Torgesen

download Torgesen

of 8

Transcript of Torgesen

  • 7/31/2019 Torgesen

    1/8

    SPRING/ SUMMER 1998

    O NE OF the m ost comp elling findings from rec entreading research is that children who get off to apo or start in reading rarely catch up .As several studieshave now d ocum ented, the p oor first-grade reader al-

    mo st invariably continu es to be a poo r reader (Francis,Shaywitz, Stuebing, Shaywitz, & Fletcher, 1996;Torge-sen & Burgess, 1998).And the c onseque nces of a slowstart in reading become monumental as they accumu-la t e expo ne nt i a l ly over t ime. As Stano vich (1986)pointed out in his well-know n pap er on the Mattheweffects (the rich get richer and the poo r get poorer)associated with failure to acquire early word readingskil ls , these co nseq uen ces range from n egat ive at ti -tudes toward reading (Oka & Paris, 1986), to reducedopp ortunit ies for vocabulary growth (Nagy, Herman, &Ander son, 1985), to missed opp ortunit ies for develop-men t of read ing comp rehen s ion s t rat eg ies (Brown ,Palinscar, & Purcell , 1986), to less actual pract ice in

    reading than other children receive (Arlington,1984).The best solu tion to the problem of reading failure

    is to allocate resources for early identification andpreven tion. It is a tragedy of the first order that whilewe kn ow clearly the costs of wai t ing too long, fewschool districts have in place a mechanism to identifyand he lp ch ildren b efore failure takes hold. Indeed, inthe majority of cases, the re is no systematic iden tifica-tion until third grade,by which time successful remedi-ation is more difficult and mo re costly.

    School -based prevent ive e f for t s should be engi -neered to maintain growth in cri t ical word reading

    skills at roughly normal levels through out th e early el-emen tary school p eriod. Although adequate develop-ment of these skills in first grade does not guaranteethat children will continue to maintain normal growth

    in second grade without extra help, to the extent thatwe allow children to fall seriously behind at any pointduring early elementary school, we are moving to a re-med ial rather than a preventive mod el of interven -tion. Once children fall behind in th e grow th o f crit icalwo rd reading skills, it may requ ire very inten sive inter-ventions to bring them back up to adequate levels ofreading accurac y (Allington & McGill-Franzen , 1994;Vaughn & Schum m, 1996), and reading fluenc y may beeven more di ff icul t to restore because of the largeamounts of reading practice that is lost by childreneach month and year that they remain poor readers(Rashotte,Torgesen, & Wagner, 1997).

    The p urp ose of this article is to pro vide p ractical ad-

    vice about methods to prevent reading failure that isgrounded in the new knowledge about read ing wehave acquired over the past two decades. My primaryfocu s w ill be o n early identification of children at riskfor p roblems in learning to read as well as meth ods formonitoring the growth of critical early reading skills.The goal is to describe procedures that will allow edu-cators to identify children who need extra help inreading before they experience serious failure and tomo nitor the early develop men t of reading skill to iden-tify children who may require extra help as readinginstruction proceeds through elementary school.

    The advice provided in this article is based on ther e s e a r c h m y c o l le a g u e s R ic h a r d Wa g n e r, C a ro l

    Rashotte, and I have been condu cting on both p redic-tion and prevention of reading disabilities (Torgesen,Wagner, & Rashott e , 1994; 1997;Wagner, e t a l. , 1994;1997) as well as the work of many other researchersthat was reviewed in an earlier issue of this magazine(Summ er, 1995). It is guided b y several important as-sumpt ions and fac ts about read ing , read ing growth ,and reading failure that w ill be d iscussed first. Follow -ing this description of assumptions and a brief outlineof som e critical dimen sion s of pre ventive instruction, Iwill describe a number of specific measures and pro-cedures that should prove useful as educators seek

    1 AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS

    CATCH THEMBEFORE THEY FALL

    Iden tification and Assessm entTo Prevent Reading Failure in Young Children

    BY JOSEPH K.TORGESEN

    Joseph K. Torgesen is curren tly a Distin gu ished Re-

    search Professor of psychology and education atFlorida State University.For the last ten years, he hasbeen part of the research effort sponsored by the Na-tional Institutes of Health to identify the nature,causes, an d best approaches to instruction for chil-dren w ith m oderate to severe readin g problem s.Theresearch conducted at Florida State University thatis cited in this article wa s supported by grants num -bered HD23340 an d HD30988 from the National In-stitute of Child Health and Human Development,an d by grants from the National Center for Learnin g

    Disabilities a nd the Don ald D. Ham m ill Fou nda tion.

  • 7/31/2019 Torgesen

    2/8

    SPRING/ SUMMER 1998

    ways to focus more intensive in-st ruct ion on chi ldren whose needsare greatest.

    Assumptions about reading,reading growth, and reading failure

    Most of the points that will be discussed in this sec-t ion are not, in fact , mere assump tions about reading,but , rather, are wel l-established facts . However, I usethe word assumption here to convey the sense eitherthat the ideas are self-evident or that they are now as-sumed to be t rue based on consistent research f ind-ings.The first o f these assump tions is, in fact, a self-ev-ident value judgment.

    Adequ ate reading com prehension is the m ost im -porta nt u ltim ate ou tcom e of effective in stru ction inreading. The ultimate purpose of reading instruction is

    to help children acquire the skills that enable learningfrom, under s t anding , and en joyment of wr it t en lan-guage.This assump tion is not c ontro versial. No matterw h a t o n e s p e r so n a l p r e f e r e n c e s fo r in s t ru c t i o n a lmethod, the end goal is to help chi ldren comp rehendwritten material at a level that is consistent with theirgen eral intellectu al abilities.

    Two general types of skill and knowledge are re-qu ired for good reading comprehen sion. Consistentwith Goughs simple view of reading (1996), com pre-hen sion o f wri t ten material requires: 1) general lan-guage comp rehen sion abi li ty; and 2) abi li ty to accu-rately and fluen tly iden tify the wo rds in p rint. Know l-edge and active application of specific reading strate-

    gies is also required to maximize reading comprehen-sion (Mast rop ieri & Scruggs, 1997) bu t most of thevariabil i ty among children and adults in comprehen-sion o f writ ten m aterial can be accounted for by m ea-suring the two broad families of skil ls identified inGough s simp le view (Hoover & Gough, 1990). That is,good general language com preh ension and good w ordreading skills are the most critical skills required for ef-fective comprehension of written material.

