelating to reading instruction, with aspecific focus on strategies that mightprevent reading failure, wascommissioned by the US Department ofEducation and the US Department ofHealth and Human Services, resulting inthe Snow et al report Preventing ReadingDifficulties in Young Children.

This report differed from the Adamsreport in that it focused much morespecifically on the implications of theresearch evidence for the teaching ofreading, with a particular emphasis onthe prevention of reading difficultiesthrough intervention programs at thepreschool level and effective teaching inthe early years of schooling. The reportidentifies specific teaching goals at thepreschool to Grade 3 level, together withrecommended teaching strategies toachieve these goals. The report alsoreviews the research evidence relating tothe effects of various school and homefactors on early literacy development, aswell as the effectiveness of differentearly intervention approaches for theprevention of reading difficulties in thecase of children who might be at risk, orfor assisting children who fail to achievesatisfactory progress in the early stagesof learning to read.

Page

18

The Report of the National ReadingPanelWhile the Snow et al report onPreventing Reading Difficulties in YoungChildren was widely acclaimed, itsfindings and recommendations were notaccepted by all, and it was felt that amore systematic review of the researchevidence relating to the teaching ofreading was required. This led to theconstitution by the US Congress of theNational Reading Panel, charged with thetask of assessing the status of research-based knowledge, including theeffectiveness of various approaches toteaching children to read. The panel wascomposed of 14 individuals, includingleading scientists in reading research,representatives of colleges of education,reading teachers, educationaladministrators, and parents. In order tocover the major topics designated thePanel established five subgroups to coverthe areas of alphabetics, comprehension,fluency, teacher preparation andcomputer-linked instruction.

The Panel developed a set of rigorousscientific standards to evaluate theresearch on the effectiveness of differentinstructional approaches used inteaching reading skills. Regional hearingswere held to allow public input, and toinform the panel of the issues that wereconsidered important by the public, andthe needs and concerns of those whowould be required to implement thePanel’s findings and determinations.

The key issues that emerged from thepublic hearings were:

• the importance of the role of parentsand other concerned individuals inproviding children with early languageand literacy experiences that fosterreading development

• the importance of early identificationand intervention for all children at riskfor reading failure

• the importance of phonemicawareness, phonics and good literaturein reading instruction, and the need todevelop a clear understanding of howbest to integrate different readingapproaches to enhance theeffectiveness of instruction for allstudents

• the need for clear, objective, andscientifically based information on theeffectiveness of different types ofreading instruction and the need tohave such research inform policy andpractice

• the importance of the role of teachers,their professional development, andtheir interactions and collaborationswith researchers, which should berecognised and encouraged

• the importance of widely disseminatingthe information developed by thePanel.

Their search of the research evidencerelating to the teaching of readingidentified a total of about 100 000 studiessince 1966, and another 15 000 publishedprior to this time. Because of the largevolume of studies, the panel selectedonly experimental and quasi-experimental studies for their review,and of these, only those that metrigorous scientific standards in reachingtheir conclusions.

The main conclusions reached by thepanel, in the various aspects of readinginvestigated, are summarised in thefollowing twp pages.

Page

19

Page

20

Phonemic awareness

Teaching children explicitly and systematically to manipulate phonemes significantlyimproves children’s reading and spelling abilities (an overall effect size of .86 onmeasures of phonemic awareness outcomes, an overall effect size of .53 on readingoutcomes, and an overall effect size of .59 on spelling outcomes, based on a total of96 comparisons from 52 studies).

The Panel’s conclusion was that the evidence on this was so clear cut that this methodshould be an important component of classroom reading instruction.

Phonics instruction

Systematic phonics instruction (as compared with nonsystematic phonics instructionor no phonics instruction) produces significant benefits for children fromKindergarten to Grade 6, and for children having difficulties in learning to read (overalleffect size of .44, based on 66 comparisons derived from 38 studies). The greatestimprovements in reading are associated with synthetic phonics instruction11 (effectsize of .45), as compared with programs based on analysis and blending of largerunits12 (effect size of .34) or programs using other systematic approaches or where thespecific nature of the approach was not specified (effect size of .27). It was also notedthat the effects of systematic phonics teaching were greater at Kindergarten and Grade1 (.56 and .54) than in Grades 2 to 6 (.27), and greater for children from low SESbackgrounds (.66) as compared with children from high SES backgrounds (.44). Effectswere also greater for children identified as ‘at risk’ (.58 at Kindergarten level and .74 atGrade 1 level), as compared with children identified as ‘reading disabled’ or where lowachievement was associated with other cognitive difficulties (.32 and .15). Systematicphonics instruction also had a stronger effect on spelling for children in Kindergartenand Grade 1 (.67) than for children in Grade 2 to Grade 6 (.09).

The Panel’s conclusion was that the evidence relating to the effectiveness of phonicsinstruction in improving reading outcomes was sufficiently strong to indicate thatsystematic phonics instruction should be a part of routine classroom instruction. Itwas however noted that because children vary in the skills they bring to theclassroom, no single approach to teaching phonics can be used in all cases, and thatteachers require training in different approaches to the teaching of phonics and howthese approaches can be tailored to meet the needs of particular groups of students.

Oral reading

Guided oral reading (that is, reading aloud to the teacher, parent, or a fellow student)is important for developing reading fluency (average weighted effect size of .41). Thehighest impact was on reading accuracy (mean effect size of .55), followed by readingfluency (mean effect size of .44) and reading comprehension (mean effect size of .35).However, there was substantial variation in the effect sizes reported for these studies(from .05 to 1.48), as well as substantial variation in the sample sizes. Because ofthe great range in the nature and design of the studies examining the effects ofguided reading, and in many cases the lack of either transfer or control data, onlyfourteen studies were found to be appropriate for inclusion in the meta-analysisfrom which the average weighted effect size was calculated.

Page

21

11 Synthetic phonics involves teaching students explicitly to convert letters into phonemes and to blendphonemes to form words.

12 Such as clusters of letters forming a subpart of the word, as in onsets, rimes, phonograms, and spellingpatterns.

Silent reading

The panel was unable to determine whether reading silently to oneself helped toimprove reading fluency. While hundreds of studies have demonstrated that betterreaders do more silent reading than poor readers, these studies are unable todetermine whether independent silent reading improves reading skills or that goodreaders simply prefer to do more silent reading than poor readers. Although notdiscouraging the practice of silent reading as a classroom technique, the Panelrecommended that this be done in combination with other types of readinginstruction such as guided oral reading.

Vocabulary instruction

The Panel was unable to identify the best method for teaching vocabulary, andconcluded that vocabulary should be taught both directly and indirectly, thatrepetition and multiple exposure to words, as well as computer technology, will assistvocabulary development, and that instruction should be based not on a singlemethod but on a combination of methods.

Reading comprehension

With regard to the comprehension of text, the Panel found that readingcomprehension is best facilitated by teaching students a variety of techniques andsystematic strategies to assist in recall of information, question generation, andsummarising of information. It also found that teachers must be provided withappropriate and intensive training to ensure that they know when and how to teachspecific strategies.

Teacher training

The Panel noted that existing studies showed that training both new and establishedteachers generally produced higher student achievement, but that the researchevidence is inadequate to draw clear conclusions about what makes training mosteffective. More quality research on teacher training was one of the major researchneeds identified by the Panel.

