Phonological skills and writing of presyllabic children

Phonological Skills and Writing of Presyllabic Children Author(s): Cristina Silva and Margarida Alves-Martins Source: Reading Research Quarterly, Vol. 37, No. 4 (Oct. - Nov. - Dec., 2002), pp. 466-483 Published by: International Reading Association Stable URL: http://www.jstor.org/stable/748262 Accessed: 02/03/2010 11:40 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ira. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. International Reading Association is collaborating with JSTOR to digitize, preserve and extend access to Reading Research Quarterly. http://www.jstor.org Reading Research Quarterly Vol. 37, No. 4 October/November/December 2002 C2002 International Reading Association (pp. 466-483) skills and writing Phonological of presyllabic children CristinaSilva Alves-Martins Margarida Instituto dePsicologia Aplicada,Lisboa,Portugal Superior U and the awareness of syllabic and intrasyllabicunits (Liberman,Shankweiler, Fischer, & Carter,1974; Sim-Sim, 1997; Treiman, 1992), may develop spontaneously over the course of the preschool years, whereas for most children proficiency in more complex abilities, such as synthesis or phonemic segmentation, which require phonemic awareness, occurs when they learn to read (Morais, 1994; Perfetti, Beck, Bell, & Hugues, 1987). The argument about the meaning of the relationship between phonological awareness and reading acquisition has been replaced gradually by a finer analysis concerning the type of minimum skills required to understand the alphabetic characteristicsof written language on the one hand and the definition of those that result from the process of learning to read on the other (Byrne, 1992, 1997; Goswami, 1998; Murray,1998; Stahl & Murray,1994, 1998). For example, works by Stahl and Murray(1994, 1998) tell us that the capacity to identify common phonemes in different words is a skill that is critical to children's understanding of the way in which written code functions, while the abilities to segment and blend phonemes essentially develop once the alphabet has been learnt. At the same time, the role phonological awareness plays in learning to read was redefined when it was conjugated with letter-name familiarityas a fundamental basis for understanding the alphabetic principle. The latter is defined as the understanding of the nature of the relationships between sounds and the letters in the written code. he discovery of the relationship between phonological awareness and the process of learning to read is one of the most important contributions of the last 30 years of research in the domain of the acquisition of literacy. Despite the significance of this discovery, for many years the texts on the subject focused exclusively on a debate about the meaning of the causal relationship between phonological awareness and learning to read (Goswami & Bryant, 1990; Morais, 1994). This theoretical discussion, in which oral analytical abilities were seen sometimes as a cause and sometimes as a consequence of learning to read, may well have arisen as a result of disparities between the evaluative tasks that were used in research in this field and the inconsistency in the way in which the concept of phonological awareness was operationalized. The concept of phonological awareness implies the existence of a general capability with multiple dimensions that influence the difficulty experienced in the various tasks. The best way to look at this capability is probably to see it as part of a continuum (Stanovich, 1992). Skills such as detecting and producing rhymes are considered to be on a lower level of the continuum as they require few analytical capabilities and simply an awareness of similar phonological sequences, while the abilities to segment and invert the phonemes in words are placed on a higher level because they imply an analytical attitude and an explicit representation of phonetic segments. Some skills, like the ability to detect rhymes 466 ABSTACT Phonological skills and writing of presyllabic children beTheobjectiveof thisstudywasto identifycausalrelationships tweenthe developmentof phonologicalabilitiesandprogressin knowledgeaboutwritingin preschoolchildren.Thestudylooked whoweredividedintothreegroups at 71 preschool-aged children, Thechildren wasgovernedbypresyllabic criteria. andwhosewriting in thefirstgroupweresubjected intervention that to anexperimental focusedon writingandwasintendedto leadthemto evolvein such Thesecondgroupwastheoba wayas to producesyllabicwriting. aimed at of syllabicunits.Thethirdset of phonological training ject children servedas a controlgroup.Allthreegroupswereat equivawitha similar number of letlentintellectual levelsandwerefamiliar ters.Thetwoexperimental intervention also programs provedequiv- alentin termsof theconceptual evolutiontheytriggered, to the exin bothexperimental tentthatthechildren groupsstartedwritingin withsyllabiccriteria; accordance thewritingof the childrenin the controlgroupcontinuedto be guidedby presyllabic criteria.In the in tests children the two phonological experimental groups achievedresultsthatrevealeda similardegreeof progress(greater thanthatattainedby the controlgroupchildren)in the initialclassification tests.Thesecsyllabledeletionandtheinitial-phoneme ondexperimental adgroupdisplayedmorestatistically significant deletiontestthaneitherof theother vancesin the initial-phoneme twogroups. Habilidades fonol6gicas y escritura en niflos pre-silabicos lasrelaciones causalesenElobjetivode esteestudiofueidentificar treel desarrollo de lashabilidades fonol6gicas y el progresode los en nifiospre-silibicos. Elestudio sobrela escritura conocimientos en tres examin6a 71nifiosen edadpre-escolar quefuerondivididos estabagobernada porcriterios pre-silibicos. gruposy cuyaescritura Losnifiosdel primergrupoestuvieronexpuestosa unaintervenci6nexperimental y tuvocomoprop6siquese centr6en la escritura to guiarloshaciala producci6nde escriturasilibica.Elsegundo fonol6gicocentradoen las unidades gruporecibi6entrenamiento silibicas.Eltercergrupode nifiosfuncion6comogrupode control. Lostresgruposteniannivelesintelectuales y estabanfaequivalentes miliarizados con unacantidadsimilarde letras.Losdos programas de intervenci6n tambienresultaron en experimentales equivalentes thrminos de la evoluci6nconceptual al puntode quepromovieron, comenzaron a esque los nifiosde ambosgruposexperimentales cribirde acuerdoa criteriossildbicos; la escritura de los nifiosdel Enlas grupode controlcontinu6guiadaporcriterios pre-silibicos. nifios los de ambos lopruebasfonol6gicas, gruposexperimentales unciertogradode progreso graronresultados querevelaron (mayor queel alcanzado porlos nifiosdel grupode control)en la omisi6n de silabainicialy en la clasificaci6n de fonemainicial.Elsegundo mostr6 mis signigrupoexperimental progresosestadisticamente en lapruebade omisi6nde fonemainicialquelos otrosdos ficativos grupos. Phonologische Fertigkeiten und das Schreiben von pra-syllabischen Kindern ZieldieserStudiewares, kausaleZusammenhdnge zwischender undFortschritten im von phonologischen Entwicklung Fihigkeiten zu identifizieren. WissenumdasSchreiben derVorschulkinder Die die in dreiGruppen Studierichtetesich auf 71 Vorschulkinder, durchpri-syllabische Kriterien eingeteiltundderenSchreibarbeiten bestimmt waren.DieKinderin dererstenGruppewurdeneinerexperimentellen Beeinflussung ausgesetzt,die sich aufsSchreiben unddazubestimmt daM konzentrierte anzuleiten, war,sie dergestalt sie syllabisches Die zweiteGruppewurde Schreiben produzierten. aufsyllabische dasObjektphonologischen Trainings, ausgerichtet Einheiten.Die dritte Auswahlan Kinderndiente als eine befanden sichaufvergleichbaren AlledreiGruppen Kontrollgruppe. intellektuellen Stufenundwarenmiteiner ihnlichenAnzahlvon Buchstabenvertraut.Die zwei experimentellenInterventions- 467 imSinnederkonzeptuellen erwiesensichalsebenbtirtig programme diesie hervorriefen, inbeiden insoweit,dagdieKinder Entwicklung, in mit Experimentiergruppen Ubereinstimmung syllabischen Kriterien zu schreibenbegannen;dasSchreiben derKinderin der wurde unter fortgesetzterAnleitungvon priKontrollgruppe Indenphonologischen Kriterien Testsersyllabischen weitergefdhrt. reichtendieKinderin beidenexperimentellen Gruppen Ergebnisse, die einen jeweils ihnlichenGradan Fortschritten offenbarten derKontrollgruppe erreicht (gr68eralsjener,dervon denKindern undbei der wurde),in deranfinglich-syllabischen Ausmerzung anfinglich-phonemischen Unterteilungder Tests. Die zweite in zeigtemehrstatistisch Experimentiergruppe auffilligeFortschritte demTestzuranfinglich-phonemischen alsjedederbeiAusmerzung denanderenGruppen. AIBSTRACT ,J?~~l3~~%~~~L~f~ ?1~5Ci >~ 7 2Dla,1L sI 71 At~ )~-c~~i0 fI tlIL?tko111-r 7, - 7 )# a?0 )VC --HfNitN:1 Y -Ik5 itAU ;5 W; 2111)znItQ` 7 h3 - kv -"*I, ?10,-l-o inLJ0R -rtf-7 Competences phonologiques et ecriture d'enfants pre-syllabiques l'?volution intellectuelle Cette etudeavaitpourobjectif d'identifier lesrelations causales entre concerne dansla qu'ilsontdclench@e, etleprogres ont mesure oiulesenfants desdeuxgroupes led&veloppement descompetences des phonologiques experimentauxcomrelatives l'acriture chezdesenfants tandisquel'ecriture connaissances d'agepresco- menc?aecrireavecdescriteres syllabiques, . dugroupe laire.L'etude a portesur71enfants entrois desenfants a descriteres a continue d'age repondre prescolaire divises contr6le . ' des dont Dans les tests les desdeux et 1'6criture enfants r6pondait criteres groupes pre-syllabiques. pre-syllabiques. phonologiques Lesenfants dupremier ont des atteint resultats manifestant desprogroupeontetesoumis. uneintervention groupes experimentaux surl'criturequiakt conduite defagon'a aceuxatteintsparles enfantsdu (sup6rieurs experimentale centr@e grescomparables aproduire Lesecondgroupe a uneecriture lesamener dansdestestsdemandant lasuppression delapresyllabique. groupe contr6le) faitl'objet d'unentrainement centre sur des unites ou du initial. Le second sylphonologique phoneme groupe mieresyllabe experimend'enfants a servidegroupe ensemble condesprogres tala presente labiques.Letroisieme statistiquement plussignificatifs quechaLestroisgroupes avaient unniveau intellectuel et cun des deuxautresgroupesdansle testde suppression du tr61e. equivalent avecunnombre delettres semblable. Lesdeux phoneme initial. etaient familiarises d'intervention se sontreve1es en ce qui equivalents programmes (1OHOJIOFHeCKHe HaBbIKH H HHCbMOy leTeHi, He 3HaIOIUHXAJCeJHHqIHa CjiorH B TOM,qTO6bIBbaIMBITI Tb pHHHHO-CJICACTBeHHbie IJenibHaCToiuero Hccine)oXBaHnJ COCTO5naU CBB3H H flPpeACTaBj1eHHI MeKAtypa3BHTHmeM y AOUKOJIbHHKOB. 