    Most child ren who becom e poor rea ders experi -ence early and continuing difficulties in learninghow to accurately identify printed words. This diffi-

    cul ty i s expressed most di -rectly on two kinds of reading

    tasks. First , children destined to bep o o r r e a d e r s a t t h e e n d o f e l e m e n t a r y

    school almost invariably have difficulties under-standing and applying the alphabetic principle in deci-ph ering un familiar w ords.These ch ildren have un usualdifficulties learning to use the regular patterns of cor-respondence between letters and sounds in words asan aid in identifying new words th ey enco unter in text(Siege l, 1989) .They have t rouble sounding out un-know n w ords. Secon d, poo r readers at all grade levelsare characterized by slowe r than n ormal developm entof a sight vocab ulary of word s they can read fluen tlyand auto mat ically. Ultimate ly, it is this difficulty in rap idword recognition that l imits comprehension in olderpo or reade rs, for the se skills allow ch ildren to focus onconst ruct ing the meaning of what they are reading

    rather than spending too many of their intellectual re-sources on trying to identify the wo rds (Adams, 1990).The strongest current theories of reading growth linkph onet ic and sight w ord reading ski lls together b yshowing how good phonetic reading skil ls are neces-sa ry in t h e fo rm a t io n o f a c c u ra t e m e m o ry fo r t h espelling p attern s that are the b asis of sight w ord re cog-nition (Ehri, in press; Share & Stanovich, 1995).

    The most com m on cau se of difficu lties acqu iringearly word reading skills is weakness in the ability to

    process the phon ological features of lan guage (Liber-man, Shankwei ler, & Liberman, 1989).This i s perhapsthe most imp ortant discovery abo ut reading d ifficultiesin the last twen ty years. Weaknesses in the p ho nologi-

    cal area of language developm ent can be m easured by avariety of nonreading tasks, but the o nes most com -monly used assess phone mic awareness, wh ich can bedefined simp ly as the ability to identify, think abo ut, ormanipulate the individual sound s in w ords. Much ofour new confidence in being able to identify childrenat risk for reading failure before reading instruction be-gins depends on the use of tests of phonemic aware-ness, since this ability has been show n to be c ausally re-lated to the growth of early wo rd reading skills (Lund -berg,Frost,& Peterson,1988;Wagner,et al.,1997).

    Discovery of the core phonological problems associ-

    AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS 2

  • 7/31/2019 Torgesen

    3/8

    SPRING/ SUMMER 1998

    ated with specific reading disabil i ty has had at leastone unanticipated consequence.The abil i ty to assessthese core language problems directly has led to thediscovery that the early word reading difficult ies ofchildren with relatively low general intell igence andverbal abi l i ty are associated wi th the same factors(weaknesses in phonological processing) that interferewith early reading growth in children who have gen-eral intelligen ce in the n orm al range (Fletch er, et al.,1994; Share & Stanovich , 1995; Stanovich & Siege l ,

    1994). So, weaknesses in phon emic awareness charac-terize children with reading problems across a broadspan of general verbal abili ty. On the one h and, manychildren enter school with adequate general verbalability and co gnitive weaknesses limited to the ph ono-logica l/ l anguage dom ain . Thei r p r imary prob lem inlearning to read involves learning to translate betweenprinted and oral language. On the other hand, anothersignificant group of poo r readers, comp osed largely ofchildren from families of lower socio-economic or mi-nori ty status, ente r scho ol signi ficant ly delayed in amuch broader range of prereading skills (Whitehurst &Lonigan, in press). Since the se children are de layed notonly in phonological but also in general oral language

    skills, the y are deficient in b oth of the c ritical kinds o fknowledge and skill required for good reading compre-hen sion. Even if these children can acq uire adequ ateword read ing sk ill, t he i r ab ilit y to com preh end themeaning of what they read may be l imi ted by thei rweak general verbal abilities.

    Children with general oral language weaknesses re-quire extra instruction in a broader range of knowl-edge and ski l l s than those who come to school im-paired only in phonological ability.What is well estab-lished at this point , though, is that both kinds of chil-dren wi l l requi re spec ia l suppor t i n the growth of early word reading skills if they are to make adequatepro gress in learning to read.

    Elements of an effective preventive

    program in readingThe most critical elements of an effective program

    for the prevention of reading disability at the elemen-tary schoo l level are: (a) the right kind and qu ality ofinstruction delivered with the (b) right level of inten-sity and duration to (c) the right children at the (d)right t ime. I wi ll briefly con sider each of these ele-men ts in turn.

    The right kind and quality of instruction. It is be-yond the scope of this article to discuss instructionalmethods for chi ldren wi th phonological processingweaknesses in any depth at all. In broad stroke, theywil l benefi t from the same approach to reading in-s t ruc t ion as ch i ldren wi th normal ab i l i t i es i n th i sareastructured, systematic, and exp licitbut for th isat-risk group, such instruction is not just bene ficial, it iscritical.As experienced teachers understand (Gaskins,et al. , 1996), we cannot assume that these ch ildren w illacquire any necessary skil l for reading words unlessthey are directly taught that skill or knowledge and re-ceive sufficient op po rtun it ies to practice it . Som e ofthe wo rd-level skills and kn ow ledge the se ch ildren w ill

    require instruct ion o n include: ph onem ic awareness,let ter-sound corre spon den ces, blending skills, a smallnum ber o f p ronun ciat ion convent ions ( i. e ., s ilen t erule), use of context to help spe cify a word once it ispartially or com pletely pho nem ically decode d, strate-gies for m ulti-syllable w ords, and au tom atic rec ogni-tion of high-frequen cy irregular wo rds. It goes almostwithout saying that this type of instruction should beembedded within as many opportunit ies for meaning-ful reading an d w riting as p ossible.