Computers and reading

With respect to computer technology, the Panel noted that there are too few definitivestudies to draw firm conclusions, but the available information indicates that it ispossible to use computer technology for reading instruction. The use of hypertext (highlighted text that links to definitions or related text) was noted as one possibleteaching strategy. It was also noted that the use of computers as word processorsmight help students learn to read, as reading instruction is most effective whencombined with writing instruction.

This 800 page report of the NationalReading Panel is the mostcomprehensive review yet of theresearch evidence relating to the factorsunderlying the acquisition of readingand the effectiveness of differentapproaches to the teaching of reading.But whether or not it will finally resolvethe debate between proponents of thewhole language approach to theteaching of reading versus the phonicsapproach, or the debate between thosewho view reading as essentially a socialprocess as against those who viewreading as a process of mapping thesounds of words to their written symbol,remains to be seen.

A possible criticism of the Report is thatit fails to address many other issues thatare of concern to teachers, educationaladministrators and the public. Theseconcerns include the social factorsrelated to the development of literacy,including the early development ofliteracy in the home and preschoolenvironment, and the importance ofhome–school partnership in developingchildren’s literacy skills and attitudes toliteracy. Broader school factorsassociated with literacy development arealso not addressed in the Report. Suchfactors would include the effects of classsize and class grouping; the effectivenessof various intervention strategies such asReading Recovery for children identifiedas having difficulty in learning to read;and what makes teachers and schoolseffective. Other issues not covered in theReport are the role of assessment andreporting in the improvement of studentoutcomes, and the setting of targets orbenchmarks for student achievement.

While these are important issues to beconsidered, there are two lines ofargument in defence of the stand takenby the Reading Panel.

It could be argued that the task ofconsidering all of these issues wasbeyond the scope of a single Committee,which was appointed for the specificpurpose of seeking answers to what wasseen as the major source of debate anddispute within the educationalcommunity with regard to theeffectiveness of different approaches tothe teaching of reading. Resolving thisdebate could then open the way toconsidering other issues in a moreobjective and dispassionate way.

It could also be argued that the researchevidence relating to these broader issueshas not yet reached a state of consensusin which implications for teachingpractice could be put into immediateeffect. One of the main findings arisingfrom the report of the National ResearchCommittee on Preventing ReadingDifficulties in Young Children13, whichprovided the basis for the work of theNational Reading Panel, was that asufficient degree of consensus had beenreached by the members of this diverseCommittee to provide an integratedpicture of how reading develops and howreading instruction should proceed.However, the limitations of this report, asnoted by the Reading Panel, was that itwas a consensus document based onbest judgement rather than a rigorousanalysis of the research evidence. It didnot specifically address how criticalreading skills are most effectively taughtand what instructional methods,materials and approaches are mostbeneficial for students of varying abilities.

Page

22 13 See Snow, Burns and Griffin (1998).

This was the issue that the Reading Panelsought to address in their Report.

Summary of US Reports onResearch Evidence Relating toEffective Instructional Practices Over the period 1967 to 2000 a series ofreports addressing the issue of effectiveinstructional practices for the teaching ofreading were undertaken in the UnitedStates. All of these reports came up withthe same conclusion; that systematicteaching of phonics was a necessary andessential component of any program forthe teaching of beginning readers.Despite the range of programs studiedand the various other factors associatedwith reading achievement, there was aconsistent finding of higher achievementbeing associated with programs thatincluded systematic teaching of phonics.

However, all these reports recognisedthat systematic teaching of phonics wasonly one aspect of a comprehensiveprogram for the teaching of reading, andthat a successful teaching programneeded to include a variety of othercomponents, including exposure to arich language and print environmentwith opportunities to develop listeningand oral language skills through storyreading and discussion, particularly inthe early stages of learning to read;exposure to interesting picture and storybooks to encourage early reading and todevelop children’s motivation to read;and the provision of stimulating andchallenging reading material matched to

their level of ability to ensure growth inreading development throughout theyears of schooling. The phonics versuswhole language debate was nottherefore a debate about two opposingand mutually exclusive approaches to theteaching of reading. It was rather adebate about whether or not systematicteaching of phonics should be includedas one aspect of the teaching of reading,in a broader program including bothphonics approaches and the literacy-based approaches favoured by wholelanguage teachers.

While in recent years there has been arecognition of the need for morebalanced approaches to the teaching ofreading14, there remains a significantnumber of teachers whose approach isbased on the theoretical assumptionsthat underlie the whole languageapproach, including the assumption thatreading is acquired through a process ofexposure to written language, in thesame way that learning to speak isacquired through exposure to thespoken language, and that no formalteaching of reading is therefore required.Such teachers continue to oppose formalsystematic teaching of phonics, in thebelief that reading for meaning does notdepend on the ability to decode text, andthat the recognition of letter/soundrelationships is only one of severaldifferent ways of deriving meaning fromtext, and of equal or even lesserimportance than other strategies that canbe applied, such as guessing fromcontext.15

Page

23

14 See, for example, Pressley (1998), and Beard (1993).

15 As a postscript to this section, it should be noted that new legislation recently introduced in the UnitedStates (the No Child Left Behind Act of 2001, which was signed into law on 8 January 2002), places particularemphasis on the improvement of reading instruction, with one billion dollars allocated to Reading First, aprogram designed to improve reading achievement through the adoption of effective teaching practicesbased on scientific research, as documented in the report of the National Reading Panel.

The evidence relating to the factorsunderlying the acquisition of readingskills is based on research in a numberof English-speaking countries, includingAustralia and New Zealand, as well asvarious other non-English speakingcountries. It can therefore be assumedthat the findings based on thisinternational body of research areapplicable to the Australian context.

In view of this it would seem appropriateto examine current practices in Australiato determine the extent to which thesepractices reflect current understandingsof the processes underlying theacquisition of reading, and what isknown about instructional strategies that are most effective in developingreading skills.

Historical OverviewOver the past 30 years, literacy teachingin Australia has been based mainly onthe principles and practices underlyingthe whole language approach to theteaching of reading. As noted by theHouse of Representatives Committee on Employment Education and Training(1992), whole language has Australia-wide support and virtually all curriculumguidelines on primary school literacyteaching are based on this approach. Ithas provided the theoretical basis for theliteracy instruction of teachers in boththeir preservice and inservice training,including the influential Early LiteracyInservice Course (ELIC), which, as notedin the House of Representatives Report,

was undertaken by virtually all teachersof early literacy throughout Australia.16

In their review of literacy instruction in Australian primary schools, van Kraayenoord and Paris (1994) tracethe historical developments that haveinfluenced literacy education inAustralia, and the emergence of theholistic approaches to literacy that haveled to a common approach to theteaching of literacy in classrooms acrossAustralia. The essential characteristic of aholistic approach to language learning,as identified by van Kraayenoord andParis, is that it does not focus instructionon component skills or delayintroduction of literacy uses untilsubskills have been mastered. Accordingto van Kraayenoord and Paris,‘communication of meaning is the maingoal, and Australian educators believethat invented spelling, incompletedecoding, text memorisation and other‘errors’ in performance should beaccepted in the child’s developmentaluses of literacy’ (van Kraayenoord andParis, 1994, page 219). It is this belief thatdistinguishes holistic and wholelanguage approaches to the teaching ofreading from approaches based on skill-based or direct teaching models.