4OHonorHecKHx CnOCO6HOCTeI o 0lHCbMe Ha 3a o6uMHM IHCJIOM H4ccnexaoBaHHe 6a3HpoBanioc Ha6nIO•eHHH TpeMtsrpynnaMn XOIKOJIbHHKOB 71 qenoBeK,qbe fHCbMOOCHOBdBaJIOCb Haa;OcJIOrOBOM iOAxoAe. eTHB nepBoi rpynne 6bumn no;aepriyTbMi 3KcepBmeHTaJIHOMy BemaTenbcTBy, HanpaBneHHOMyHa nIHCMO H rpynna 6bina npeHa3HaqeHHOMy AI nojrOTOBKHaeTeii K CHJan6HqeCKOMy nHcbMy.BTOpaW HaAH o6,eKTOM 4oHonoriOmecKoro o6yenHRa, nauejiennoro 4epeHunnaiumo cHIIa6HqecKHIX Bce TpHrpynnb iMeJIHOaHHaKOBbIi Tpemsi rpynna aeTei 6Ema KOHTpOJIbHOH. eanHHHL. ypoBeH]H 3HaJIH OAJHaKOBOe KOJIHieCTBo 6yKB. 06e 3KCiiepHMeHTaJILHble nHHTenneKTyaJhHbmI HHTepBeHIHOHHble nporpaMMbL TaI•e 6un paBHno3HnatHbIB HiiaHe 3BOJOKHH nIOIHIATITHi, KOTOpble B OHHpa3BHBaJIH. B HTore AeTH B o6eHX 3KCIiepHMeHTaJibHblX rpynniax Hala1HInica B TOBpeMA KaKAITH H3 KOHTpOJlbHOOi COOTBeTCTBHH C CHUIa6HieCKHMHKpHTepHSIMH, rpylnnl• npoAojDKaI pyKOBO;CTBOBaThCS upinuHCbMe aocjIOrOBbNMH KpHTepH[MH.B KIaccH4HKaIu4HOHHbIX TeCTaXAeTH H3 AByX 3KCiepeHMeHTaJbHbIX rpynn He OnyCKaJHHalaJibHbiai Cam OF rH onpeaeMiH HaqajibHyIo OHemy OaHHaKOBO ycIelHIHO(IIOKa3aB6oJiee BbICOKHe pe3yiqETaTbI, ieM aeTH H3 KOHTpOJIbHOirrpyllnil). BTopas rpynna noKa3ana6onee CTaTHCTHwiecKH3HatHMylo TeCTOBBbSIBJIeHI4OHatlailbHOi (OHeMbi, leM OCTaJIbHbIe AHHaMHKy IIpH BbIrImOJIHeHHH rpynnb. (OHojIorHIecKHX 468 Byrne(1998)showed thatthe awarenessof phonewith lettersdo not of themselves mic unitsand familiarity will a child that manageto make use of the alguarantee in real contexts,but ratherthat reading phabeticprinciple it is the combinationof the formertwo thatprovescrucial to the discoveryof the latter.In turn,understandingthe alphabeticprincipleenableschildrento gain access to decoding proceduresthatformpartof the readingactivity. Phonemicawarenessis thus now seen as a skillthat is necessaryto obtain,albeitof itselfinsufficientto ensure, a full conceptualunderstandingof the alphabetic principle;childrenmay possess an adequatenotion of the phonemicstructureof oral speech withoutthis in itself leadingto an understandingof the alphabeticnature of writtenlanguage.This understandingnot only requires phonemicanalyticalskills,but also meansthatchildren must possess a concept of writingthatleads them to representationof the phonemicstructureof words. In the 1970s,Read(1971, 1975)was alreadycalling attentionto the creativespellingof preschool-agechildren,showing thatthey are not all at the same level in termsof the way in which they look at writtenlanguage. Sincethen, a largenumberof works (Ferreiro,1988; Pontecorvo& Orsolini,1996;Read,1986;Sulzby,1989) have shown thatan understandingof the abstractrules thatunderliethe organisationof alphabeticsystemsis a processthatbegins earlyon, via the informalcontacts thatchildrengraduallymake with writtenlanguage.In theireffortsto understandthe meaningsof graphicmarks and to interactwith others(both peers and adults),children ask themselvesquestionsaboutthe correspondences between objectsand writingand aboutthe relationshipsbetween the oral and the writtenformsof language.In this way they build up unconventionalideas aboutthe propertiesof writingand what it represents. They constructa series of conceptualhypothesesthatcan be more or less close to the real alphabeticsystem. Ferreiro'swork (1988) on preschoolchildren'sinvented spellingssuggeststhatchildren'sknowledgeof writtenlanguageevolves throughthree essentiallevels of conceptualisation.The firstlevel can be characterisedby the searchfor criteriathatmake it possible to differentiate between drawingsand writtenlanguage,and by the gradualperceptionthata sequence of lettersconstitutes an object that stands in for the real thing. In parallel with this differentiation process, the child also elaborates criteria that make a series of letters into something that can transmit a message. He or she considers that there must be a minimum number of letters for it to be possible to read and write a message, and also that the letters must vary. The latter criterion leads him or her not to use the same sequence of letters to spell different words. Phonological skills and writing A second level involvesrefiningthe formsof qualitaof the orderof lettersin his or her attive (diversification (the minimumnumber temptsat writing)and quantitative differenof lettersneeded to makewritingunderstandable) tiationbetweenchainsof letters,in such a way as to ensuredifferencesin how differentwordsare represented. The ideas thatchildrenformaboutwritingreflect the imagesthatthey build up fromthe examplesthey see aroundthem, in which words are genericallycomposed of variouslettersand those lettersare set out in different sequences for differentwords. It is these merelygraphic characteristics thatchildrenseek to take into accountin theirinventedspellings.At these levels childrenhave not yet establishedany relationshipbetween oraland written language.Genericallyspeaking,it is possible to call these firsttwo levels presyllabic. On a thirdlevel childrenbegin to relateoral language to writtenlanguage.This level consistsof a numberof stages,beginningwith the searchfor correspondencesbetween lettersand the syllabicsegmentsin words. In theirinventedspellingschildrenwritea letter with which to representeach of the syllablesin a word, even thoughthey establishthe letter/syllablecorrespondence on a purelyrandombasis-syllabic writingwithout phonetization. Subsequentlythey establishthis correspondenceusconventional lettersto representone of the sounds in ing a given syllable-syllabic writingwith phonetization.This syllabicwritingis particularly apparentin languagessuch as Spanishand Portuguese.The frequencywith which this type of writingappearsin Portugueseprobablyis explainedby its structurein which thereare manypolysyllabic words and the syllabicstructurethatpredominates is one of open syllablesof the consonant/voweltype (Andrade& Viana,1993;Vigario& Fale, 1993). Laterstill,childrenbegin to analyseoral languagein a way thatgoes beyond the syllabiclevel. Thisgives rise to inventedspellingsin which childrenrepresentall the phonemes in some of the syllablesof a word yet continue to use single lettersto denote othersyllablesin the same word-syllabic-alphabeticwriting. The thirdconceptuallevel culminatesin an understandingof the alphabeticnatureof writtenlanguagereflected in inventedspellingsin which the phonetic structureof the word is fullylearntand codified,even though not all the applicable orthographic conventions are respected-alphabetic writing. This evolutionary path, which has been identified in relation to a variety of languages-French (Besse, 1993, 1995, 1996; Chauveau & Rogovas-Chauveau, 1994; Fijalkow, 1993), Portuguese (Alves-Martins,1994; AlvesMartins& Mendes, 1987), Italian (Pontecorvo & Zuchermaglio, 1988, 1995), Hebrew (Tolchinsky, 1995; 469 Tolchinsky& Levin,1988),and English(Sulzby,1989)has not, however,been takeninto accountin the majority of researchinto phonologicalawareness. This developmentalpath does not requirepassing throughall the variousstages.Theremay even coexist formsof writingthatimplydifferentways of lookingat the writtencode (Besse, 1995;Sulzby,1986).Sulzby,for example,considered"thatthereis no one identicaldevelopmentalsequence in the way in which childrenuse the writtensystem"(p. 70). Inasmuchas certainwrittenforms appear,disappear,and reappearduringthe courseof the processof acquiringwriting,she arguedfor the existence of a developmentaldialecticthatis both continuousand discontinuous. Thereare pointson this evolutionarypaththatare similarto the stages in the developmentof writingput forwardby Hendersonand Beers(1980) and Gentry (1982).The latteridentifiedfive stages in children's spelling.Duringthe firstor precommunicative stage, children randomlycombinelettersand pseudoletterswithout payingattentionto correspondencesbetween lettersand sounds.Therewould appearto be similaritiesbetween this initialstage and the firsttwo levels definedby Ferreiro(1988)-that is to say the presyllabiclevels. Duringthe second or semiphoneticstage, children begin to phoneticallyrepresentsome of the components in words by choosing letterswith which they are familiar. The lettersthatare employedin this way may represent both sounds and syllablesin the word in question. This second stage has similaritiesto Ferreiro's (1988) initialphases of the phonetizationof writingnamelythose thatinvolve syllabicwritingwith phonetization, in which childrenuse conventionallettersto representone of the sounds in a syllable,and those that writing. correspondto syllabic-alphabetic The thirdor phoneticstage is characterisedby a systematiccorrespondencebetween lettersand sounds, but withoutany respectfor orthographicconventions.It is thus equivalentto the alphabeticwritingnoted by Ferreiro(1988). The next two stages formthe backdropto children'sprogressin relationto theirlanguage'sorthography-aspects thatFerreiro(1988) did not address.In this transitionalphase childrenbegin to take accountof the fact that the way in which they are supposed to write may be influenced by morphemic factors. Finally they reach the last stage of the process, which is to say that they write correctly. To the extent that it constitutes a form of oral language analysis, preconventional writing introduces a metalinguistic practice that undoubtedly has some important consequences in learning the oral segments of words. These consequences probably are enhanced by the 470 processof phonetizingwriting,in which children'sattemptsto matchthe lettersin and the sounds of words constitutea concretesupportfor the gradualidentification of the successivephonemesthatformpartof the words.Variousauthors(Stahl& Murray,1998;Treiman, 1998)consideredthatchildren'sinventedspellings,in which a few conventionallettersrepresentsounds, stimulate phonemicawareness.Withinthe contextof invented writing,the activationof lettersseems in turnto be govin the erned by a word'sphonologicalcharacteristics, sense thatit is easierwhen the words containphonetic sequences thatactuallyrepresentthe namesof the letters (Mann,1993;Quintero,1994).Set againstthis backactivitiesmay lead to a processof ground,writing-related or appropriating constructingthe alphabeticprinciplevia a complex developmentalinteractionbetween a child's capabilityto segmentwords into phonemesand the use of the graphicsupportunderlyingthe lettersin orderto representthem. Consequentlychildren'spreconventional