    The lesson from recent large-scale prevention stud-ies (Brown & Felton, 1990; Foorm an, et al ., 1998;Torge-sen, et al., 1998;Vellutino, et al., 1997) is that it is po ssi-ble to maintain critical word reading skills of most c hil-dren at risk for reading failure at roughly average levelsif this type of instruction is provided beginning some-time during kindergarten o r first grade. However, it isalso true th at in all studies cond ucted to date, substan-t ia l p ro p o r t io n s o f c h il d re n w it h t h e m o s t s e v e rewe aknesses rem ain significantly imp aired in the se cr iti-cal skil ls fol low ing inter vent ion. For exam ple, i f weadopt the 30th percenti le as a standard for adequatereading progress, then the p ropo rtion of the total pop -ulation remaining at risk in spite of the best interven-

    tions tested to date ranges from 5 p ercent to 7 p ercent(Torgesen , 1998).Altho ugh th ese results are clearly bet-ter than the 30 percent to 60 percent of children whofrequently fall below these standards without specialinterventions, they n evertheless suggest that the re is acore of disabled readers in the population for whomwe have not yet solved th e reading pu zzle.

    It is almost certain that some additional answers tothis question will come as we direct our attention tothe q uali ty and intensity, as well as the con tent, of ourinst ruct ion. For example, Juel (1996) has show n theimpo rtanc e of a particular kind o f scaffolded interac-t ion between teacher and chi ld in increasing under-standing and application of phonemic reading skil ls,

    and these types of interactions are also prescribed inthe teacher manuals of at least two widely used in-structional programs designed for children with read-ing disabili t ies (Lindamo od & Lindamo od, 1984; Wil-son, 1988).We turn n ow to a brief consideration of is-sues surrou nd ing inten sity of instruction.

    The right level of intensity. Greater intensity and d u-ration of instruction is required because the increasedexplicitness of instruction for children who are at riskfor reading failure requires that more things be taughtdirectly by the teach er. Intensity of instruc tion is in-creased primarily by reducing teacher/student ratios.Unless beginning reading instruction for children withphonological weaknesses is more intensive (or lasts

    s ign i fi can t ly longer) t han norm al ins t ruc t ion , thesechildren will necessarily lag significantly behind theirpe ers in reading grow th. An e ffective p reventive pro-gram may involve several levels of instructional inten-sity ranging from small-group to one-on-one instruc-t ion, depe nding upon the severity of the r i sk factorsfor each child.

    The right children a t the right tim e. These factorsare considered together because th ey are both t ied di-rectly to th e availability of accurate identification pro -cedu res at various age levels. That is, to be most effi-

    3 AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS

  • 7/31/2019 Torgesen

    4/8

    SPRING/ SUMMER 1998

    cient, a preventive program should be focused on th echildren who are most in need of special instruction.The efficiency of the entire process will be improvedif procedures are avai lable to accurately target theright children very early in th e p roce ss of reading in-st ruct ion. Al tho ugh t iming issues w i th regard topreventive instruction have not b een comp letelyresolved by research (Torgesen, et al ., 1998), wed o k n o w , fo r e x a m p l e , t h a t in s t r u c t io n i nphonological awareness during kinder-

    garten can have a positive effect onreading growth after formal readinginstruction begins in the first grade(Lundberg, Frost , & Peterson, 1988).T h u s, I h a ve p r o c e e d e d o n t h e a s -sumpt ion that i t wi l l be useful toidentify high-risk children at sometime during the kindergarten yearso that preventive work may beginas early as p ossible.

    How accurate are currently

    available early iden tification

    procedures?As stated earlier, the p rimary pu rp ose of this article is

    to make some practical suggestions about proceduresand tests that can be used to identify children for pre-ventive reading or p rereading instruction. From th e ou t-set, how ever, it is imp ortant to rec ognize th at our abilityto predict which children will have the most seriousreading d ifficulties is still far from p erfect. For exam ple,in a recent comprehensive review of early identificationresearch (1998), Scarborough po inted o ut that all stud-ies continue to report substantial levels of two kinds ofpred iction errors.

    False positive errors are made when chi ldren who

    will eventually becom e good readers score b elow th ecut -of f score on the pred ic t ive ins t rument and a refalsely iden tified as at risk. In gen eral, the pro po rtionof this type of error has ranged between 20 percentand 60 percen t, with an average of around 45 perce nt.That is, almost h alf of the children identified d uringkindergarten as at risk turn out n ot to h ave seriousreading problems by the end of first grade. False n ega-tive e r ro r s o c c u r w h e n c h i ld re n w h o la t e r e x h ib i treading problems are identified as not being at risk.Typical perce ntages of false n egative e rror s range from10 percent to 50 p ercent , wi th an average of around22 percen t.That is, on average, current p rocedures failto identify about 22 percent of children who eventu-

    ally end up w ith serious re ading difficulties.In any given study,th e relative pro po rtion o f false p os-itive and false negative errors is somewhat arbitrary,since it depen ds on the level of the c ut-off score. For ex-ample, we reported a significant reduction in the per-centage of false negative errors within the same sampleof chi ldren by doubl ing the number of chi ldren weident i fied as at r i sk (Torgesen , in press; Torgesen &Burgess, 1998). Our goal was to identify, during the firstsemester of kindergarten, the ch ildren m ost at risk to bein the bo ttom 10 p ercent in w ord reading ability by thebeginning of secon d grade. When w e selected the 10

    p e r c e n t o f ch ild r e n w h o s c or e dlow est on our predictive tests, ourfalse negative rate was 42 percent(we missed almost half the chil-d r e n w h o b e c a m e e x t r e m e lyp o o r r e ad e r s) . H o w e ve r, w h e nwe identified the 20 percent ofchildren who scored lowest onour measures, the false nega-tive rate w as reduced to 8 per-

    cent. As a practical matter, ifschools desire to maximizetheir chances for early in-

    tervention with the m ost im-paired children, they should pro-

    vide th is interven tion to as many ch ildren as possi-ble.This is less of a waste o f resource s than it might

    seem at first glance, because, although many ofthe falsely identified children receiving interven-

    tion may not be among the most seriously disabledreaders, most o f them are likely to b e b elow-average