The dominance of a single approach tothe teaching of reading in Australia isattributed by van Kraayenoord and Paristo two main influences. First, thepreservice and inservice training ofteachers, which in Australia tends to bevery uniform because the same sources

Page

24

CURRENT APPROACHES TO THE TEACHINGOF READING IN AUSTRALIA

16 A more detailed analysis of the philosophy and practice of whole language teaching in Australia and NewZealand is provided by Hempenstall (1996), and Tunmer and Chapman (2002).

of ideas and exemplary models are usedby most institutions. And second, theinfluence of a few key individuals, who,as a result of the relatively smallpopulation and vast distances, have hada powerful influence in promotingparticular models of literacy instruction.Among those mentioned by vanKraayenoord and Paris as having had akey role in influencing approaches to theteaching of literacy in Australia are MarieClay (1972), Brian Cambourne (1988),Don Holdaway (1979), Donald Graves(1983), and Michael Halliday (1975), all of whom espoused holistic and wholelanguage approaches to the teaching of reading.

While there seems to be an increasingrecognition in Australia of theimportance of phonemic awareness andthe need for more balanced approachesto the teaching of reading, there is as yetlittle evidence of a major shift from theholistic and whole language approachesthat continue to dominate literacyteaching practices in Australia. While the terms ‘balanced’ and ‘structured’ havebeen applied to describe some of thenew literacy programs and initiatives thathave been developed to address theliteracy needs of children in the earlyyears of schooling, these programs donot generally include direct systematic or explicit teaching of phonics. Theseprograms would therefore correspond towhat Foorman et al (1998), in their studyof the role of instruction in learning toread, describe as the implicit or indirectapproach, in which instruction in thealphabetic code is seen as incidental tothe act of making meaning from print(see also, Pressley, 1998, for a descriptionof whole language versus balancedprograms of literacy instruction).

Evidence Relating to the Use ofPhonics in Australian Programs ofReading Instruction Precise information on the extent towhich systematic teaching of phonics isincluded as a part of the regular teachingprogram in Australian schools is difficultto locate. In the case of New SouthWales, Center, Freeman and Robertson(1998) note the lack of uniformity in earlyliteracy instructional practices in thekindergarten year, and the fact that morerecently trained teachers tend to use awhole language approach, because thisis the philosophy that has been adoptedby most teacher training institutions overthe past 15 years. They observe that moreexperienced teachers are more likely touse a more phonically based approach,but because early literacy teaching doesnot occupy a large part of a teacher’spreservice training, many recentlygraduated teachers tend to learn on thejob, adopting the methods that arecharacteristic of the specific school towhich they are appointed.

Some evidence on the literacy programsprovided to support students withlearning difficulties is available from thesurvey of schools undertaken as a part ofthe DETYA funded study on primarystudents with learning difficulties inliteracy and numeracy (Rohl, Milton andBrady, in Louden et al, 2000). While thefocus of the study was on students withlearning difficulties, the results of thissurvey do provide some information thatis indicative of the nature of the teachingprograms provided by schools.

Data from the 377 schools thatresponded to this survey indicated thatthe most widely used programs tosupport the literacy learning of studentswith learning difficulties are First Steps

Page

25

(49 per cent), Reading Recovery (46 percent), and the Victorian Early YearsLiteracy Program (19 per cent). Inaddition, 32 per cent of schools indicatedthat they were using some form of directinstruction. The direct instructionprograms listed included somephonics–based programs such asTHRASS; the Macquarie UniversityReading Program, SWELL; and theHeather Harvey Intensive ReadingProgram. However these programs werementioned by less than one per cent ofthe schools that responded to the survey.

From these responses it is clear that thepreferred approach to the support ofstudents with reading difficulties areclassroom and intervention programswhich are based essentially on a wholelanguage approach to the teaching ofreading (that is, meaning emphasis asopposed to code emphasis approaches,where the teaching of phonics is implicitrather than explicit). This includesReading Recovery, which is based on amore intensive one-to-one version ofwhole language teaching strategies.17

Given that students with learningdifficulties are the group most likely torequire direct instruction in readingskills, the overall total of 32 per cent ofschools providing this form of teachingfor students who have been identified ashaving a learning difficulty is likely toprovide an overestimate of the extent towhich direct instruction is used in regularclassroom teaching. It should also benoted that the kinds of programsincluded under the general term ‘directinstruction’ are not further specified, andthat this term probably covers a widerange of different teaching strategies with

varying degrees and types of instruction,which may not necessarily include directteaching of the alphabetic code.

In addition to the data from the schoolsurvey, the DETYA funded study onprimary students with learning difficultiesalso collected information from systemsacross all states and sectors on theprograms and strategies that were beingimplemented in each system to assiststudents with learning difficulties. A listing of the main programs andstrategies mentioned by systems isprovided by Rivalland and House (inLouden et al, 2000, pages 138–140). While not exhaustive, this listing is said toprovide a general picture of the patternsof support provided throughout Australia.

A total of 164 programs and strategieswere listed. Of these, 27 (16 per cent)referred to programs which had a focuson phonics. An analysis of this listingindicated that phonics-based programswere more likely to be included amongthe programs and strategies listed by theindependent sector (30 per cent of the 50 programs listed), as compared withthe Catholic sector (14 per cent of the 51 programs listed) or the Governmentsector (8 per cent of the 63 programslisted).

These results, while indicative only, arenevertheless consistent with theevidence from other sources that wholelanguage approaches to the teaching of reading still dominate the Australianscene, and while specific programsbased on alphabetic code instruction arebeginning to be used, particularly inindependent schools, these remain theexception rather than the rule.

Page

2617 See, for example, Chapman, Tunmer and Prochnow (1999), and Chan and Dally, in Louden et al, (2000),page 214–215).

There have been few studies in Australiawhich have been designed specifically to look at the effectiveness of a phonics-based versus a whole language approachto the teaching of early reading. Somerelatively small school-based studies ofthe effects of direct instruction programsthat include specific teaching of phonicshave been reported, but these havegenerally been based on very smallsamples, and results are generallyreported in terms of actual gains versusexpected gains based on norm-referenced measures, without anycontrol to serve as a comparison group(Lockery and Maggs, 1982). However,some data relating to the effectiveness of more structured programs includingsome explicit teaching of phonics isavailable from the ACER research-basedevaluation of the Victorian pilot projecton multi-age grouping (de Lemos, 1999),while a longitudinal study of theeffectiveness of training in phonemicawareness at the preschool level hasbeen reported by Byrne and Fielding-Barnsley (Byrne, 1998). Furtherinformation on the effectiveness of aprogram which includes specificinstruction in sound/symbolcorrespondences and phonemicawareness is also provided by theevaluation of the Schoolwide EarlyLanguage and Literacy Program (SWELL),which was developed by a group ofMacquarie University researchers incollaboration with researchers at theJohn Hopkins University in the UnitedStates (Center, Freeman and Robertson,1998).

Data from the Victorian Multi-ageProjectThe ACER research-based evaluation ofthe Victorian pilot project on multi-agegrouping was designed to examine theimpact of multi-age grouping in the firstthree years of schooling on children’ssocial development and academicachievement. For the purposes of thisstudy teachers in both the pilot sample(with multi-age classes) and in thecontrol sample (with predominantlysingle-grade classes) were asked tocomplete a questionnaire on variousteaching and class grouping practicesadopted in their teaching program. Twoof the questions in this questionnairerelated to the teaching of literacy.