writingmay constitutea way in which to organiseand expand oralanalyticalcapabilitiesand promotean even more preciseunderstandingof the relationshipsbetween the oraland the graphicelements.This point of view can be confirmedby Richgels'sstudy(1995),which showed thatpreschool-agechildrenwho were classifiedas being good at phonetizingin theirinventedwritingfound it easierto readwords thatwere simplefroma phonetic standpointthan did childrenwhose writtenoutputwas not as rich. Along similarlines, variousauthors(Adams,1998; Mann,1993;Treiman,1998)began to view children'sinvented spellingsas an understandingof the alphabetic principle,but the knowledgeaboutwrittencode that childrenrevealin theirattemptsat writinghas often been seen as justone more indicatorof phonologicalawareness (Mann,1993;Vale & Cary,1998).Consequently,few studies(Alvarado,1998;Pontecorvo& Orsolini,1996; Vernon,1998)have assessedthe relationshipbetween the evolutionof knowledgeaboutwritingon the one hand and progressin phonologicalskillson the other. Alvarado(1998) tested childrenat differentconceptual levels in writingby givingthem the task of deleting the initialphoneme in words presentedin both oraland written form. As expected, children at the alphabetic level achieved the highest success rates-around 50%in the oral format and 95% in the written format. The children at the syllabic-alphabetic and syllabic with conventional sound correspondence conceptual levels attained similar degrees of success-60% and 57%respectively-when the words were presented to them in conjunction with the written form. However, these levels fell significantly-to 16%and 8%---inthe exclusively oral format. READING RESEARCH QUARTERLY October/November/December 2002 37/4 The comparison between children whose writing is governed by letter/syllable correspondence on a purely random basis and those who choose conventional letters with which to represent one of the sounds in a syllable suggests the importance of the role of conventional letters as a factor that increases phonemic awareness. The former group displayed success rates of only 15%and 7% (with and without the written version) in the phonemic test. Children at the presyllabic level achieved zero rates of success in both versions of the tests. SimilarlyVernon (1998) showed that children at a less evolved conceptual level display fewer analytical responses to a task involving the segmentation of words than do their more advanced counterparts. This study also confirmed that children perform better when they are asked to segment words if the words are presented to them in a written format rather than an oral format. These studies, which show that phonological capabilities gradually improve until alphabetic understanding of writing is achieved, do suffer from a limitation as they centre their analysis on a single phonological task and consequently do not look upon phonological awareness as a multidimensional ability (Schatschneider, Francis, Foorman, Fletcher, & Mehta, 1999; Stahl & Murray,1994; Stanovich, 1992). This point of view may, however, be put into perspective if we bear in mind the fact that some elementary forms of phonological awareness, such as sensitivity to syllables and rhymes and the capability to detect the initial phonemes in words, can develop more or less spontaneously over the course of the preschool years (Libermanet al., 1974; Treiman, 1992). More sophisticated phonological skills, such as segmentation or phonemic deletion, may be stimulated at these ages by training programs that are oriented toward fostering this type of capacity (Goswami & Bryant, 1990; Lundberg, 1991; Troia, 1999). It is also important to mention that some-albeit not many-studies have shown that phonological training leads to progress in children's invented writing (Manrique, 1997; Tangel & Blachman, 1992). Manrique conducted a phonological intervention program in a preschool class that included rhyming games, games that involved the initial and final sounds of words, and phonemic analysis games. At the end of the year, 65% of the children not only had evolved significantly in terms of their phonological skills, but also were capable of writing words and texts in accordance with alphabetic criteria. If the results obtained by Alvarado (1998) and Vernon (1998) are combined with the latest set of data, it is possible to see that children's phonological abilities may have some effect on the way in which they look upon writing and that their phonological skills may develop as they build up more knowledge about written Phonological skills and writing code. It seems to us thatthis hypotheticalreciprocalability-enhancingactionhas not yet receivedmuch attention fromresearchersand thatit is thereforeuseful to analyse it, startingwith the initiallevels of knowledgeaboutwriting and workingwithinthe frameworkof a multidimensional concept of phonologicalabilities. Researchquestions We have formulatedthe followingresearchquestions: 1. Willthe conceptualhypothesesaboutwrittenlanguage of presyllabicchildrenevolve when they are subjectedto a writingtrainingprogramintendedto alterthe way in which they representwriting? 2. Willthe conceptualhypothesesaboutwrittenlanguage of presyllabicchildrenevolve when they are subjectedto a phonologicaltrainingprogramdesignedto work on syllabicunits? 3. Arethere any differencesbetween the two types of trainingin termsof the effectsthey have on the conceptualhypothesesaboutwrittenlanguage? 4. Willthe phonologicalskillsof presyllabicchildren undergoevolutionwhen they are subjectedto a writingtrainingprogramintendedto alterthe way in which they representwriting? 5. Willthe phonologicalskillsof presyllabicchildren undergoevolutionwhen they are subjectedto a phonologicaltrainingprogramdesignedto work on syllabicunits? 6. Arethereany differencesbetween the two types of trainingin termsof the effectsthey have on phonological skills? Method Participants Participantswere 90 middle-class Portuguese children with an average age of 5 years and 8 months, a standard deviation of 3 months, a lower limit of 5 years and 1 month, and an upper limit of 6 years and 4 months. They attended various kindergartensand had not received any formal training in reading and writing. In those kindergarten classes there were no regular classroom activities or instructionrelating to phonological awareness or invented spelling. None of the children knew how to read-a fact that was verified by means of individual reading tests. All children's names are pseudonyms. Only children who displayed presyllabic invented spellings in the pretest were selected for the study. The children were then randomly divided up into three groups-two experimental groups and one control group. 471 Figure 1 Example of Samuel's presyllabic writing C)TU (URSO) (RATO) oB Nineteen children ended up not taking part in the entire experiment and were therefore eliminated from the results. The final groups were thus made up as follows: experimental group 1 (N= 21); experimental group 2 (N= 23); control group (N= 27). Design This was an experimental study in which children were assessed on a pre- and posttest basis, both in terms of the way in which they saw written language and of their phonological skills. In between the pre- and the posttests, experimental group 1 was subjected to a writing training program intended to lead to a restructuring of the way in which they saw and employed written language, while experimental group 2 was subjected to a phonological training program designed to work on syllabic units. The control group took part in exercises involving the categorisation of geometric figures in accordance with criteriasuch as shape, size, and colour. The experimental- and control-group programs bea gan week after the pretest and lasted for 2 weeks. In order to ensure that any progress made was of a lasting nature, the posttest was carried out 1 month after the programs were concluded. Because a knowledge of letters can serve as an intermediary and an instrument that makes it easier to become aware of phonemic entities (Stahl & Murray,1998; Treiman & Cassar, 1997), we checked how many letters the children were familiarwith. We also determined the level of their intelligence. Procedure Adult: Tryto write urso(bear). The evaluation of children's conceptual levels In order to assess the children's conceptual levels, we asked them to spell their name and then to spell a set of words to the best of their ability. After spelling each 472 word they were asked to read what they had written. The verbal responses that frequently accompanied the act of spelling were recorded. We dictated 21 pairs of words: words of identical size from a linguistic point of view but refer to items of different size (e.g., urso/rato [bear/mouse]); words of a different size from a linguistic perspective but refer to items of a similar size (e.g., hipop6tamo/boi [hippopotamus/ox]); and singular and plural pairs of words (see Appendix A). These choices were designed to enable us to understand whether the reasoning behind the children's writing was based on the items to which the words refer, or on the linguistic characteristicsof the words themselves. If a child used nonlinguistic or presyllabic criteriato govern his or her writing, it might be expected that inasmuch as bears are big animals and mice are small ones, he or she wrote the word urso (bear) with more or bigger letters than he or she used for the word rato (mouse) and justified his or her choice by referringto the size of the reference items. If the form of reasoning that governs the writing was linguistic, we expected that both words would be written using approximately the same number of letters and that this would be justified (for example) with reference to the size of the words concerned and especially to the number of syllables they contain. Our classification of the children's responses was inspired by the classification grids drawn up by AlvesMartins(1994), Besse (1995), and Ferreiro(1988). We employed the following categories: presyllabic writing, syllabic writing without phonetization, and syllabic writing with phonetization. Presyllabic writing. This category was used to classify spellings in which the children made no attempt to connect oral and written language. In spelling the different words children took account of factors such as a minimum number of letters for each word and a different combination of letters with which to discriminate between different words. The way in