    reader s (Torgesen & Burgess, 1998).Two other pieces of information are relevant to the

    selection of procedures for early identification of chil-

    dren at risk for reading difficulties. First, pre diction ac-curacy increases significantly the longer a child hasbee n in scho ol. Prediction of reading disabilities fromtests given at the beginning of first grade is signifi-cantly more accurate than from tests administered d ur-ing the first semester of kindergarten (Scarborough,1998;Torgesen, Burgess, & Rashot te , 1996). Given thewidely varying range of childrens preschool learningopp ortun ities, many children m ay score low o n earlyident i f i ca t ion ins t rument s in the f i r s t semeste r o f kindergarten simply because the y have no t had the op -po rtuni ty to learn the skill s. Howe ver, if pre readingskills are actively taught in kinde rgarten, some of thesedifferences may be reduced by the beginning of the

    second semester of school. Thus, I would recommendthat the screening p rocedures described h ere not beadministered unti l the beginning of the secon d seme s-ter of kindergarten, at which t ime they will be m uchmore efficient in identifying children who will requiremore intensive prevent ive inst ruct ion in phonemicawaren ess and othe r early reading skills.

    Second , although batteries co ntaining mu ltiple testsgenerally provide better prediction than single instru-men ts, the inc rease in efficienc y of mu lti-test batteriesis generally not large en ough to w arrant the extra timeand resources required to administer them (Scarbor-ough, 1998). Thus, I recommend an ident ificat ion pro-ced ure involving administration o f two tests: 1) a test

    of knowledge of let ter names or sound s; and 2) a mea-sure of ph onem ic awareness. Measures of let ter know l-edge continue to be the best single predictor of read-ing difficult ies, and m easures of pho nem ic awarenesscontribute addit ional predictive accuracy. In our expe -rience, tests of let ter name know ledge are most p redic-tive for kinde rgarten ch ildren , and tests of letter-soun dknow ledge are mo st predictive for first grader s. Sincereading growth is influenced by noncognitive factorssuch as at tent ion/mot ivat ion and home background(Torgesen, et al. , 1998), as well as spec ific know ledge

    AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS 4

  • 7/31/2019 Torgesen

    5/8

    SPRING/ SUMMER 1998

    and skills, scores from th ese o bjective testsm i gh t p ro f it a b ly b e su p p le m e n t e d w i t hteacher ratings of behavior and attentionto identify children most at risk for sub-seque nt d ifficulties in learning to read.

    How should pho nemic

    awareness be assessed?S i n c e r e se a rc h e r s f i r s t b e g a n t o

    s t u d y p h o n o l o g ic a l aw a re n e s s i n t h eearly 1970s, mo re than twe nty differen t taskshave been used to measure awareness of ph onem esin words.These measures can be grouped into threeb ro a d c at e go r ie s : so u n d c o m p a r iso n , p h o n e m e se g -mentat ion, and p honeme blending.

    s Soun d com parison tasks use a number of differ-ent formats that all require children to make compar-isons betwe en the soun ds in different w ords. For ex-ample, a child might be asked to indicate w hich wo rd(of several) begins or ends with the same sound as atarget word (i .e. ,Which word begins with the samefirst sound as cat: boy, cake, o r fan ?). Additio na lly,

    tasks that require children to generate words that havethe same first or last sound as a target word would fallin th is category. Sound com parison tasks are amongthe least difficult measures of phonemic awareness,and thus are particularly appropriate for kindergartenage children .s Phoneme segmentation tasks involve counting,

    pron ounc ing,d eleting, adding, or reversing the individ-u a l p h o n e m e s i n w o rd s . C o m m o n e x a m p l e s o f th i st y p e o f t a s k r e q u i r e p r o n o u n c i n g t h e i n d i v i d u a lphonemes in words (Say the sounds in cat one at atime.),d eleting soun ds from wo rds (Say cardwithoutsaying the /d/ sound.), or co unting sounds (Put on emarker on the l i ne for each sound you hear in the

    word fast .)s Phonem e blending skill has only been measuredby one kind of task.This is the sound-blending task inwhich the tester pronounces a series of phonemes inisolation and asks the child to blend them together toform a word (i .e. ,What word do these sounds make,

    / f/ - / a/ - / t/ ?). Easie r varian ts o f th e so un d-ble ndin gtask can be produced by allowing the child to choosefrom two or th ree p i c tures the word tha t i s repre-sented by a series of phone mes.

    In general, these different kinds of ph onem ic aware-ness tasks all appear to be measuring essentially thesame cons t ruc t , o r ab i li ty. Al thou gh some research(Yop p, 1988) h as indicated th at the tasks may involvedifferen t levels of inte llectu al com p lexity, and th eremay be some differences between segmentation andblending tasks at certain ages (Wagner, Torgesen, &Rashotte, 1994), for the most part , they all seem to bemeasuring growth in the same general ability (Hoien,et al. , 1995; Stanovich, Cunningham, & Cramer, 1984).Sound comparison measures are easiest and are sensi-t ive to emergent levels of phonological awareness,while segmentation and blending measures are sensi-tive to differences among children during later stagesof developm ent involving re finem ents in exp licit levelsof awarene ss. Measures o f sensitivity to rhym e (Which

    word rhymes with cat:leg or m at?) are notincluded as measures of pho nem ic aware-ness because they appe ar to be measuringsome thing a little differen t, and less pre -dictive o f reading disabilities, from th osemeasure that ask children to attend to

    in d i v id u a l p h o n e m e s . Fo r t h e s a m ereason, measure s of syllable aware-ness are not included in this group.

    Measures of phonemic awareness

    that are suited for early identificationp u rp o se s i n c l u d e t h e fo l l o w i n g t h re ew idely used tests:

    The Phonological Awareness Test (Ro b e r t so n &Salter, 1995). This test contains five differen t me asuresof pho nem ic awarene ss, plus a measure of sensit ivityto rhyme. The five measures of phonem ic awarenessare segmenta t ion of ph onem es , ph onem e iso la t ion ,p h o n e m e d e le t io n , p h o n e m e s u b st it u t io n , a n dphon eme blending.The phon eme isolation test, wh ichrequires children to pron ounc e the first , last , or middlesounds in wo rds, would appear to have the most ap-propriate level of difficulty for kindergarten screening(the test should be easy enough so that only the mostdelayed ch ildren w ill do p oorly on it), and any of theothers could be used for first- or second-grade assess-ments. The Phonological Awareness Test is nationallynorm ed on children from age five through nine, and i tcan b e orde red from LinguiSystems, 3100 4th Avenue ,East Moline, IL 61244-0747. Phone: 800-776-4332. Thecos t o f a t est manual , t es t supp lies , and f ift een t es tboo klets is $69.