The first of these questions askedteachers of Prep to Year 2 children toindicate whether their approach to theteaching of literacy was based on awhole language approach or a structuredapproach. Of the 272 teachers whoresponded to this question 77 per centindicated that their teaching of literacywas based on a whole languageapproach, 17 per cent indicated that theirprogram was based on a balanced or amixed approach, incorporating elementsof both the whole language approach anda structured approach, while 6 per centindicated that they followed a structuredprogram, based on the identification andsystematic teaching of the subskillsunderlying the development of readingand writing. There were somedifferences between the teachers in thepilot and control schools in terms of theproportion who followed a wholelanguage approach (87 per cent in the

Page

27

AUSTRALIAN RESEARCH ON EFFECTS OF PHONICSVERSUS WHOLE LANGUAGE INSTRUCTION

multi-age group and 69 per cent in thecontrol group) and the proportion whofollowed a structured approach (3 percent in the multi-age group and 9 percent in the control group).

The second of the two literacy questionsasked teachers about their approach tothe teaching of phonics. In response tothis question, 51 per cent of the teachersindicated that there was no specificteaching of phonics in their program, 22 per cent indicated that they includedteaching of phonics as and whennecessary (implicit phonics), while 27 percent of teachers indicated that theyincluded systematic teaching of phonicsas a part of their teaching program.Again there was a difference betweenthe pilot and the control schools interms of the proportion of teachers whoindicated that their program did notinclude any specific teaching of phonics(70 per cent in the case of the pilotschools as compared with 36 per cent inthe case of the control schools).

Data from this project also providedinformation on the relationship betweenteaching program and student outcomesin reading, since all the children in theevaluation study were assessed in theirfirst year of school, and again at the endof their first, second and third years ofschooling, on various measures of earlyliteracy and reading skills.

In order to examine the effect ofteaching program on student outcomes,a school measure based on overallteacher responses to these twoquestions was created, distinguishingbetween schools which in generaladopted a whole language approach tothe teaching of literacy with little or noemphasis on the teaching of phonics,and schools which adopted a more

structured approach to the teaching ofliteracy with some emphasis on phonics.Comparisons of student outcomes inthese two groups indicated little or nodifferences in the first year of school onthe measures of early literacy, with thewhole language group scoring higher onthe Language Profile of the AGS EarlyScreening Profiles administered in thesecond term of school (effect size of -.12), and the structured program groupscoring higher on the teacherassessment of reading skills at the end ofthe first year of school (effect size of .07).There was also little or no differencebetween these two groups on themeasures administered in the secondyear of schooling, with the wholelanguage group scoring higher on aWord Recognition test administered inthe third term of school (effect size of -.05), and the structured program groupscoring higher on the teacherassessment of progress in reading at theend of the school year (effect size of .04).However, in the third and fourth years ofschooling the structured program groupscored higher than the whole languagegroup on the standardised measures ofreading comprehension administered inthe third and fourth years of schooling(with an effect size of .38 in Year 2 and aneffect size .30 in Year 3). These findingsof a long term positive effect for a morestructured program and systematicteaching of phonics in the early years ofschooling are consistent with theoverseas evidence.

While not nationally representative, thedata from this sample of schools providesome indication of the relative emphasison whole language versus structuredapproaches to the teaching of reading inwhat is probably a fairly typical sample ofAustralian schools, and the extent toPage

28

which explicit teaching of phonics isincluded as a part of the regular classprogram. It also provides some data onthe effects of more structured versus lessstructured programs on studentoutcomes. These data were collected in1995, so it is possible that there mighthave been some shift in emphasis as aconsequence of the increasingrecognition of the role of phonics andphonemic awareness in the early stages oflearning to read. However, this study doesprovide a basis for examining changes inattitude to the systematic teaching ofphonics over the last half decade.

Data on the effects of training inphonemic awareness at thepreschool levelAustralian data on the effectiveness oftraining in phonemic awareness at thepreschool level is available from theseries of studies by Byrne and Fielding-Barnsley (Byrne, 1998; Byrne andFielding-Barnsley, 1991a, 1991b, 1993,1995; Byrne, Fielding-Barnsley and Ashley,2000). In this carefully designedlongitudinal study samples of children infour preschools were randomly allocatedto two groups, one of which underwenttraining in phonemic awareness, usingthe Sound Foundations program (Byrneand Fielding-Barnsley, 1992). A controlgroup was exposed to the same set ofmaterials for the same amount of time,but instead of training in phonemicawareness they were given practice inclassifying items from the set of materialson attributes such as shape, colour,animacy, and edibility. The experimentalgroup comprised a total of 64 children,with 62 children in the control group(after two children in the initial controlsample left the region). There were somedifferences between the two groups on

the various measures administered priorto the training sessions. The controlgroup scored marginally higher on thePPVT (effect size of -.10), the ability toname letters (effect size of -.04) and therecognition of common signs such asMcDonald’s (effect size of -.08). Theexperimental group scored marginallyhigher on the rhyming task (effect size of .14) and the identification of letterscorresponding to a given sound (effectsize of .16). There was no differencebetween the two groups on the ClayConcepts of Print test. At the conclusionof the program the children in theexperimental group were found toperform better on measures ofphonemic awareness (effect sizes of .43 and .39 on measures of phonemeidentity and phoneme elision) and on a structured test of printed worddecoding.

At the end of their first year of schooling(Kindergarten) the children in theexperimental and control groups wereassessed again on the measures ofphonemic awareness, and also onmeasures of alphabet knowledge, wordidentification, pseudo-wordidentification and spelling. The childrenin the experimental group againperformed better on the measures ofphonemic awareness (effect sizes of .81 and .71 on measures of phonemeidentity and phoneme elision), and alsoon a measure of pseudo-wordidentification (effect size of 1.1). Theexperimental group also scored higheron the measures of real wordidentification (effect size of .18) andspelling (effect size of .29), althoughthese differences were not statisticallysignificant. There was however nodifference on the measure of alphabetknowledge (effect size of 0). Page

29

The superior performance of theexperimental group on various measuresof word reading, spelling, readingcomprehension and listeningcomprehension was still evident at theend of Year 1 and at the end of Year 2. At the end of Year 1 the differencesbetween the experimental and controlgroups were significant on the pseudo-word reading task (effect size of 1.27),and were substantially higher (but notstatistically significant) on the measure ofreal word reading (effect size of .67). Thedifferences were however less markedon the measures of spelling (effect sizesof .03 to .15). At the end of Year 2 thedifferences between the experimentaland control group were significant onthe pseudo-word reading task (effectsizes of .98 and .95), and substantiallyhigher (but not statistically significant)on the measures of reading real words,both regular and irregular (effect sizes of.55, .46 and .60), and on the measures ofreading comprehension and listeningcomprehension (effect sizes of .69 and .40).

A further follow-up of these children wasundertaken at the end of Year 3 (Byrne,1998) and again at the end of Year 5(Byrne, Fielding-Barnsley and Ashley,2000). The Year 3 follow-up was based on105 of the original 126 children (57 in theexperimental group and 48 in the controlgroup), while the Year 5 follow-up wasbased on 103 children, one less in eachof the groups.

At the end of Year 3 the differences weresignificant on the pseudo-word readingtest, but not on the other tests includingmeasures of irregular word reading,reading comprehension and listeningcomprehension, and a measure of printexposure (the Title Recognition Test,adapted from Cunningham and

Stanovich, 1990). Standard deviations arenot reported for these results, so effectsizes cannot be calculated for thesedifferences.