which they spelled some words took account of the properties of the reference items-for example, by using more letters for words that referred to large items. Generally speaking, the children did not verbalise at all when spelling, and they read their words globally, as a single whole. Figure 1 shows the writing of Samuel, a presyllabic child. The following are some extracts from Samuel's interview. Samuel: (Writes five letters.) Adult: Now readwhat you wrote to me and show me with your finger. Samuel: Urso(reads globally). Adult: Whydid you writethe word ursolike that? READINGRESEARCHQUARTERLY October/November/December 2002 37/4 Samuel: Because ursois a big animal. Adult: Now write rato(mouse). Samuel: (Writes2 letters.) Adult: Now readwhat you wrote to me and show me with your finger. Samuel: Rato(readsglobally). Adult: Why did you writethe word ratolike that? Samuel: Because ratois a littleanimalit has to be written with a few letters. Syllabic writing withoutphonetization. This categowas used for spellings in which the correspondence ry between oral and written language was based on syllabic units. When they produced this type of spelling, the children often syllabically segmented what was said to them orally before they spelled anything and then represented each syllable using a random letter. They read the words syllabically. Now let us look at some extracts from an interview with Isabel who wrote syllabically. Figure 2 portrays some of her writing. Adult: Tryto write urso. Isabel: Ur(writesI ) so (writesA). Adult: Now read what you wrote to me and show me with your finger. Isabel: Ur(points to the I) so (points to the A). Adult: Why did you writethe word ursolike that? Isabel: (Readsthe word out again,pointingto each of the letters)ur/so,there are two. Adult: Now write rato. Isabel: Ra (writesS) to (writesF). Adult: Now readwhat you wrote to me and show me with your finger. Isabel: Ra (pointsto the S) to (points to the F). Adult: Why did you writethe word ratolike that? Isabel: (Readsthe word out again,pointingto each of the letters)ra/to,there are two. Syllabic writing with phonetization. This category was reserved for syllabic writing where the children phonically analysed part of the spoken words by selecting a letter from their repertoire that adequately represented the sounds they had identified. They continued to read words syllabically, and in their verbal output the children tried to identify not only the quantity of letters that they ought to use, but also-in some cases-the most suitable ones. Finally we have some extracts from an interview with Pedro in which some of the words were written in this way. Figure 3 illustrates his writing. Adult: Now write urso. Pedro: Ur...ur...u(writesU) so (writesI). Adult: Now read me what you have justwrittenand show me with your finger. Pedro: Ur/so(points to the U and then to the I). Phonological skills and writing Figure 2 Example of Isabel's syllabic writing without phonetization (URSO) (RATO) Figure 3 Example of Pedro's syllabic writing with phonetization I (URSO) (RATO) Adult: Pedro: Now write rato. Ra...ra... ra, it's A (writes A), ra/tu, tu it's U (writes U). Adult: Now read me what you have justwrittenagain and show me with your finger. Pedro: Ra/to(pointsfirstto the firstletterand then to the last one). When we analysed the protocols, we calculated the number of words in each of the categories. On the basis of this analysis we determined the percentile distribution of the various types of spelling between the protocols and classified the children by evolutionary level-for a child to qualify for a given level, around 90% of his or her spellings had to match the applicable criteria. The evaluation ofchildren 'sphonological awareness In order to evaluate the children's phonological awareness, we gave them a battery of four phonological subtests with differing levels of difficulty (so as to address the heterogeneous nature of this particularcapability). When we created this battery of tests, we sought to take account of the size of the units (syllables and phonemes) and the phonological properties of the initial phonemes 473 in each word. The battery included two classification and two deletion tests (see Appendix B). Each of the classification tests, one of which was based on the initial syllable and the other on the initial phoneme, was composed of 14 items preceded by 2 examples. In each item the children were presented with four drawings representing four oral words (there were no written words). In the case of the initial-syllableclassification test two of the words in each item began with the same syllable, while the other two started with different ones (e.g., girafa [giraffe]/panela [cooking pot]/cenoura [carrot]/palhago[clown]), and the children had to identify the words that began with the same syllable. In the initial-phoneme classification test two of the words in each item began with the same phoneme, whereas the others started with different ones, and the children again had to identify the words that began with the same one (e.g., fivela [buckle]/telhado [roofll/janela[window]/ fogueira [bonfire]). The children were given the following instructions when they took the initial-syllable classification test: "We'regoing to play a kind of cardgame in which we have to find two words thatbegin with the same syllable, fromamonga totalof fourwords. First,can you tell me The childwas what each of these drawingsrepresents?" shown the firstexample.Whenhe or she found it difficult to identifyany of the words thatcorrespondedto the drawings,the researchernamedthe word and asked the child to repeatit as manytimes as was necessaryuntilhe or she was automaticallyable to identifyall fourwords. The adultthen went on by saying,"Twoof these words in the drawingsbegin with the same syllable.Let'splay, but this time it doesn'tcount-it's justfor you to see how the game works.Whichare the two words thatbegin with the When the childgot it wrong the researcher same syllable?" asked him or her to pronounceall the words slowly, emphasisingthe initialsyllableand leadinghim or her to realise the similaritybetween the initialsyllablesof two of the words.This procedurewas then repeatedfor the second example.When both exampleswere over the researcheradded, "NowthatI've told you how the game is played, let's start.But be careful-this time you have to find out which words begin with the same syllableon your own." The instructions for the initial-phoneme classification test were similar. They only differed in one respectinstead of saying that the children had to find the words that began with the same syllable, they were told that they had to find those that began with the same sound. When a child got the first example item wrong, the researcher asked him or her to pronounce all the words slowly, emphasising the initial phoneme and leading him or her to see the similarity between the initial phonemes 474 in two of the words. This procedure was then repeated for the second example item. In the tests involving the deletion of an initial syllable and an initial phoneme, the children were asked to pronounce in isolation each of the syllables or phonemes of words that were presented in a figurative format (i.e., each oral word was again represented by a drawing), and to say what remained of the word without the initial segment, the deletion of which might result in something that was not a word in terms of the Portuguese language (e.g., no/ta; ta, in the initial-syllable deletion test; or r/io; io, in the initial-phoneme deletion test). The initial-syllable deletion test was made up of 14 items, half of which were composed of two syllables each and the other half of three syllables each, preceded by two examples; the initial-phoneme deletion test was made up of 24 items (also preceded by two examples), half of which were composed of monosyllabic words and the remainder were disyllabic. The children were given the following instructions for the initial-syllable deletion test: "Doyou see these cards?Do you know what this drawing shows?"When a child found it difficultto identifyone of the words thatcorrespondedto the drawings,the researchernamedthe word and got the child to repeatit as manytimes as was necessaryuntilhe or she was automatically able to identifythe word. "Thisgame is like this: We'regoing to take away the firstsyllableof the word and say what'sleft. Let'splay, but this time it doesn'tcountit'sjustfor you to see how the game works.Tell me what the firstsyllableof the word nota (banknote) is."If the child was unableto answer,he or she was asked to pronounce the word slowly along with the researcherin such a way as to isolatethe firstsyllable."Nowlet's take the no away fromthe word. What'sleft of the word nota if we take away no?"In the event thatthe childwas unableto complywith the instructionthe researchersaid, "Well,if we take away the firstsyllableof the word nota we're left with ta. Listencarefully:no/ta; ta."The same procedure was then followed for the word in the second example,after which the researcheradded, "Nowthatyou've understood how the game works,let's play for real.You're going to tell me the firstsyllablein each word and then say what is left of the word once we have takenaway this littlebit." The instructions for the initial-phoneme deletion test were similar, except that instead of saying that the idea was to take away the first syllable, we said that we were going to take away the first sound. In the event that a child was unable to complete one of the example items correctly, he or she was asked to pronounce the word slowly along with the researcher in such a way as to isolate the initial phoneme and then say what was left of the word without that sound. READINGRESEARCHQUARTERLY October/November/December 2002 37/4 Whenever the researcher was in any doubt as to whether the child had understood the term "syllable,"he or she always used the term "littlebit" instead. In all the tests one point was awarded for each correct answer. The evaluation of children's knowledge of letters In order to determine how many and which letters the children were familiarwith, they were give a set of cards bearing the letters of the alphabet in capitals (K, W, and Ywere excluded, as they are not formally part of the Portuguese alphabet), which they were asked to name. They were then asked to write down the letters they had recognised. The total number of possible points in this test varied between 0 and 23. The evaluation of