    The Test of Phonological Awareness (Torgesen &Bryant, 1994). This test was designed as a group-admin-istered test of phonemic awareness for kindergartenand first-grade ch ildren . It was sp ecifically con structe dto be most sensit ive to children with w eaknesses in d e-

    velopm ent in this area, wh ich he lps make i t appro pri-ate for iden tifying at-risk children . The kind ergarte nversion of the test requires children to notice whichwords (represented by pictures) begin with the samefirst soun d, w hile the first-grade version asks them tocomp are words on the b asis of their last sounds. It canbe easily administered to grou ps o f five to te n ch ildrenat a t ime. The Test of Phonological Awareness is na-t ionally norm ed, and i t can b e ord ered from PRO-EDPublishing Com pany, 8700 Sho al Creek Blvd., Austin,TX 78757-6897. Phone : (512) 451-3246. The cost of atest manual and a supply of fifty test forms (twenty-fivekinde rgarten version, twen ty-five elementary schoolversion) is $124.

    Th e Yopp-Singer Test of Phon em e Segm enta tion(Yop p, 1995) is a brief test of childrens ability to iso-late and pron ounc e the individual phone mes in words.This is a task that has been widely used in research onph onem e awareness over the past twenty years, and itis highly correlated with other measures of phonemea w a r e n e s s . T h e t e s t w a s d e s ig n e d fo r c h i ld r e n i nk inde rgar t en , but i t shou ld a lso be app ropr ia t e foridentifying ch ildren wh o are w eak in p hon emic aware-ness during first grade.The test has twenty-two itemsthat are all of the same type and that ask the child topronounce each of the phonemes in words that varyfrom two to three phonemes in length.The test does

    5 AMERICAN EDUCATOR/AMERICAN FEDERATION OF TEACHERS

  • 7/31/2019 Torgesen

    6/8

    SPRING/ SUMMER 1998

    no t have nor ms w ith it, but it is available free in vol-ume 49 (1995) of the widely read journ al The ReadingTeacher, pp.20-29.

    The measurement

    of letter knowledgeIn a ll o f o u r r e se a rc h , w e h a ve m e a su re d l e t t e r

    k n o w l e d ge in t w o w a ys . We m e a su re let ter na m eknowledge by presenting each letter in simple upper-

    case type on a single card and asking for its name.Thescore on this test is simply the number of letters forwh ich the ch i ld can g ive the app ropr i at e name. Wemeasure letter-sound knowledge by presenting all let-ters in lower-case type and asking for the sound theletter makes in w ords. If a conson ant let ter can com-monly represent two d if fe ren t sounds ( i. e ., c , g ) weprobe for the second sound, and we also ask for thelong and short p ronun ciation o f each vowel.The scoreis the total num ber o f soun ds the c hild can give. Wehave foun d that letter-nam e kno w ledge is a more sen si-t ive predictor for kindergarten ch ildren , w hile let ter-sound knowledge is a better predictor for children infirst grade. Two tests that p rovide nationally standard-

    ized n orms for p erformanc e on let ter-name and letter-sound know ledge are:

    Th e letter identification subtest of the WoodcockReading Mastery Test-Revised (Woodcock, 1987).Thistest does not measure simple let ter-name knowledgein the w ay we assess it , because i t presents let ters inseveral differen t fonts, some o f which m ay be u nfamil-iar to ch ildren . It also allow s children to give either th ename or the sound the l e tt e r makes in w ords. How-ever, children w ho perform p oorly in kindergarten (donot know the names of very many let ters) wi l l notr e a c h t h e m o re d i ff ic u l t it e m s , so t h a t t h e i r s c o reshould be quite com parable to a more straightforwardtest of let ter-name knowledge. The Reading Mastery

    Test-Revised is available from American Guidance Ser-v ice , 4201 Woodland Road , Circ le P ines , MN 55014-1796. Phone (800) 328-2560. The cost for the manualand forms is $314.95.

    Th e graphem es subtest of the Phonological Aware-ness Test (Rob ertson & Salter, 1995). This test providesa comprehensive assessment of let ter-sound knowl-edge exten ding from single co nsonants (i .e. , b,c,k,m)through vow el digraph s and diphthon gs (i.e., ea, ai, ow,oy).As mention ed be fore, it is standardized on ch ildrenfrom aged five through nine.

    Is it necessary for a test to be

    nationally standardized for it to beuseful in early identification ?

    This issue is important because of the potential ex-pense of employing standardized measures in large-scale scree ning efforts. Nation ally based no rms are notrequired to ident i fy which chi ldren wi thin a givenclassroom or school are weakest in phonemic aware-ness and letter knowledge. However, the prop ortion ofch i ldren wh o co me to school w ith w eak sk il ls andknowledge in these areas wi l l depend somewhat onspecific aspects of their preschool language and liter-

    acy environment and will almost certainly vary fromscho ol to schoo l across differen t comm un ities. Testswith national norms can help to pinpoint classes orschools in w hich a special effort mu st be made to en-hance p hon ological awareness in children p rior to, orduring, reading inst ruct ion. For example, a classroomin which 75 percent of the children performed belowthe 20th percent i le ( in the bot tom 20 percent of al lchildren), will require more instructional resources toprepare children for learning to read than a classroom

    in which only 10 percent of the children scored thatlow.Withou t nor ms, it is po ssible to identify weak ch il-dren w ithin a given class or schoo l, but it is not p ossi-ble to determine what proport ion of chi ldren in theentire school may require intervention because of rela-t ively weak pre reading skil ls and know ledge. On th eone hand, if classroom resources allow extra help foronly a fixed num ber o f children (say, 20 p ercent to 30percen t) , then me asures wi thout nat ional norms canbe used to identify the group of children within theclassroom m ost in nee d of intervention. On the o therhand , i f t he goa l i s t o de t e rm ine th e amoun t of re -sources that may be needed to help all children withrelatively weak skills in these areas, the n no rm ative

    measures will be required.The com bination o f let ter know ledge and ph onem ic

    awareness tests I have recommended should take nomore than ten to fifteen minutes pe r child to adminis-ter.The tests do not require highly trained personnel toadminis t e r them , a lthough anyone wh o t es t s youngchildren must be very familiar with the tests and beable to establish a supportive rapport.