At the end of Grade 5, the experimentalgroup scored significantly higher thanthe control group on three of the eightmeasures administered: the Word Attacksubtest from the Woodcock ReadingMastery Tests – Revised (Woodcock,1987), the Castles irregular word readinglist (Castles, 1993), and the combinedmeasure on the three word reading lists(nonwords, regular words and irregularwords). The effect sizes for thesedifferences were .34, .39 and .33.Although the experimental group alsoscored higher on the other testsadministered (word identification; nonword and regular word reading; Spelling;and Title Recognition), these differenceswere not statistically significant.

These results have been quoted at somelength because they indicate substantialand lasting effects of phonemic trainingat preschool level on subsequent readingskills, without any further differentialtraining between the experimental andcontrol groups at school level. Thisfinding is consistent with the findingsreported by the US National ReadingPanel, and support the accumulatingevidence that phonemic awareness isone of the critical factors that underliechildren’s success in learning to read.

Evaluation of the Schoolwide EarlyLanguage and Literacy Program(SWELL)The only other Australian evaluation of a program that includes systematicinstruction in phonics that could belocated is the evaluation of the SWELLprogram in NSW (Center et al, 1998,Page

30

2001a; 2001b). This evaluation wasundertaken in 1995 to 1996, and wasbased on a sample of six schools, all ofwhich qualified for DisadvantagedSchools funding on the basis of parentaloccupation. Two of these schools hadstarted implementing the SWELLprogram in Kindergarten and Year 1 theprevious year (1994), and four startedimplementing the SWELL program at the beginning of 1995.

The SWELL program is based onStanovich’s interactive-compensatorymodel of reading acquisition (Stanovich,1980, 1984), and comprises three phases.The first phase, the Emergent LiteracyProgram, is designed to develop earlyconcepts of literacy in the first three to six months of the kindergarten yearthrough traditional top-down processes:that is, through exposure to a literaturerich environment, including story-tellingand retelling; Big Book activities (readingtogether); development of expressiveand receptive language abilities; activitiesdesigned to develop sensitivity tophonological and syntactic awareness;and other early writing activities (such asusing symbolic representation as indrawing, scribble, letters and words (withinvented and conventional spelling) forcommunication of ideas. The secondphase of the program, Becoming Literate,is introduced from about the middle ofthe second term of school, andcomprises a formal literacy instructionprogram which supplements theongoing emergent literacy activities, andfocuses mainly on bottom-up processes,including sound/symbolcorrespondences (phonologicalrecoding); phonological and phonemicawareness; and specific instruction inspelling and recognition of irregularwords. This phase continues to about the

end of Year 1, by which time mostchildren would be expected to havemastered the grapheme-phonemecorrespondences that are necessary forindependent reading, and the ‘learningto read’ stage, as defined by Chall (1983),has therefore been completed. The thirdphase of the program, Towards LiteracyCompetence, then returns to a top-downorientation, with a focus on developinglistening and reading comprehensionstrategies as children move into the nextstage of ‘reading to learn’.

For the purposes of the evaluation, thechildren in Year 1 in the year precedingthe introduction of the SWELL programformed the control group. Thesechildren, who had not had any exposureto the SWELL program in theirkindergarten year, were assessed at thebeginning of Year 1. Children who hadparticipated in the SWELL program in thekindergarten year were then assessed atthe beginning of Year 1 in the followingyear. This provided a control sample of249 children and an experimental sampleof 366 children. The comparison betweenthese two groups was then used as thebasis for evaluating the effectiveness ofthe SWELL program. The measures usedfor this purpose included the Burt WordTest, a standardised word recognitiontest, the Passage Reading Test, a measureof reading accuracy developed by Deno,Mirkin and Chiang (1982), and theExpressive Word Attack Skills Test, acriterion-referenced test of phonologicalrecoding developed at the SpecialEducation Centre at MacquarieUniversity (1991). The results of thisevaluation indicated that the children inthe experimental group, after 12 monthsin the SWELL program, scored higherthan the control group from the previousyear on all three measures, with these Page

31

differences being statistically significanton the Passage Reading Test and theExpressive Word Attack Skills Test, butnot on the Burt Word Test. Using the dataon mean scores and standard deviations(adjusted for school effects) as reportedby Center et al, the differences in termsof effect sizes were .08 on the Burt WordTest, .24 on the Passage Reading Test and.40 on the Expressive Word Attack SkillsTest. A further evaluation wasundertaken six months later, at thebeginning of the third school term, atwhich point the experimental group ofchildren had spent 18 months in theSWELL Program. By this time two schoolshad dropped out of the study, so thissecond evaluation was based only onthose children in the initial control andexperimental samples who remained atthe schools that were still participating inthe evaluation study, providing a sampleof 47 children in the control group and84 children in the experimental group.This second evaluation included thePassage Reading Test; the ExpressiveWord Attack Skills Test; theDevelopmental Spelling Test, developedby Tangel and Blachman (1992) tomeasure spelling proficiency by takinginto account the number of phonemesrepresented and the level oforthographic representation; and theDiagnostic Reading Test (Waddington,1988), which is a standardised testrecommended for classroom use by theNSW Department of School Education.This latter covers word decoding, sight-word knowledge, understanding ofwords, comprehension of sentences, theuse of indirect cues, and obtainingmeaning from direct speech. The resultsof this second evaluation at 18 monthsindicated that the children in theexperimental group continued to scoreat a higher level than the control group

on all the measures used, and that thesedifferences were statistically significanton all tests except the Passage ReadingTest. Effects sizes calculated on the basisof the mean scores and standarddeviations provided (adjusted for effectsof school and 12 month measures)indicated an effect size of .24 on thePassage Reading Test, .58 on theExpressive Word Attack Skills Test, .40 onthe Diagnostic Reading Test and .52 onthe Developmental Spelling Test.

While this study was based on arelatively small sample of schools andstudents, and encountered somedifficulties in implementation, the resultsare nevertheless consistent with theoverseas research evidence whichindicates that explicit instruction in thealphabetic code leads to improvedoutcomes in reading achievement, withthe effect sizes obtained in this studygenerally comparable to those reportedin the United States literature.

Limitations of Training in PhonemicAwareness and Phonics-basedInstruction While the evidence indicates thatphonemic awareness is a strongpredictor of success in learning to read,and that explicit instruction in phonics isa more effective teaching strategy thanprograms which provide little or nosystematic instruction in phonics, thisdoes not mean that all children willsucceed in reading provided that theyhave acquired an understanding of thephonetic structure of words and havereceived explicit instruction in phonics.There will always be some children whowill have difficulty in learning to read,whatever instructional strategy isadopted.Page

32

In discussing the results of their six yearfollow up at Grade 5, Byrne et al (2000)point out that early training in phonemicawareness at the preschool level is not initself sufficient to prevent subsequentreading difficulties. That is to say, it doesnot have a vaccination effect in terms ofpreventing reading difficulties. Theyidentified a small number of children(nine out of 56) in their experimentalsample who were successful on thephonemic awareness training (that is,they achieved the pass criterion on thepost intervention measures of phonemeidentity and decoding), but neverthelessexperienced difficulties in learning toread, and by Grade 5 were performingbelow the 5 per cent cut-off used todefine children with a reading difficultyon one or more of the five reading testsadministered at the end of Grade 5.