children's intelligence The level of the children's intelligence was evaluated using the coloured version of Raven's Progressive Matricestest (1947-as revised in 1956), because it is not very dependent on verbal aspects. The writing training program The writing training program was organised around situations that led the child to think about the rules of writing from two points of view: His or hers, and that of a hypothetical boy or girl of the same age whose spellings were syllabic. In each session the children were asked to spell a set of words to the best of their ability. After spelling each word they were asked to read out what they had written. They were then confronted with the same word, as spelled in accordance with syllabic criteria-one letter/one syllable correspondence-by a hypothetical child from another school. The children were asked to say which of the two versions was better. Before deciding, they had to try to slowly read both what they had written and the writing of the hypothetical child and to try to justify why they had written the word as they had and why the other child had written it in another way. From the second session onward, in addition to these instructions, before they wrote down the words the children were asked to say in advance how many letters they were going to use. In this way, we created the conditions that would arouse a cognitive conflict in the children between two approaches to writing, inasmuch as the situation itself led them to confront the two ways of writing and to try to expound the hypotheses underlying each of the two approaches. The objective of these spelling situations was to lead children with presyllabic writing to evolve toward syllabic writing. Phonological skills and writing Thephonological training program The phonological training program involved oral and aural exercises of identification, segmentation, and manipulation of syllabic units. It did not include any written words. In designing it we sought to take account of a number of principles that we considered to be important to research in this domain. They included factors such as the use of drawings to represent sounds (Ball & Blachman, 1991); the use of an object to denote each sound (O'Connor,Jenkins, & Slocum, 1993); the modelling of sounds, with their reproduction by the child and with the researcher calling attention to their articulatory characteristics(Lie, 1991); explicit instructions for the tasks employed; and the graduation of linguistic complexity, particularlyin terms of the size of the words and the articulatoryproperties of the phonemes. The following tasks were performed over the course of the eight sessions that made up the training program: 1. Syllabic segmentation games in which the child first had to pronounce each syllabic unit and denote it by beating time with a pencil, and then identify the number of syllables that made up each word. The children were given the oral words in a series of four. Each series was composed of words that varied between one and four syllables (e.g., sol [sun]/fada [fairy]/camelo [camel]/borboleta [butterfly]).Whenever a child was unable to carry out the task, the correct response was modelled and he or she was again asked to reproduce it. 2. Initial-syllableclassification games in which the child had to identify two words that began with the same syllable from among four that were presented to him or her in figurative form (each oral word was represented by a drawing). This task was organised into 12 series. Each one was made up of six items of four words each. In one of the series the initial syllables were vowels (e.g., anel [ring]/irmd [sister]/avd [grandfather]/6culos[spectacles]; in the others the target syllables in all six items began with the same consonant (e.g., bota [boot]/carro [car]/rato [mouse]/raquete [racket];or circo [circus]/rosa [rose]/mapa [map]/rede [net]). We employed the following procedure in playing this type of game. The children were asked to indicate the initial syllable of each of the four words in such a way as to identify words that possessed the same initial syllable. When the children found it difficult to carry out this operation, the researcher got them to repeat the word slowly and to stop after they had said the first syllable. Once they were able to isolate the syllable they pointed to each of the words, indicating only the first syllable and identifying the words that shared the common initial segment. 475 Table 1 Number of children whose writing was presyllabic and syllabic at the posttest Experimental Group 1 (n = 21) Experimental Group 2 (n = 23) Control Group (n = 27) Children with presyllabic writing Childrenwith syllabic writing 0 0 26 21 23 1 3. Initial-syllabledeletion games in which the children were asked to pronounce in isolation each of the syllables of 20 words, and to say what remained of the word without the initial segment. It should be noted that in order to facilitate the children's task, the deletion of the first syllable of the words that we used always gave rise to a new word. For 10 of the words we used a game with a figurative support (each oral word was represented by a drawing). For the remaining 10, the words were presented orally (without pictures). We employed the following procedure in playing this type of game. The children were told that there was a word hidden within another word and that they could find it if they took away the first "littlebit."They had to try to find out which of three words that were presented to them corresponded to the hidden one (e.g., for the word fivela [buckle] they were given the words vela [candle]/vaca [cow]/dado [die];while for espinha [bone] they were given linha [line]/pinha [pine]/pote [pot]). Even when the child was immediately successful, the researcher modelled the operation that the child had just carried out in order to show him or her that he or she was right. When the child was unable to give the correct answer, the researcher led him or her to identify the first syllable, modelled the deletion operation, and made the child repeat it. The control group program We organised a set of exercises with the control group using material of the logical blocks type. The children were asked to classify geometric shapes in accordance with criteriasuch as identical shape, size, or colour. The three programs involved eight approximately 15-minute sessions and were individually conducted by us with the children over the course of 2 weeks. Results We will begin by presenting the data concerning the knowledge of letters and the level of intelligence of the children in the three groups. We carried out ANOVAwith the group as the independent variable and either the number of letters known or the level of intelligence as the 476 dependent variables. We obtained a figure of 1(2,68) = .02, p = .978 for the number of letters known and 1(2,68) = 1.80, p = 1.78 for the level of intelligence. There were no statisticallysignificant differences between the three groups. In order to assess the effect of the two types of training on the children's concepts of written language (research questions 1, 2, and 3), we looked at the number of children in the three groups whose writing was more highly evolved in conceptual terms at the posttest moment. Table 1 suggests that both forms of training contributed to an expansion of the children's knowledge about written code, inasmuch as the posttest writing of the children in both these groups was systematically governed by linguistic principles. When we look at this table we do not need any statisticalprocedure to tell us that the two types of training had a similar impact in terms of their role as a factor in the progress of the children's concepts about written language. Analysis of the children's individual protocols at the posttest shows that all the children in the two experimental groups began to govern their attempts at writing in accordance with syllabic criteria.Only two children in experimental group 1 and four in experimental group 2 began to phonetize some of the syllables in the words. These phonetization procedures principally began to be applied to the first syllable in a word and were used more often in relation to vowels than to consonants. At the same time they almost always occurred with words in which the initial syllable was accentuated. These children were the ones who were familiarwith the greatest number of letters. The children in the control group continued to obey presyllabic criteriain their spelling. Just one child in this group spelled a few words using syllabic criteria (more precisely, 22% of all the words she wrote). Figure 4, which contains extracts from Vera's (a member of experimental group 1) protocol, exemplifies the presyllabic forms of writing that she produced in the pretest. Adult: Now write urso(bear). Vera: (Writesin silence.) Adult: Now readme what you have justwrittenand show it to me with your finger. Vera: (Readsgloballywhile simultaneouslypassingher fingerover the writing.) Adult: Now write rato(mouse). Vera: (Againwritesin silence.) Adult: Readit to me againand show me with your finger at the same time as you read. Vera: (Readsas before.) Adult: Why did you use differentlettersfor ursoand rato? READINGRESEARCHQUARTERLY October/November/December 2002 37/4 Figure 4 Exampleof Vera'spresyllabicwritingin the pretest Table 2 Meanscores and standarddeviationsfor the initial-syllableclassificationtest Pretest hc T Ei ijr D (URSO) Experimental Group 1 Experimental Group 2 Control Group 0 Vera: (RATO) (URSO) (RATO) Now let us look at extracts from Vera's posttest, which are illustrativeof the characteristicsof the children's spellings at that point. Vera spelled all the words in accordance with the syllabic hypothesis. Figure 5 shows some of the words she spelled. Now write urso. Adult: Ur(writesO) so, so (writesA). Vera: Now read me what you wrote and show me Adult: with your finger. Ur(points to the O) so (points to the A); it Vera: has to be two, don'tyou see...ur/so(readsthe word again). Adult: Now write rato. Ra (writesO) to (writesS). Vera: Now readwhat you wrote to me and show Adult: me with your finger. Ra (points to the O) to (points to the S). Vera: Researcher:Why did you writethe word ratolike that? Vera: (Readsthe word out again,pointingto each of the letters)ra/to,there are two of them. In order to assess the effect of the two types of training on the children's phonological skills (research questions 4, 5, and 6) we used ANOVAwith