    Mon itoring grow th

    in early reading skillsOnce reading instruction begins, the be st predictor

    of future reading growth is current reading achieve-

    men t, and th e m ost cri t ical indicators of good progressin learning to read during the early elementary periodare measures of word reading skil l. Children w ho endup as poor readers at the en d of elemen tary school area l m o s t i n v a r i a b l y t h o se w h o f a i l t o m a k e n o rm a lpro gress in th ese skills dur ing the first years of eleme n-tary scho ol . These ch i ldren are m ost frequ ent ly im-paired in both the ability to apply phonetic strategiesin reading new words and in the abi l i ty to ret r ievesight w ords from m emo ry. They no t on ly have di ff i-cul ty becoming accurate in the appl icat ion of thesepro cesses b ut also the y frequ ently have spe cial difficul-t ies w ith becom ing fluent in their app lication. Beforediscussing specific methods for the diagnostic assess-men t of these w ord reading skills, one gen eral issue re-garding read ing assessmen t requ ires discussion.

    First , the assessment that w ill be recom mende d h ereis very different from the authentic l i teracy assess-mentthat is currently advocated by many reading pro-fessionals (Paris, et al ., 1992). Authe ntic assessmen t isdifferent in at least two ways from the reading assess-men t measures we w ill be discussing. First, the goal ofauthe ntic assessmen t is to measure childrens app lica-tion of broad literacy skills to auth en tic tasks, like gath-er ing inform at ion for a repo r t , use of li t e racy as amed ium for social interactions, or ab ility to read a selec-

    AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS 6

  • 7/31/2019 Torgesen

    7/8

    SPRING/ SUMMER 1998

    t ion and then wri te a response to i t . I t also seeks tomeasure childrens en joyment, own ership, and involve-men t in literacy activities both at school and at hom e.

    This kind of assessment is a clear complement tothe type of assessments we will describe for monitor-ing growth in word level reading skills.All of the liter-acy outcome s that are p art of authentic assessment areimportant parts of a total literacy assessment program.After all, if a child can re ad, bu t does n ot en joy readingand does not apply important literacy skills to every-

    day tasks, then some imp ortant goals of l iteracy in-struction h ave no t been attained.H o w e v e r, s in c e t h e se p ro c e d u re s a r e fo c u se d o n

    high-level reading ou tcomes, they cann ot p rovide p re-cise information about level of performance on impor-tant su bski ll s in re ading. If a chi lds overal l pe rfor-manc e o n auth entic l iteracy tasks is l imited, i t is fre-quently difficult to obtain a precise estimate of thespec ific comp one nt skills that are we ak. The goal oft he k ind of assessment s we wi l l d i scuss here i s t oquantify the degree of skill a child possesses in wordidentification p rocesses that have been show n to be acritical foun dation for o verall reading suc cess.

    Commo nly used diagno stic

    measures of word reading abilityIt is beyond the scope of this article to identify all

    the available tests of w ord level reading skills. Rather, Iwill provide examples of measurement strategies fromthe mo st common ly used me asures.

    Sight word reading ability. Two measures are w idelyused in this area, and bo th involve the same assessmen tstrategy.The Word Identification subtest from the Wood-cock Reading Mastery Test-Revised (Woodcock, 1987),and the reading subtest of the Wide Range Achieve-m ent Test-3 (Wilkinson , 1995) both req uire children toread lists of words that gradually increase in length and

    complexi ty whi le decreasing in frequency of occur-rence in printed English. For examp le, the easiest threewords on the Word Identification subtest are go, the,an d m e, words of mid-level difficulty are pioneer, in -qu i re , a n d weal th , a n d t h e h a r d e s t t h r e e a r eepigraphist, facetious, an d shillelagh.

    Neither of these widely used tests place stringentt ime p ressure on students, so both p honet ic decodingprocesses and sight word processes can be used toident i fy words on these l is t s. Both t es t s have be ennor me d nationally, and on e of their strengths is thatthey allow a direct assessment of childrens ability toi d e n t i fy w o r d s s o l e ly o n t h e b a s is o f t h e w o r d sspelling.When reading text, children also have con textclues available to assist w ord identification, and thustext-based me asures, although they m ay be more au-thenticin on e sense, are less direct in the ir assessmentof the kinds of word-processing skills that are particu-larly deficien t in children w ith reading pro blems.

    Phonetic reading ability. The single best measure ofchildrens ability to apply knowledge of letter-soundcorresponden ces in decoding words i s p rovided b ymeasures of nonword read ing (Share & Stanovich ,1995).Th e Word Attack sub test o f the Woodcock Read-ing Mastery Test-Revised (Woodco ck, 1987) is a good

    examp le of this kind of diagnostic test. It consists of aseries of increasingly complex nonwords that childrenare asked to soun d ou t as best the y can. The th reeeasiest items on the test are ree, ip, an d din; items ofmoderate difficulty are rejune, depine, an d viv; an dthe three hardest i tems are pn ir, ceism in adolt, an dbyrcal. Because the words are presented out of con-tex t, the y stress the c hilds ability to fully analyze eachword to produce the correc t p ronuncia t ion . On theother h and, measures such as this do not allow an as-

    sessment of childrens ability to combine phonetic de-coding with use of context to arrive at a words cor-r e c t p ro n u n c i a tio n . H o w e v e r, s in c e b o t h g o o d an dpoor readers appear able to use context equally well( a s lo n g a s t h e c o n t e x t is u n d e r s t o o d , Sh a r e &Stanovich, 1995), this is not an imp ortant omission ona diagnostic me asure o f word reading ability.