On the basis of a more detailedexamination of this group of children,they identified what they termed a ‘rateof responsiveness’ or ‘learning rateparameter’, measured in terms of thetime it took the children to reach a stableunderstanding of phonemeunderstanding in the initial trainingsessions, as indicated by the Session ofLast Error measure (that is, the numberof sessions it took before the child madeno errors in the activities following theinitial training on the target phoneme).This measure was found to be a strongpredictor of subsequent readingprogress over and above the finaloutcome measure of phoneme identity.

On the basis of these findings Byrne et alpostulate a more general learningparameter, which determines not onlythe rate of responsiveness to the originalinstruction, but also the acquisition ofhigher level orthographic coding. That is,children who are slow to grasp ideasearly in reading development, eventhough they finally grasp them, are alsoslow to acquire other principles thatdetermine their rate of progress inlearning to read.18 Training in phonemicawareness or direct teaching of phonicsis therefore not in itself sufficient toovercome reading difficulties in the caseof children who are slow to respond orwho have underlying problems incognitive processing.

Page

33

18 It would seem likely that this general learning parameter is also related to speed of processing, which isassumed to be the main factor underlying individual differences in general reasoning ability or the g factorin intelligence, see for example Anderson (1992) and Jensen (1998).

This review has focused specifically on the research findings relating to theprocesses underlying the acquisition ofreading and the evidence of instructionalpractices that are effective in improvingreading outcomes. It has not addressed arange of other factors that are associatedwith literacy development.

This focus has been adopted for tworeasons.

The first reason is that it is seen asproviding a balance to the prevailingAustralian literature on literacydevelopment, in which the dominantemphasis has been on the social andcultural factors associated with literacydevelopment. This emphasis has led tothe assumption that in order to effectchanges in children’s literacydevelopment it is necessary either tochange the social and culturalenvironment of the child, or to adapt theschool environment to more closelyreflect the values and experiences of thechild’s home environment.

There would seem to be two major flawsin this approach.

One is that there is little evidence toindicate that any changes in the child’ssocial and cultural environment, of thekind that might be brought about in theshort term, would have any measurableeffect on literacy outcomes.

The second is that there would seem to be a logical flaw in arguing that poorliteracy outcomes, as measured inconventional terms, are associated withdiscrepancies between the values andexpectations of the school as comparedwith the values and expectations of the

home, and at the same time to argue thatthe way of improving literacy skills is toadapt the school environment to make itmore like the home environment. Whilethis might be seen as amisrepresentation of this argument,there is nevertheless an element of truthin it which needs to be addressed. At what point does one distinguishbetween respect for different culturalattitudes and practices, as well as theobvious need to make the schoolenvironment as friendly, accepting andsupportive as possible, and the need toensure that children are exposed to thekinds of learning experiences that aremost likely to develop the skills andconcepts that are necessary to achieveliteracy and numeracy?

The second reason for focusing onevidence relating to the instructionalstrategies that might improveachievement in literacy learning is thatthis is more directly relevant to the issueof how literacy outcomes might beimproved.

The teaching of reading is the specificresponsibility of the school. While homefactors clearly play a part in developingthe language skills that form the basis forlearning to read, and children enterschool with varying levels of skills andfrom varying kinds of background, theresearch evidence tells us that it ispossible to provide children withexperiences that will develop the kind ofskills that they need to be able to read,regardless of their current level of skilland their home background. While someflexibility in teaching approach is neededto cater for children who are at different

Page

34

RELEVANCE OF RESEARCH FINDINGSFOR TEACHING PRACTICE

levels, and whose rates of progress maydiffer, the research evidence indicatesthat the kinds of skill that underlie thereading process are universal, and applyto all children regardless of theirparticular cultural, language, or socio-economic background.

While parental values and attitudes mayvary, and while these do need to berecognised and accommodated in theschool program, most parents want theirchildren to learn how to read, and expectthat the school will teach them how toread.

Learning to read is only one aspect of the broader educational program. It isnevertheless an important and crucialpart of the program, and failure to learnhow to read can have long term negativeconsequences.

In reviewing the research evidencerelating to the effectiveness of differentinstructional practices on learningoutcomes, it is of interest to note thatthe effect sizes associated with effectiveclass teaching practices are in generalhigher than the effect sizes of broaderschool-related initiatives designed toimprove school effectiveness or to assiststudents who are having difficulty inlearning to read. For example, effectsizes associated with the reduction ofclass size are generally substantiallylower than the effect sizes associatedwith effective teaching practices, whilethe effectiveness (in terms of effect sizes)of one-to-one intervention programssuch as Reading Recovery have yet to beestablished on the basis of well designedstudies meeting rigorous scientificcriteria.

At the same time, it should be noted thatthe adoption of effective classroompractices is, in comparison with other

strategies such as the reduction of classsize or the implementation ofintervention programs such as ReadingRecovery, much more cost effective.

Teachers have to be trained andemployed. It costs no more to trainteachers to use effective teachingpractices than to train them to useineffective teaching practices. And itcosts no more to employ effectiveteachers than it costs to employineffective teachers.

From this point of view, the researchevidence relating to the extent to whichthe introduction of effective instructionalpractices can improve student outcomesis encouraging. While time andresources might be required to ensurethe implementation and ongoingevaluation of the impact of the teachingpractices that have been identified aseffective, there is good reason to believethat the outcomes will be positive.

At the same time, the evidence indicatesthat in some cases reading difficulties areassociated with underlying cognitivedeficits, and that in such cases studentswill require ongoing intensiveintervention and support to achievesatisfactory outcomes.

Page

35

There is increasing recognition in theeducational community that it is time tomove away from policies and practicesbased on philosophical beliefs aboutwhat should work, to evidence-basedpolicies and practices based on theresearch evidence as to what does work(see, for example, Masters, 1999).

In the area of reading literacy, thereappears to be a discrepancy between theresearch evidence as to ‘what works’, andthe teaching strategies that form thebasis of most current teaching programs.

In order to improve standards of literacy,it would seem that there is a need torethink current approaches to theteaching of reading literacy in Australia,and to implement an ongoing programof research to evaluate the effectivenessof different approaches and strategiesdesigned to improve student outcomes.

As noted at the beginning of this review,a range of new initiatives designed toimprove literacy learning has beenimplemented in Australia over the pastdecade. However, no systematicmechanism has been set in place toevaluate the effectiveness of theseprograms in terms of improving studentoutcomes.

The adoption of evidence-based policiesand practices in the area of literacydevelopment implies a greater emphasison the evaluation of programs designedto improve literacy outcomes, usingsoundly based research designs andvalid and reliable research tools. In thisway it would be possible to evaluate theextent to which the various educationalinitiatives introduced are meeting theirstated objectives, and whetheralternative (and possibly less costly)strategies might be equally if not moreeffective than those currently in place.

Page

36

THE WAY FORWARD

Adams, M. J. (1990). Beginning to Read:Thinking and Learning About Print.Cambridge, Mass.: MIT Press.

Ainley, J., Fullarton, S., Frigo, T., andOwen, J. (1999). Literacy Advance:Reviewing the Literature. Melbourne:Catholic Education Commission ofVictoria.

Ainley, J., and Fleming, M. (2000).Learning to Read in the Early PrimaryYears. Melbourne: Catholic EducationCommission of Victoria.

Anderson, M. (1992) Intelligence andDevelopment: A Cognitive Theory.Oxford, UK, Cambridge, MA: Blackwell.

Andrews, S. (1989). Psycholinguistics andreading acquisition: The argument fordecoding. New South Wales Journal ofSpecial Education, 10, 15–20.