repeated measures to compare the children's performance in the skills and writing SD Mean SD 11.47 9.43 10.55 2.24 3.57 2.72 13.85 13.73 12.11 .35 .31 2.27 Table 3 Meanscores and standarddeviationsfor the initial-phonemeclassificationtest Pretest Experimental Group 1 Experimental Group 2 Control Group Posttest Mean SD Mean SD 5.95 6.73 5.33 2.61 2.41 1.50 8.28 9.86 5.60 1.90 1.96 2.10 This test was composed of 14 items. Becausea bear is a bear-it's not the same thing as a mouse, is it? Phonological Mean This test was composed of 14 items. Figure 5 Exampleof Vera'ssyllabicwritingin the posttest 07 Posttest various phonological tasks that the children were set on the pre- and posttest. In order to deepen our understanding of the differences between the groups, we subsequently carried out a post hoc analysis using a version of the Tukey test that applies to situations involving different-sized groups. In those cases in which we could not assume that variances were homogeneous, we performed the same analysis using the Games-Howell procedure. We will now present the children's results in each of the tests. In the initial-syllable classification test, the descriptive statistics included in Table 2 suggest that the results obtained by the two experimental groups on the posttest are close to one another and slightly better than those of the control group. The results of the ANOVAreveal that there was a only statisticallysignificant evolution in the children's results in this classification test between the two evaluation moments, F(1,68) = 68.84, p = .001. There were no statisticallysignificant differences between the groups. When we come to the initial-phoneme classification an test, analysis of Table 3 enables us to see that the avthat the two experimental groups obtained at the erages are close to each other and are higher than those posttest of the control group. There had not been much difference between the average results achieved by the three groups in the pretest. The results of the ANOVA show that the moment F(1,68) = 46.72, p = .001; group F(2,68) = 16.38, p = .001; and group x moment interaction F(2,68) = 9.72, p = .001 variables produced statistically significant effects. The post hoc procedure revealed that both types of training led to statisticallysignificant effects on the chil- 477 Table 4 Meanscores and standarddeviationsfor the deletiontest initial-syllable Pretest Experimental Group 1 Experimental Group 2 Control Group Posttest Mean SD Mean SD 4.76 3.69 3.59 4.42 3.59 3.72 12.90 11.86 6.70 1.99 2.18 5.23 This test was composed of 14 items. Table 5 Meanscores and standarddeviationsfor the initial-phonemedeletiontest Pretest Mean Experimental Group 1 Experimental Group 2 Control Group 3.14 5.52 1.44 Posttest SD 6.49 6.49 3.64 Mean SD 3.95 13.78 1.74 7.12 7.12 2.95 This test was composed of 24 items. dren'sperformancein this phonologicaltask,inasmuch as therewere statisticallysignificantdifferencesbetween the resultsachievedby the two experimentalgroupson the one hand and those of the controlgroup on the other, whereastherewere no differencesbetween the results of the experimentalgroupsthemselves. The averagesachievedin the initial-syllable deletion test (see Table4) clearlyindicatethatthe two experimental groupsobtainedmuch betterresultson the posttest than did the controlgroup.The same table also shows thatthe averagesthe threegroupsscored on the pretest did not differgreatlyfromone another,albeitExperimentalGroup1 did slightlybetterthanthe others. The resultsof the ANOVAcorroboratethe idea that the children'sresultsin this test evolved significantly, froma statisticalpoint of view, between the pretestand posttestmomentsF(1,68) = 202.44,p = .001.They also show thatthereare statisticallysignificantdifferencesdependingon the groupF(2,68) = 8.48,p = .001 and the momentx group interactionF(2,68) = 14.05,p = .001. Post hoc analysisshowed thatboth types of trainingproduced an impact,inasmuchas we recordedstatistically significantdifferencesbetween the experimentalgroups on the one hand and the controlgroupon the other,but not between the two experimentalgroups. As regardsthe initial-phonemedeletiontest, we can see fromTable5 thatexperimentalgroup 2's resultsin the posttestwere farsuperiorto those of the othertwo groupsand therewas a considerableevolutionbetween the two evaluationmoments.The averagesattainedby 478 the other two groups were practically identical on the pretest and the posttest. The results of the ANOVAshow that the children's performance in the initial-phoneme deletion test evolved significantly, from a statisticalpoint of view, between the pretest and posttest F(1,68) = 29.95, p = .001. The group F(2,68) = 13.52, p = .001 and moment x group interaction F(2,68) = 20.43, p = .001 variables also produced substantial effects. Post hoc analysis showed that this difference in evolution is due to the impact of the phonological training, inasmuch as the averages that the members of experimental group 2 obtained in this test differ significantly, from a statisticalpoint of view, from those obtained by experimental group 1 and the control group. On the other hand the same analysis reveals no differences between the averages obtained by experimental group 1 and those of the control group. Discussion One of the aspects we sought to assess in this study was to what extent presyllabic children who are subjected to a writing training program evolve, not only in their conceptualisations of writing, but also in their phonological skills. Our data confirm that a writing training program designed to alter the way in which presyllabic children see and use writing such that they begin to employ syllabic criterianot only leads presyllabic children to progress in their invented spellings, but also improves their phonological skills. The children's performance in phonological tasks on the posttest did indeed improve significantly from a statistical point of view, both in terms of their manipulation of syllabic units (as we can see from the results they obtained in the initial-syllable deletion tests) and as regards their awareness of the presence of common phonemic elements in different words (shown by the results of the initial-phoneme classification test). The progress we recorded in the latter test is particularly significant in light of the fact that children's ability to identify common phonemes in different words is critical to an understanding of the alphabetic nature of writing (Stahl & Murray,1998). The results obtained in this test also suggest that establishing systematic relationships between segments of speech and written units may contribute to the conceptual emergence of phonemic entities along lines that are similar to the way put forward by Byrne and Fielding-Barnsley(1989), who looked upon evolution in the performance of this task as concept formation. The evolution we observed in the phonological tasks, in the children who underwent the writing training program, fits into the overall line of thought that Treiman READINGRESEARCHQUARTERLY October/November/December 2002 37/4 (1998) and Adams (1998) suggested about the role invented writing plays in enabling children to grasp the oral structure of words. We also sought to investigate whether presyllabic children who underwent a phonological training program would evolve, not only in their phonological abilities, but also in their conceptualisations of writing. The data we obtained indicate that the phonological training program led not only to statisticallysignificant progress in the children's phonological skills (as may be seen from the results they obtained in the initial-syllable deletion, initial-phoneme classification, and initial-phoneme deletion tests), but also to changes in the characteristicsof their invented spellings. Our study confirms the data that Manrique (1997) and Tangel and Blachman (1992) produced that phonological training leads to progress in the nature of children's invented spelling. Finally, we sought to investigate whether the two types of training would have identical effects on both children's conceptualisations of writing and their phonological skills. When it came to the way in which the children represented writing, the data show that the two types had equivalent effects (at the posttest all the children in both experimental groups had begun to govern their writing in accordance with the syllabic hypothesis). One element that warrants particularnote is the case of the children who, besides beginning to write in accordance with the syllabic hypothesis, also started to employ conventional letters to represent some of the sounds in words. As we said earlier, these children were those members of the experimental groups who were familiar with the largest number of letters. Besides this, on the pretest these children had achieved success rates that were above the average scored by the other children in the tests involving classification based on common phonemes and the deletion of the initial phoneme. It was probably the conjugation of their knowledge of letters and a higher level of phonemic awareness that led to the appearance of phonetization processes in these children. The fact that phonetization procedures only occurred in these children confirms Byrne's (1998) view about the need for coordination between knowledge of letters and phonological abilities if children are to make progress in learning the written code. However, in our case we found that this was already true of children at a less evolved level. It is also important to add that in the relatively isolated examples of phonetization processes we came across, vowels were used more often than consonants. This can probably be explained by the fact that in Portuguese the sound value of some vowels in syllables corresponds to the name of the letter concerned (e.g., the i in livro [book] or the u in urso [bear]).This match be- Phonological skills and writing tween the name of a letter and its sound value occurs much more often in vowels than it does in consonants. This circumstance makes it easier to remember the letter in question, given that various authors have shown that one of the factors that influence children's ability to establish a relationship between letters and phonemes is that the phoneme forms part of the name of the letter in question (Mann, 1993; Treiman & Cassar, 1997; Treiman et al., 1998). As far as phonological skills are concerned, the data show that there were both similarities and differences