    Word reading fluency. Word reading fluency mea-sures have typ ica l ly measured ra t e of read ing co n-nected text. One o f the m ore w idely used measures inthis area is the Gray Oral Reading Test-3rd Edition .(Wiede rho lt & Bryant, 1992). This test con sists of thir-teen increasingly difficult p assages, each followe d byfive comprehension questions.A measure of oral read-

    ing rate is obtained by recording the time it takes forthe ch ild to read each passage. One p otential problemw ith the Gray Oral Reading Test is that it doe s no t p ro-vide a very sen sitive m easure of individual differen cesin word reading abil i ty at very low levels of perfor-mance, such as those found in beginning first graders,or d i sab led readers th rough second grade . The pas-sages simp ly begin at to o h igh a level for c hildren w ithvery poor or undeveloped reading skills to display thewo rd reading skills they ac tually po ssess.

    In an effort to provide measures of fluency and ac-curacy in w ord re ading skill that are simp le to adm inis-ter and sensi t ive to individual di fferences across abroad ran ge of reading skills, we are c urre ntly develop -

    ing simple measures ofWord Reading Efficiency an dNon Word Effi cien cy (Torgesen & Wagner, 1997) . Inboth of these measures, children are sho wn lists of in-creasingly difficult words and nonwords and asked toread as many words as possible in forty-five seconds.There are two forms to each test , and the childs scoreis simply the average number of words read in forty-five secon ds. Ini t ial evaluat ion s ind icate that thesemeasures are very reliable (parallel form reliabilitiesvary between .97 and .98 for kindergarten throughfifth grade).They are also highly correlated with corre-sponding measures from the Woodcock Reading Mas-ter Test-Revised at early grades (when children oftenrun out of words they can read before they run out of

    t ime, correlations range from .89 to .94) and slightlyless correlated (.86 to .88) at fourth grade, wh en flu-ency of word reading processes becomes more impor-tant to pe rformanc e on the tests.These tests have beenstandardized nationally and will be available from PRO-ED pub lishing com pany in late summ er 1998 . If a sin-gle form o f each test is adm inistered, it will provide in-dices of growth in phonetic decoding and sight wordreading that can be administered several times duringthe year and th at take a very short amoun t of t ime togive.

    7 AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS

  • 7/31/2019 Torgesen

    8/8

    SPRING/ SUMMER 1998

    To sum marize, adequate mon itoring of the growthof childrens wo rd read ing abilities sho uld include o ut-of-con text measure s of wo rd reading ability, ph one ticdecoding ability (as measured by ability to read non-wo rds), and word reading f luenc y. The fluenc y mea-sures become more important after about second toth i rd grade , wh en ch ildren h ave acqui red a fund of word reading skills they can apply with reasonable ac-curacy.Measures that involve out-of-con text w ord re ad-ing more di rect ly assess the kinds of word reading

    ski l l s that are part icularly problemat ic for chi ldrenwith reading disabil i t ies because they eliminate thecontextual suppo rt on wh ich th ese children rely heav-ily.To obtain a complete picture of overall reading de-velopm ent, how ever, it is also important to o bserve thew ay that the child integrates all source s of informationabout w ords in text , and this can only be estimated bycarefully observing children as they read connectedpassages.

    REFERENCESAdams, M.J. (l990).Begin nin g to read :Thin kin g an d learnin g a bou t

    prin t. Cambridge, MA: MIT Press.Al ling ton , R.L. ( l984) . Conten t coverage and con tex tual read ing in

    reading groups.Jou rnal o f Read in g Behav ior, 16,85-96.Allington, R.L. , & McGill-Franzen, A. (1994). Reading and the mildlyhandicapped. Intern ation al Encycloped ia of Education , Oxford,U.K.:Pergamon.

    Brown, A. L. , Palincsar, A. S. , & Purcell, L. (l986). Poor readers:Teach,dont label. In U. Neisser (Ed.) The School Achievem ent of Minor-ity Children: New Perspectives. (pp . 105-143) NJ: Lawrence Erl -baum Associates.

    Ehri, L.C. (in press). Grapheme -pho neme know ledge is essential forlearning to read w ords in English. In J. Metsala & L. Ehri (Eds.).Word recognition in beginning reading. Hillsdale, NJ: LawrenceErlbaum Associates.

    Fletch er, J. M., Shaywitz, S. E., Shankw eiler, D. P., Katz,L.,Liberm an, I.Y.,Stuebing, K.K., Francis, D. J. , Fowler,A. E. , & Shaywitz, B.A. (1994).Cognitive pro files of reading disability: Comparisons of discrep-ancy and low achievement definitions. Jou rn al of Edu cat ionalPsychology, 86, 6-23.

    Foorman, B.R., Francis, D.J., Fletch er, J.M., Schatsch neider, C., & Mehta,P. (1998). The role of instruction in learning to read: Preventingreading failure in at-risk children.Jou rn al of Education al Psychol-ogy.

    Francis, D.J., Shaywitz, S.E., Stueb ing, K.K., Shaywitz, B.A., & Fletch er,J.M. (1996). Developme ntal lag versus de ficit mo dels of reading dis-ability:A longitudinal, individual grow th curves analysis.Jou rn al ofEducation al Psychology,88,3-17.

    Gask ins , I .W. , Ehri , L.C. , Cress , C ., OHara, C., & Donnelly, K. (1996) .Procedures for word learning:Making discoveries about words. TheRead in g Teach er, 50,312-327.

    Gough, P. B. (1996).How children learn to read and w hy they fail .An-na ls of Dyslexia, 46, 3-20.

    Hoien , T. , Lundberg , I ., Stanovich , K. E. , & Bjaalid , I . (1995) . Compo-nents of phonological awareness.Read in g an d Writ in g:An Inter-disciplina ry Journ al, 7,171-188.