Andrews, S. (1992). A skills approach:Optimising initial reading instruction. InA.J. Watson and A. M. Badenhop (Eds),Prevention of Reading Failure. Sydney:Ashton Scholastic.

Anderson, R.C., Hiebert, E.H., Scott, J.A.,and Wilkinson, I.A.G. (1985). Becoming aNation of Readers: The Report of theCommission on Reading. Washington,DC: National Academy of Education,Commission on Education and PublicPolicy..

Beard, R. (Ed.) (1993). Teaching Literacy,Balancing Perspectives. London: Hodderand Stoughton.

Bond, G. and Dykstra, R. (1967). Thecooperative research program in firstgrade reading. Reading ResearchQuarterly, 2, 5–142.

Bowey, J. A. (2000). Recent Developmentsin Language Acquisition and ReadingResearch: The Phonological Basis ofChildren’s Reading Difficulties. AustralianEducational and DevelopmentalPsychologist,17, 1, 5–31.

Britton, J. (1970). Language and Learning.Harmondsworth: Penguin.

Britton, J., Burgess, T., Martin, N.,McLeod, A., and Rosen, H. (1975). TheDevelopment of Writing Abilities 11–18.London. Schools Council Publications.

Burgess, C., Burgess, T., Cartland, L.,Chambers, R. Hedgeland, J., Levine, N.,Mole, J., Newsome, B., Smith, H., andTorber, M. (1973). UnderstandingChildren Writing. Harmondsworth:Penguin.

Byrne, B. (1998). The Foundation ofLiteracy: The Child’s Acquisition of theAlphabetic Principle. Hove, East Sussex,UK: Psychology Press Ltd.

Byrne, B., and Fielding-Barnsley, R. (1989).Phonemic awareness and letterknowledge in the child's acquisition ofthe alphabetic principle. Journal ofEducational Psychology, 81, 313–321.

Byrne, B. (1991a). The role ofphonological awareness in readingacquisition. Australian Journal ofReading, 14, 133–139.

Byrne, B., and Fielding-Barnsley, R.(1991b). Evaluation of a program to teachphonemic awareness to young children.Journal of Educational Psychology, 83,451–455.

Byrne, B., and Fielding-Barnsley, R. (1992).Sound Foundations. Artamon, Australia:Leyden. Page

37

REFERENCES

Byrne, B. and Fielding-Barnsley, R. (1993).Evaluation of a program to teachphonemic awareness to young children:A 1-year follow-up. Journal of EducationalPsychology, 85, 104–111.

Byrne, B., and Fielding-Barnsley, R. (1995).Evaluation of a program to teachphonemic awareness to young children:A 2- and 3-year follow-up and a newpreschool trial. Journal of EducationalPsychology, 87, 488–503.

Byrne, B., Fiedling-Barnsley, R. andAshley, L. (2000). Effects of preschoolphoneme identify training after six years:Outcome level distinguished from rate ofresponse Journal of EducationalPsychology, 92, 4, 659–667.

Cambourne, B. (1988). The Whole Story:Natural Learning and the Acquisition ofLiteracy. Auckland: Ashton-Scholastic.

Castles, A. (1993). Varieties ofdevelopmental dyslexia. Unpublisheddoctoral dissertation. MacquarieUniversity, Sydney. (As quoted by Byrneet al, 2000).

Center, Y., Wheldall, K., Freeman, L.,Outhred, L. and McNaught, M. (1995). Anevaluation of Reading Recovery. ReadingResearch Quarterly, 30 (2), 240–263.

Center, Y., Freeman, L., and Robertson, G.(1998). An evaluation of schoolwide earlylanguage and literacy (SWELL) in sixdisadvantaged schools. InternationalJournal of Disability, Development andEducation, 45 (2), 143–172.

Center, Y., Freeman, L., and Robertson, G.(2001a). The relative effect of a code-oriented and a meaning-oriented earlyliteracy program on regular and lowprogress Australian students in Year 1classrooms which implement ReadingRecovery. International Journal ofDisability, Development and Education,48 (2).

Center, Y., Freeman, L., and Robertson, G.(2001b). A longitudinal evaluation of theSchoolwide Early Language and LiteracyProgram (SWELL). In Slavin, R andMadden, N (Eds) Success for All:Research and Reform in ElementaryEducation. Mahwah, NJ: LawrenceErlbaum Associates, Inc.

Chall, J. (1967). Learning to Read: TheGreat Debate. New York: McGraw-Hill.(Second edition, 1983; Third edition,1996).

Chall, J. (1983). Stages of ReadingDevelopment. New York: McGraw Hill.

Chan, L.K.S. and Dally, K. (2000). Reviewof Literature. In Louden et al., Mappingthe Territory: Primary Students withLearning Difficulties: Literacy andNumeracy. Canberra: Department ofEducation, Training and Youth Affairs.

Chapman, J.W., Tunmer, W.E. andProchnow, J.E. (1999). Success in ReadingRecovery Depends on the Developmentof Phonological Processing Skills. NewZealand: Ministry of Education.

Clay, M. (1972). Reading: The Patterningof Complex Behaviour. Auckland:Heinemann.

Coltheart, M. (1980). When can childrenlearn to read – and when should they betaught. In T.G. Waller and G.E.MacKinnon (Eds) Reading Research:advances in theory and practice (Vol. 1,pages 1–30). New York, Academic Press.

Page

38

Crévola, C. and Hill, P. (2001). CLaSS:Children’s Literacy Success Strategy: AnOverview (Revised Edition). Melbourne:Catholic Education Office.

Crévola, C. and Hill, P. (1998). Initialevaluation of a whole school approach to prevention and intervention in earlyliteracy. Journal of Education for StudentsPlaced at Risk, 3 (2), 133–157.

Cunningham, A.E. and Stanovich, K.E.(1990). Assessing print exposure andorthographic skill in children: A quickmeasure of reading experience. Journalof Educational Psychology, 82, 733–740.

de Lemos, M. (1999). A research-basedevaluation of the Victorian First StepsPilot Project for the First Three Years ofSchooling. Melbourne: ACER.

de Lemos, M. (August, 1997). The literacydebate. Paper presented at the ACERThursday Forum, Melbourne.

de Lemos, M. and Harvey-Beavis, A.(1995). The Development andAssessment of Literacy: Factors Relatedto the Development of Literacy:Background Paper 1 for the NationalSchool Literacy Survey. Unpublishedpaper, ACER. Melbourne. .

Deno, S.L., Mirkin, P.K., and Chiang, B.(1982). Identifying valid measures ofreading. Exceptional Children, 49, 36–45.

Department of Employment, Educationand Training. (1991a). Australia’sLanguage: The Australian Language andLiteracy Policy. Canberra: AustralianGovernment Publishing Service.

Department of Employment, Educationand Training. (1991b). Australia’sLanguage: The Australian Language andLiteracy Policy. Companion Volume tothe Policy Paper. Canberra: AustralianGovernment Publishing Service.

Department of Employment, Education,Training and Youth Affairs. (1998).Literacy for All: The Challenge forAustralian Schools. Canberra: AustralianGovernment Printing Service.

Dixon, J. (1967) Growth through English.Huddersfield, Yorkshire: OxfordUniversity press for the NationalAssociation for the Teaching of English.

Flesch, R. (1955). Why Johnny Can’t Read.New York: Harper and Row.

Foorman, B.R., Francis, D.J., Fletcher, J.M.,Scatschneider, C. and Mehta, P. (1998).The role of instruction in learning toread: Preventing reading failure in at-riskchildren. Journal of EducationalPsychology, 90, 1, 37–55.