between the groups. Let us begin by emphasising the fact that when set against the results of the control group, neither of the two training programs seems to have had a statisticallysignificant impact on the initial-syllable classification test. The relatively good results that the children in the control group achieved in this test indicate that these types of skills develop spontaneously over the course of the preschool years and do not involve any development in the way in which children look at and use writing, as has been shown in many other studies (Libermanet al., 1974; Sim-Sim, 1997; Treiman, 1992). At the same time the two programs had more or less identical effects on the initial-syllable deletion test and the initial-phoneme classification test, but not on the initial-phoneme deletion test. In the latter the results obtained by the children who underwent the phonological training program (experimental group 2) were clearly superior to those achieved by the group that took the writing training program (experimental group 1). If we bear in mind the fact that the training only addressed syllabic units and did not include any form of activity aimed at appropriating segmental units, these results raise a number of questions. There is not a lot of literatureto help interpret them, given that when phonological intervention programs employ games with syllables or other suprasegmental units, most of them do so as a prior step to the use of games that address phonemic units (Byrne & Fielding-Barnsley, 1989; Lundberg, Frost, & Peterson, 1988). Reviews of recent works on phonological training programs (Bus, Marinus,& Ijzendoorn, 1999; Troia, 1999), in which the characteristicsand the impact of intervention programs conducted as part of the last 20 years' research in this domain are analysed, do not contain a single reference to a program that has been exclusively centred on syllabic units. The results that the children in experimental group 2 obtained in the initial-phoneme deletion test may perhaps have been precipitated by the way in which the deletion operation was modelled at syllable level. In light of the fact that "the concept of phoneme identity is closely related to the ability to segment a part of a spoken word" (Stahl & Murray,1998, p. 83), this aspect of the 479 program, which was coordinated with the children's progress in forming the identity of phonemes, may have contributed to the statisticallysignificant evolution in their scores in this test. In addition, the progress we observed in the performance of this task may also have been facilitated by the characteristicsof the test itself. The test included many items that were composed of monosyllabic words in which the initial phoneme corresponds to the onset of the syllable. Treiman (1992) said that in these circumstances preschool-age children can achieve a certain amount of success in this test on the basis of the intrasyllabic components. In summary, the results of this study confirm the point of view put forward by various authors (Byrne, 1997, 1998; Pontecorvo & Orsolini, 1996; Stahl & Murray,1998; Tolchinsky & Teberosky, 1998) who said that there is a complex interactionbetween the processes involved in becoming aware of the oral units in speech and understanding the way in which written code works. This interaction, which has primarilybeen noted when children begin to see conventional correspondences between the sounds of words and the letters they need to use (Adams, 1998; Treiman, 1998), seems to be true even at earlier levels. This conclusion is one of the most innovative aspects of the results we obtained, inasmuch as the interactive dynamic between oral learning processes and the processes involved in understanding writing have previously been taken into consideration only when children have begun to establish some graphic-phonetic correspondences in their preconventional writing (Treiman, 1998). Our study confirms that the skills that enable children to think about oral language and the way in which children represent the written code begin to influence one another at quite early stages. This means that within the context of a language like Portuguese, in which there are a lot of polysyllabic words and open syllables of the consonant-vowel type, phonological training programs that work on syllabic units lead children to establish relationships between speech and writing that are themselves based on syllabic units. At the same time, inducing children to think about syllabic components as units that codify writing favours the development of phonological skills. It should be noted that this evolution also occurs at the level of the awareness of phonemes-a skill that Byrne (1998) and Stahl and Murray(1998) consider critical to grasping the alphabetic characteristicsof writing. Our study also reveals that there may be two ways in which it is possible to transition from forms of presyllabic writing to syllabic writing formats: one derived from the application to writing of the various ways to manipulate syllabic units, and one of reflection on the relation- 480 ship between the various parts of the written word and the word as a whole. From the pedagogical point of view our results strengthen the stance that various authors (Adams, 1998; Treiman, 1998) have taken about the importance of the early stimulation of invented spelling activities to promote both phonological awareness and the gradual learning of the alphabetic principle. This implication becomes even more importantif we remember that although they were not yet establishing graphic-phonetic correspondences, the children who underwent the syllabic writing program experienced statisticallysignificant evolution in terms of their awareness of phonemic entities-a capability that is considered fundamental to an understanding of the alphabetic principle (Byrne, 1998; Stahl & Murray,1998) At the same time the results of our study also seem to indicate that programs designed to train syllabic unit identification, segmentation, and manipulation capabilities may have an impact at the phoneme awareness level, quite apart from the impact they have in terms of children's representations of written language. Given that both these types of program are easily applied to children in this age group, it seems to us that they should be more systematically used to help children understand the alphabetic principle. To conclude, we would also like to point out some limitations of this study and a few directions for future research. One of the limitations is that we did not control the general language ability of the children in our sample group-a factor that can be important, given that some studies have shown the existence of a relationship between this variable and phonological awareness (e.g., Webster & Plante, 1992). Another limitation is that although there is some similaritybetween the pre- and posttest and the training programs, the children may simply have imitated learned behaviours rather than reorganized their conceptual understandings. However, it seems to us that the fact that the posttest was carried out a month after the training program offers at least some guarantee that what took place really was a number of conceptual reorganisations. At the same time, the fact that the writing training program had an impact on the phonological tests and that the phonological training program had an effect on the children's representations of writing would appear to confirm that the outcome did not constitute an imitation of learned behaviours. As far as developments for future research are concerned, it would be interesting to manipulate the number of letters with which the experimental groups are familiar in such a way as to be able to specify the mediating function this knowledge performs in terms of the training's effectiveness. Another pertinent aspect that re- READINGRESEARCHQUARTERLY October/November/December 2002 37/4 quiresmore in-depthinvestigationin the futureis the need to clarifythe capabilitiesthat,within the framework of this type of intervention,may well have enabled some childrento evolve to the use of phonetizationprocesses in theirwriting. REFERENCES and learningaboutprint. ADAMS,M.(1998). Beginningto read: Thinking Cambridge,MA:MITPress. M. (1998). Conscienciafonol6gicay escrituraen nihos ALVARADO, preescolares:la posibilidadde omitirel primersegmento. Lecturay Vida,3, 42-50. M.(1994). Conceptualizag6esinfantissobre a linALVES-MARTINS, guagem escritae aprendizagemda leitura.Discursos,8, 53-70. ALVES-MARTINS, M.,&MENDES,A. (1987). Evolugdodas conceptualiza96es infantissobre a escrita.AnalisePsicol6gica,5, 499-508. ANDRADE,E., &VIANA,M.C.(1993). Sinerese, dierese e estruturasildbica. InActas do IXEncontroNacionalda AssociagdoPortuguesade Linguistica (pp. 31-42). Coimbra,Portugal:APL. B.A.(1991). Does phonemeawareness trainingin BALL,E., &BLACHMAN, make a differencein earlywordrecognitionand developmental kindergarten spelling?ReadingResearch Quarterly,24, 49-66. BESSE,J.-M.(1993). De I'ecriture productricei la psychogenese de la langueecrite. InG. Chauveau,M.R6mond,&E. Rogovas-Chauveau(Eds.), L'enfantapprenti-lecteur. L'entr6edans le systbme ecrit(pp. 73-82). Collection CRESASn- 10. InstitutNationalde RecherichePedagogique:L'Hartmattan. BESSE,J.-M.(1995). Lecrit,1'6coleetl'illettrisme.Paris:Magnard. BESSE,J.-M.