    Hoover,W.A., & Gough, P.B. (1990).The simple view of reading.Read -ing an d Writing, 2,127-160.Juel, C. (1996).Wh at makes literacy tutoring effective?Reading Re-

    search Quarterly, 31,268-289.Liberman, I.Y. , Shankwe iler, D., & Liberman, A.M. (1989). The alpha-

    betic principle and learning to read. In Shankwe iler, D. & Liber-man , I .Y. (Eds. ), Phonology and reading disability: Solvin g thereading puzzle (pp .1-33).Ann Arbor, MI:U. of Michigan Press.

    Lindamood, C.H., & Lindamood, P. C.(l984).Aud itory Discrim in ationin Depth. Austin, TX:PRO-ED, Inc.

    Lundb erg, I.,Frost, J., & Peterson, O. (l988).Effects of an extensive pro-gram for stimulating phonological awareness in pre-school chil-dren.Read in g Research Qua rterly, 23,263-284.

    Mastropieri, M.A. & Scruggs,T.E. (1997). Best p ractices in p romotingreading comprehension in students with learning disabilities:1976-

    1996.Rem edial an d Special Edu cati on , 18,197-213.Nagy, W.E. , Herman , P.A. , & Anderson , R.C. (1985) . Learn ing words

    from context.Read in g Research Qua rterly, 20,233-253.Oka, E., & Paris, S. (1986). Patterns of motivation and reading skills in

    unde rachieving children. In S. Ceci (Ed.),Han dbook of cognitive,social, and neu ropsychological aspects of learnin g disabilities(Vol.2). Hillsdale, NJ: Erlbaum.

    Rashotte , C.A.,Torgesen, J.K., & Wagner, R.K. (1997). Growth in read-ing accuracy and fluency as a result of intensive intervention.

    Paper presented at the annual meetings of the Florida Branch ofthe International Dyslexia Association,Miami,FL, Sept emb er.

    Scarborough, H.S. (1998). Early identification o f children at risk for

    r ead in g d i sab i l it i e s : P h o n o lo g ica l aw aren e ss an d so m e o th e rprom ising pred ictors. In Shapiro, B.K., Accardo, P.J., & Capute, A.J.(Eds.) Specific Readin g Disability:A view of the spectrum . (pp .75-120).Timon ium, MD:York Press, Inc.

    Share, D.L.,& Stanovich, K.E. (1995). Cognitive p rocesses in e arly read-ing developm ent:A mode l of acquisition and individual differen ces.Issues in Education :Con tribu tions from Education al Psychology,1, 1-57.

    Siegel,L.S. (1989). IQ is irrelevant to th e de finition of learning d isabili-ties.Jou rnal of Learn in g Disab ilit ies, 22, 469-479.

    Stanovich, K.E., & Siegel,L.S. (1994).The ph enotyp ic performanc e pro-file of reading-disabled ch ildren: A regression-based test o f theph onological-core variable-differen ce mode l.Jou rn al of Edu ca-tional Psychology, 86,24-53.

    Stanovich, K.E. (l986).Matthew effects in reading:Some conseque nce sof individual differences in the acquisition of literacy.Readin g Re-search Quarterly, 21,360-406.

    Stanov ich , K.E. , Cunningham, A.E. , & Cramer , B.B. ( l984) .Assessingph onological awarene ss in kindergarte n children: Issues of taskcomparability.Jou rnal of Experi m en tal Chi ld Psychology, 38 ,175-190.

    Torgesen , J.K. ( in p ress) .Assessment and instruct ion fo r pho nemicawareness and w ord reading skills. In H.W. Catts and A.G. Kamhi(Eds.) Langua ge Ba sis of Readin g Disabi lities. Needham Heights,MA:Allyn & Bac on .

    Torgesen, J.K. (1998) Indiv idu al differen ces in respon se to read in gintervention. Paper presented at the Pacific Coast Research Con-ference, LaJolla, CA, Febru ary.

    Torgesen, J.K., & Burgess, S.R. (1998). Consistency of reading-relatedpho nological proc esses throughout early childhood : Evidenc e fromlongitudinal-cor relational and instru ction al studies. In J. Metsala & L.Ehri (Eds.). Word Recognition in Beginning Reading. Hillsdale, NJ:Lawrence Erlbaum Associates.

    Torgesen, J.K. , Burgess, S. , & Rashotte, C.A. (1996). Predicting phono-logically based reading disabilities:What is gained by waiting ayear? Paper presented at the annual meetings of the Society for theScien tific Stud y of Reading, New York, Apr il.

    Torgesen, J.K.,Wagner,R.K., & Rashotte, C.A. (1994). Longitudinal stud-ies of phonological processing and reading. Jou rn al of Learn in gDisab ilit ies, 27,276-286.

    Torgesen, J.K. ,Wagner, R.K., & Rashotte, C.A. (1997).The preventionand rem ediation of severe reading disabilities: Keeping the en d inmind. Scientific Studies of Reading, 1, 217-234.

    Vaughn, S. & Schumm, J.S. (1996) Classroom ecologies: Classroom in-teractions and implications for inclusion of students with learningdisabilities (p p. 107-124). In D. L. Speec e & B.K. Keogh (Eds.),Re-search on classroom ecologies. Mahwah, NJ:Lawren ce Erlbaum As-sociates.

    Wagner, R.K.,Torgesen, J.K., & Rashotte, C.A. (1994).The developmentof reading-related ph onological processing abilities: New evidence

    of bi-directional causality from a latent variable longitudinal study.Developm ental Psychology, 30,73-87.

    Wagner, R.K.,Torgesen, J.K., Rashotte, C.A., Hecht, S.A., Barker, T.A.,Burgess, S.R. , Donahue, J. , & Garon,T. (1997). Changing causal rela-tions between phonological processing abilities and word-levelreading as children develop from beginning to fluent readers:Afive-year longitudinal study. Develop m en ta l Psychology, 3 3 , 4 6 8-479.

    Whitehurst , G.J. & Lonigan , C.J. ( in p ress) . Ch i ld developme nt andeme rgent literacy. Child Development, 69 .

    Wilson , B.A. (1988) .In structor m an ual. Millbur y, MA: Wilson Lan-guage Training.

    Yopp , H.K. (l988).The validity and reliability of phone mic awarene sstests.Read in g Research Qua rterly, 23,159-177.

    AMERICAN EDUCATOR/ AMERICAN FEDERATION OF TEACHERS 8