Freebody, P. and Gilbert, P. (1999)Research into language and literacy. InKeeves, J. and Marjoribanks, K. AustralianEducation: Review of Research 1965–1998.Camberwell: ACER Press.

Gough, P. (1995). The New Literacy:caveat emptor. Journal of Research inReading, 18, 2, 79–86.

Graves, D. (1983). Writing: Teachers andChildren at Work. Portsmouth:NH:Heinemann

Gunn, S. (compiler) (1999). Children’sLiteracy National Projects: ConferenceReport and Executive Summary,Researcher’s Conference, 1998. Canberra:Department of Employment, Education,Training and Youth Affairs (DETYA),Literacy Section of the Literacy andSpecial Programmes Branch

Halliday, M. A.K. (1975). Learning how toMean: Explorations in the Developmentof Language. London: Edward Arnold.

Page

39

Hempenstall, K. (1996). The wholelanguage approach to reading: Anempiricist critique. Australian Journal ofLeaning Disabilities, 3, 1, 22–32.

Holdaway, D. (1979). Foundations ofLiteracy. Sydney, NSW: Ashton-Scholastic.

House of Representatives StandingCommittee on Employment Educationand Training. (1992). The literacychallenge: Strategies for earlyintervention for literacy and learning forAustralian children. Canberra: AustralianGovernment Publishing Service.

Jensen, A.R. (1998). The g Factor: TheScience of Mental Ability. Westport,Conn; London: Praeger.

Juel, C., Griffith, P. and Gough, P. (1986).Acquisition of literacy: A longitudinalstudy of children in first and secondgrade. Journal of Educational Psychology,78, 243–255.

Keeves, J. and Marjoribanks, K. (1999).Australian Education: Review of Research1965–1998. Camberwell: ACER Press.

Lockery, M., and Maggs, A. (1982). Directinstruction in Australia: a ten-yearanalysis. Educational Psychology, 2, 3–4,263–288.

Louden, W., Chan, L.K.S., Elkins, J.,Greaves, D., House, H., Milton, M.,Nicols, S., Rivalland, J., Rohl, M., and vanKraayenoord, C. (2000). Mapping theTerritory: Primary Students with LearningDifficulties: Literacy and Numeracy.Canberra: Department of Education,Training and Youth Affairs.

Masters, G. (November/December 1999).Towards a National School ResearchAgenda. Keynote Address presented atthe joint NZARE/AARE Conference,Melbourne.

Masters, G. and Forster, M. (1997).Mapping Literacy Achievement: Resultsof the 1996 National School EnglishLiteracy Survey. Canberra: DEETYA.

National Reading Panel (2000). TeachingChildren to Read: An Evidence-BasedAssessment of the Scientific ResearchLiterature on Reading and ItsImplications for Reading Instruction.http://www.nichd.nih.gov/publications/nrp/report.htm (Posted 13 April 2000,accessed 15 April 2000, last modified 16May 2000).

Nelson, N. and Calfee, R.C. (1998). Thereading-writing connection viewedhistorically. In Nelson, N. and Calfee, R.C.(Eds). The Reading-Writing Connection.Chicago, Ill: National Society for theStudy of Education.

Nicholson, T. (2000). Reading the Writingon the Wall: Debates, Challenges andOpportunities in the Teaching ofReading. Palmerston North, NZ:Dunmore Press.

Pearson, P.D. (1999). A historically basedreview of Preventing Reading Difficultiesin Young Children. Reading ResearchQuarterly, 34 (2), 231–246.

Pflaum, S., Wahlberg, H.J., Karegianes,M.L. and Rasher, S.P. (1980). Readinginstruction: A quantitativeanalysis.Educational Researcher, 9, 12–18.

Pratt, C. and Coltheart, M. (1996). GuestEditorial, Australian Journal ofPsychology, 48, 3,

Pressley, M. (1998). Reading Instructionthat Works: The Case for BalancedTeaching. New York: The Guildford Press.

Page

40

Rivalland, J., and House, H. (2000).Mapping System Provision for LearningDifficulties. In Louden et al., Mapping theTerritory: Primary Students with LearningDifficulties: Literacy and Numeracy.Canberra: Department of Education,Training and Youth Affairs.

Rohl, M., Milton, M., and Brady, D. (2000).Survey of Schools. In Louden et al.,Mapping the Territory: Primary Studentswith Learning Difficulties: Literacy andNumeracy. Canberra: Department ofEducation, Training and Youth Affairs.

Rosen, C. and Rosen, H. (1973). TheLanguage of Primary School Children.Harmondsworth: Penguin.

Share, D.L. (1995). Phonological recodingand self-teaching: sine qua non ofreading acquisition. Cognition, 55,151–218.

Snow, C.E., Burns, S., and Griffin, P.(1998). Preventing Reading Difficulties inYoung Children. Washington, DC:National Academy Press.

Special Education Centre. (1991).Expressive Word Attack Skills Test(Revised). Sydney: Macquarie University,Special Education Centre.

Stanovich, K.E. (1980). Toward aninteractive-compensatory model ofindividual differences in thedevelopment of reading fluency. ReadingResearch Quarterly, 16, 32–71.

Stanovich, K.E. (1984). The interactive-compensatory model of reading: A confluence of developmental,experimental and educationalpsychology. Remedial and SpecialEducation, 5, 11–19.

Stanovich, K.E. (1986). Matthew effects inreading: some consequences ofindividual differences in the acquisitionof literacy. Reading Research Quarterly,21, 360–406.

Tangel, D. and Blachman B. (1992). Effectof phoneme awareness instruction onkindergarten children’s inventedspelling. Journal of Reading Behaviour,24, 233–261.

Tasmania, Department of Education andthe Arts. (1994). Learning to Read andWrite. Hobart: Department of Educationand the Arts.

Tunmer, W.E. and Chapman, J.W. (2002).The theoretical contexts of reading: HowChildren Learn to Read (and Why SomeDon’t). In P. Adams and H. Ryan (Eds.),Learning to read in Aotearoa NewZealand (pp. 51–64). Palmerston North,New Zealand: Dunmore Press.

Tunmer, W.E. (November/December,1999). Science Can Inform EducationPractice: The Case of Literacy. 1999Herbison Lecture, presented at the jointNZARE/AARE Conference, Melbourne.

van Kraayenoord, C.E. and Paris, S.G.(1994). Literacy instruction in Australianprimary schools. The Reading Teacher, 48, 3, 218–228.

Waddington, N.J. (1988). DiagnosticReading and Spelling Test. Ingle Farm,South Australia: Waddington EducationalResources.

Woodcock, R.W. (1987). WoodcockReading Mastery Test – Revised. CirclePines, MN: American Guidance Service.

Page

41

For some decades, world-wide, there have been national initiatives toimprove literacy rates and standards. During the same period,concentrated research studies have been undertaken to find out howbest to achieve the desired improvements. Two main thrusts inteaching and learning how to read and write have emerged, often incontroversy. One is generally known as the 'whole language'approach and the other concentrates more on instruction in phonics.What works? This paper focuses on the theoretical assumptionsunderlying these two approaches to the teaching of literacy, and thestudies which have been undertaken, in the international arena, tofind out how children progress, from their earliest educational years,in attaining both initial reading skills and lifelong literacy.

PRESS

9

ISBN 0-86431-584-8

780864 315847