(1996). An approachto writingin kindergarten.InC. Pontecorvo,M.Orsolini,B. Burge,&L. Resnick(Eds.), Children'searlytext construction(pp. 127-144). Hillsdale,NJ:Erlbaum. V. (1999). PhonologicalawareBUS, A., MARINUS, H., &IJZENDOORN, ness and earlyreading:A meta-analysisof experimentaltrainingstudies. Journalof EducationalPsychology,91, 403-414. BYRNE,B. (1992). Studies in the acquisitonprocedurefor reading: Rationale,hypothesesand data. InP. Gough, L. Ehri,&R. Treiman(Eds.), Readingacquisition(pp.1-34). Hillsdale,NJ:Erlbaum. of the alphabeticprinciple:Children'siniBYRNE,B. (1997). The learnability tialhypotheses abouthow printrepresentsspoken language.Applied Psycholinguistics,17, 401-426. BYRNE,B. (1998). Thefoundationof literacy.Hove, England:Psychology Press, Ltd. R. (1989). Phonemicawareness and BYRNE,B., &FIELDING-BARNSLEY, letterknowledgein the child'sacquisitionof the alphabeticprinciple.Journalof EducationalPsychology,81, 313-321. E. (1994). Les cheminsde la CHAUVEAU, G., &ROGOVAS-CHAUVEAU., lecture.Paris:Magnard. avantla la lettre.InH. Sinclair(Ed.),La E. (1988). L'ecriture FERREIRO, produtiondes notationschez le jeune enfant(pp. 18-69). Paris:Presses Universitairesde France. J. (1993). Entrerdans l'6crit.Paris:Magnard. FIJALKOW, GENTRY,J.R. (1982). An analysisof developmentalspellingin GNYSAT WRK.TheReading Teacher,36,192-200. U. (1998). The roleof analogies in the developmentof word GOSWAMI, recognition.InJ.L.Metsala&L.C.Ehri(Eds.), Wordrecognitionin beginningliteracy(pp. 41-63). London:Erlbaum. GOSWAMI, U., &BRYANT,P. (1990). Phonologicalskillsand learningto read Hove, England:Erlbaum. HENDERSON,E.H.,&BEERS,J.W. (1980). Developmentaland cognitive aspects of learningto spell:A reflectionof wordknowledge.Newark,DE: International ReadingAssociation. LIBERMAN, D., FISCHER,F.W.,&CARTER,B. I.Y.,SHANKWEILER, (1974). Readingand the awareness of linguisticsegments. Journalof ChildPsychology, 18, 201-212. Experimental LIE,A. (1991). Effectsof a trainingprogramforstimulatingskillsin word 26, 234-250. analysis in first-gradechildren.ReadingResearch Quarterly, LUNDBERG, I. (1991). Phonemic awareness can be developed without reading instruction. In S. Brady & D. Shankweiler (Eds.), Phonological processes in literacy (pp. 47-53). Hillsdale, NJ: Erlbaum. Phonological skills and writing LUNDBERG, I., FROST,J., &PETERSON,O.-P. (1988). Effectsof an extensive programforstimulatingphonologicalawareness in preschoolchildren. 23, 263-283. ReadingResearch Quarterly, MANN,V. (1993). Phonemeawareness and futurereadingability.Journal of LearningDisabilities,26, 259-269. A.M.B.(1997). Alfabetizaci6nemergente:DiferenciassocioMANRIQUE, culturales.Unpublisheddoctoraldissertation,Universidadde Buenos Aires, Buenos Aires,Argentina. MORAlS,J. (1994).Lartde lire.Paris:EditionsOdileJacob. B. (1998). Gainingalphabeticinsight:Is phonememanipulation MURRAY, skillor identityknowledgecausal? Journalof EducationalPsychology,90, 461-475. O'CONNOR,R., JENKINS,J., &SLOCUM,T. (1993). Teachingphonological awareness to youngchildrenwithlearningdisabilities.ExceptionalChildren, 59, 532-546. PERFETTI, C.A.,BECK,I., BELL,L., &HUGUES,C. (1987). Phonemic knowledgeand learningto readare reciprocal:A longitudinal studyof firstgrade children,Merrill-Palmer Quarterly,33, 283-319. M.(1996). Writingand writtenlanguage PONTECORVO, C., &ORSOLINI, in children'sdevelopment.InC. Pontecorvo,M.Orsolini,B. Burge,&L. Resnick (Eds.), Children'searlytextconstruction(pp.3-23). Hillsdale,NJ:Erlbaum. C. (1988). Modesof differentiaPONTECORVO, C., &ZUCHERMAGLIO, tion inchildren'swritingconstruction.EuropeanJournalof Psychologyof Education,3, 4, 371-398. C. (1995). A passagem paraa alPONTECORVO, C., &ZUCHERMAGLIO, fabetizago: aprendizadonumconceitosocial. InY. Goodman(Ed.),Comoas criangasconstroema leiturae a escrita(pp. 67-101). PortoAlegre,Brasil:Artes Medicas. G. (1994). El uso y funci6nde las letrasen el periodopre-alQUINTERO, fab6tico,Lecturay Vida,15,28-38. READ,C. (1971). Pre-schoolchildren'sknowledgeof Englishphonology. HarvardEducationalReview,41, 1-34. READ,C. (1975). Children'scategorizationof speech sounds in English. Research Rep. No. 17. Urbana,IL:NationalCouncilof Teachersof English. READ,C. (1986). Children'screativespelling.London:Routledge&Kegan Paul. D. (1995). Inventedspellingabilityand printedwordlearningin RICHGELS, kindergarten.ReadingResearchQuarterly,30, 96-109. SCHATSCHNEIDER, C., FRANCIS,D., FOORMAN, B., FLETCHER, J., & of phonologicalawareness:An applicaMEHTA,P. (1999). The dimensionality tionof item response theory.Journalof EducationalPsychology,91, 439-449. SIM-SIM,I. (1997). Avaliagdoda linguagemoral:Umcontributopara o conhecimentodo desenvolvimentolinguisticodas criangasportuguesas.Lisboa, Portugal:FundagdoCalousteGulbenkian. B. (1994). Definingphonologicalawareness and its STAHL,S., &MURRAY, relationshipwithearlyreading.Journalof EducationalPsychology,86, 221-234. B. (1998). Issues involvedin definingphonological STAHL,S., &MURRAY, awareness and its relationto earlyreading.InJ.L. Metsala&L.C.Ehri(Eds.), Wordrecognitionin beginningliteracy(pp. 65-85). London:Erlbaum. K. (1992). Speculationson the causes and consequences of STANOVICH, individualdifferencesin earlyreadingacquisiton.InP. Gough,L. Ehri,&R. Treiman(Eds.), Readingacquisition(pp. 307-347). Hillsdale,NJ:Erlbaum. SULZBY,E. (1986). Writingand readingas signs of oraland writtenlanguage organizationin youngchildren.InW.H.Teale & E. Sulzby(Eds.), and reading(pp.50-87). Norwood,NJ:Ablex. Emergentliteracy:Writing SULZBY,E. (1989). Assessment of emergentwritingand children'slanin real time: guage whilewriting.InL. Morrow&J. Smith(Eds.), Writing Modellingproductionprocesses (pp. 83-109). New York:Longman. TANGEL, D.M., & BLACHMAN,B.A. (1992). Effect of phoneme awareness instruction on kindergarten invented spelling. Journal of Reading Behaviour, 24, 233-262. TOLCHINSKY,L.L. (1995). Desenvolvimento da alfabetiza9&o e suas implica96es pedag6gicas: evid~ncias do sistema hebraico de escrita. In Y. Goodman (Ed.), Como as crianpas constroem a leitura e a escrita (pp. 36-53). Porto Alegre, Brasil: Artes Medicas. TOLCHINSKY,L.L., & LEVIN,I. (1988). O desenvolvimento da escrita em criangas israelenses pre-escolares. In E. Ferreiro & M. Palacio (Eds.), Os processos de leitura e escrita (pp. 143-158). Porto Alegre, Brasil: Artes Medicas. 481 A. (1998). The developmentof word TOLCHINSKY, L.L.,&TEBEROSKY, segmentationand writingin two scripts.CognitiveDevelopment,13,1-24. R. (1992).The roleof intrasyllabic unitsin learningto readand TREIMAN, spell. InP. Gough,L. Ehri,& R. Treiman(Eds.), Readingacquisition(pp. 65-106). Hillsdale,NJ:Erlbaum. R. (1998). Whyspelling?The benefitsof incorporating TREIMAN, spelling intobeginningreadinginstruction.InJ.L.Metsala&L.C.Ehri(Eds.), Word recognitionin beginningliteracy(pp. 289-313). London:Erlbaum. de I'orthographe en TREIMAN, R., &CASSAR,M.(1997). L'acquisition anglais. InL. Rieben,M.Fayol,&C. Perfetti(Eds.), Des orthographeset leur acquisition(pp.79-99). Paris:Delachauxet Niesti. R., BRODERIK, V., TINCOFF,R., RODRIGUEZ, K.,MOUZAKI, TREIMAN, A., & FRANCIS,D. (1998). The foundationof literacy:Learningthe sounds of letters.ChildDevelopment,69,1524-1540. TROIA,G. (1999). Phonologicalawareness interventionresearch:A critical reviewof the experimentalmethodology.ReadingResearchQuartely,34, 28-52. VALE,A.P., &CARY,L. (1998). Escritainventadae detec9dofonemicaem leitoresprincipiantes: Preditoresdo desempenhoulteriorem leiturae escrita.RevistaPortuguesade Psicologia,32, 29-56. VERNON,S. (1998). Escrituray conscienciafonol6gicaen nihos hispanoparlantes.Infanciay Aprendizaje,81, 105-120. VIGARIO, M.,&FALE,I. (1993).A sflabado portugu6s:Umadescrigaoe algumasconsiderag6esde ordemte6rica.InActas do IXEncontroNacionalda AssociagAoPortuguesade Lingufstica(pp.465-478). Coimbra,Portugal:APL. WEBSTER,P.E., &PLANTE,A.S. (1992). Effectsof phonologicalimpairmenton word,syllableand phonemesegmentationand reading.Language, Speech, and Hearing,23,176-182. ReceivedMarch5, 2001 FinalrevisionreceivedNovember6, 2001 AcceptedJanuary7, 2002 APPENDIX A Words used to evaluate children's conceptual levels Pairs of words that are of an identical size from a syllabic point of view, but which refer to items that possess different sizes urso/rato (bear/mouse) vaca/mosca (cow/fly) galo/pinto (cock/chick) tigre/sapo (tiger/toad) elefante/lagarto (elephant/lizard) dinossauro/borboleta (dinosaur/butterfly) cavalo/formiga (horse/ant) Pairs of words that are of a different size from a syllabic perspective, but which refer to items of a similar size raposa/cao (fox/dog) hipop6tamo/boi (hippopotamus/ox) leopardo/dragdo (leopard/dragon) macaco/foca (monkey/seal) sardinha/ra (sardine/frog) gafanhoto/vespa (grasshopper/wasp) abelha/pulga (bee/flea) Singular/pluralpairs of words ovelha/ovelhas (sheep/sheep) zebra/zebras (zebra/zebras) janela/janelas (window/windows) uva/uvas (grape/grapes) navio/ navios (ship/ships) igreja/igrejas(church/churches) gato/gatos (cat/cats) 482 READINGRESEARCHQUARTERLY October/November/December 2002 37/4 APPENDIX B Battery of phonological tests Initial-syllable classification test Cards with representative drawings: Examples: bolo/nariz/navio/moinho (cake/nose/ship/mill) igreja/ouro/arroz/iogurte (church/gold/rice/yogurt) Items: uva/asa/unha/ilha (grape/wing/nail/island) enxada/ourigo/agulha/apito (hoe/urchin/needle/whistle) rolo/sapo/figo/roupa (roll/toad/fig/clothes) coelho/machado/piano/macaco (rabbit/axe/piano/monkey) garrafa/galinha/pijama/moeda (bottle/hen/pyjama/coin) tesoura/casaco/moinho/cavalo (scissors/coat/mill/horse) vaso/pipa/mesa/vaca (vase/barrel/table/cow) chupa/fato/faca/bico (lolly pop/suit/knife/brook) janela/menina/tomate/torrada (window/girl/tomato/toast) girafa/panela/cenoura/palhago (giraffe/pot/carrot/clown) bota/jarro/ninho/bola (boot/jar/nest/ball) saco/sapo/burro/mota (bag/toad/donkey/motorcycle) laranja/medalha/lagarto/pinheiro (orange/medal/lizard/pine) sino/data/dado/folha (bell/date/die/leaf) Initial-phoneme classification test Cardswith representative drawings: Examples: colher/chave/chuva/bola (spoon/key/rain/ball) j6ia/n6/jipe/pd (jewel/knot/jeep/shovel) Items: alce/urso/arca/ovo (moose/bear/arch/egg) orelha/alface/drvore/igreja (ear/lettuce/tree/church) raposa/regador/viola/boneca (fox/watering can/guitar/doll) mala/peixe/chucha/mota (bag/fish/doll/motorcycle) sumo/gola/leite/gato (juice/collar/milk/cat) buzina/cegonha/vassoura/veado (horn/stork/broom/deer) serra/copo/cama/lupa (saw/glass/bed/magnifying glass) fivela/telhado/janela/fogueira (buckle/roof/window/bonfire) boca/tigre/selo/tacho (mouth/tiger/stamp/pot) pato/p ra/milho/chuva (duck/pear/corn/rain) tijolo/bolacha/seringa/banana (brick/biscuit/syringe/banana) cebola/toalha/gaveta/cigarro (onion/towel/drawer/cigarette) lata/luva/roda/fita (can/glove/wheel/ribbon) Phonological skills and writing desenho/camisa/domin6/novelo (drawing/shirt/domino/ball of wool) Initial-syllable deletion test Cardswith representative drawings: Examples: nota (bank note) ameixa (plum) Items: av6 (grandmother) orelha (ear) radio (radio) morango (strawberry) gorila (gorilla) vela (candle) caneta (pen) foca (seal) tapete (carpet) passaro (bird) boca (mouth) seta (arrow) laio (tie) dedal (thimble) Initial-phoneme deletion test Cardswith representative drawings: Examples: noz (nut) b6ia (life buoy) Items: rio (river) rosa (rose) mel (honey) mola (clothespin) gas (gas) galo (cock) vale (valley) vila (village) cdo (dog) capa (overcoat) fio (thread) fava (bean) torre (tower) telha (tile) pdo (bread) pipa (barrel) boi (ox) bolo (cake) sal (salt) sumo (juice) lua (moon) lula (squid) dente (tooth) dedo (finger) 483