DOCUMENT RESUMEED 080 179PS 006 664AUTHORArcher Philip D Sewall Michae - PDF document

1 DOCUMENT RESUMEED 080 179PS 006 664AUTHO
DOCUMENT RESUMEED 080 179PS 006 664AUTHORArcher, Philip D.; Sewall, Michael B.TITLECompensatory Prekindergarteners' I.Q._Gain Correlatedwith Third Grade Reading Achievement._PUB DATEFeb 73NOTETOp.; Paper presented at the Annual Meeting of theAmerican Educational Research Association (NewOrleans, Louisiana, February 25 - March 1,1973)EDRS PRICEMF-$0.65 HC Not Available from EDRS.DESCRIPTORS*Achievement Gains; *Compensatory Education Programs;Correlation; *Intelligence Quotient; *IntelligenceTests; Preschool Education; *Preschool Programs;*Reading Tests; Standardized Tests; TechnicalReportsIDENTIFIERSNew York State Pupil Evaluation Program; StanfordBinet Intelligence TestABSTRACTThe purpose of this study was to investigate thecorrelation between gain on the Stanford-Binet Intelligence Test(S-B) in compensatory prekindergarten programs, and scores on the NewYork State Pupil Evaluation Program (PEP), Reading Section, overthree years later. Four hundred and five subjects showed acorrelation of 0.42 for the pre-post S-B gain correlated Ath PEP,when corrected for regression, versus 0.13 for uncorrected S-B gain._Additional correlations were compute

2 d for restricted IQ ranges. T4isstudy le
d for restricted IQ ranges. T4isstudy lends support to compensatory education programs attempting coraise IQ's. [Not available in hard copy due to marginal legibility oforiginal document.] (Author) FILMED FROM BEST AVAILABLE COPYt.U S OEPARTMENTOF HEALTHEDUCATION a WELFARENATIONAL INSTITUTE OFEDUCATION--,sDC,ENT HAS BEEN REPROJt/CED EX:CTLY AS RECEivED F RO%,t,E PERSON OR CRG.,N.ZATIONOR,GNA't^dC.PO,N'S OF VIEW OR OPINIONSSTATED DO NO, NECESSARILY REP1ESENT Ovv,C,L NATIONAL INS TITUTE OFED,CA1 ION POSIT ON OR POLICYCory Prekindergarteners 'I .Q . Gail CorrelatedWith Third Grade Reading AchievementSCOPE OF INTEREST NOTICEThe ERIC Facility has assignedthis document for pr a:sumo)PstoPIn ow lodgement, thus documentnatio of interest to the dewing-poutnoted to the right. Index-ing should reflect their spacialpoints of view.Philip'rcherWake Forest UniversityardMichael B. Sewall-Mohawk Valley Community CollegePresented nt the American Fduct. ion Research AvociationNew 07-1eansFebruary 25-!4 reh 1, 1973 uctionDaring the past few years, there has leeen a gced :sealof criticism leveled at, compensatory prekilelergarten pro-grams that attempt to raise

3 IQ (e.g. Zieeler).Much ofthis eriticism
IQ (e.g. Zieeler).Much ofthis eriticism suggests that these programs merely acceleratethe normal developmental process, and that children not in,this type of prekindergarten program quickly catch up.Mixis being popularly called the "fade-out phenomenon."Karnes(1968) reports that dramatic increases in academic programshad faded by the end of first grade.Another contention is that there is little evidence tonuppert the transfer of training for such mental processesas measured by IQ.This would then suggest that reportedgains in IQ by these subjects is nct correlated with laterachievement as measured by a standardized achievement test.This does not mean to imply that gains reported by manyHead Start programs were accurately reported.This actuallyis a mute point if there is no correlation between the twoin the first place.The purpose of, this study is to invesigate the re-lationship between gain (or change) scores on standardizedintelligence tests, during prekindergarten compensatoryed4cation programs, and future performance on standardizedachievement tests.This study le imoortant for at least two reasons.First, compensatory programs which use IQ

4 gain as their criteria for auaceS3are -
gain as their criteria for auaceS3are -prevalent throughout theliterature(DiLorenzo, Waikard, Gray and Klaus, Beiiter ara Ehgelmann,Miller, etc.),Secondly, the methodology used for determining thegall-%scores ersed is one which the authorsfind seldom implementedin the literature.It takes into account regression whichhas often been a major contributor to the gainshown by sub-jects in compensatory programs (Jensen).-Review of the Literatureis well known, many studies have -Leen conductodwhichshow a high correlation between IQ andAchievement (Lennon;Woodrow; Manolakes; George and Sheldon;Birch,L.G.).This ofcourse is an important relationshipwhich has been acceptedby many coordinators of compensatory education programs(Di Lorenzo, Weikart, Gray and Klaus, Bereiter. and Engelmarn,Alpern, Karnes, Kohlberg, Phillips, and Reidore).One ofthe objectives of their programs was to raise IQ.It wasfelt if this could be accomplished then Vas IQgain wouldbe positively correlated with a gain in laterachievement.Much criticism has been leveled at thiscontention, forhas not been empirically demonstrated.SubiectsThe sample for this study was drawn fromtive geoP

5 ;raphloareao: Nlw York State in which147
;raphloareao: Nlw York State in which1476 children participated in a nrekinderarten evalAation*duriag1965-66 or 1966-47(Table 1).Aporoxim.stel., 84% of these children, weredisadvanta-t3I.The Crier criterion for the idcnt;_fi-cationdiscONantaged and nondisadvantag,ed childrenwas the flther's occupationalrating on the Warner Scale.When there was no father in the home, themothertr occu-pation or the general eccaomic statue ofthe family wasthe index 1171r.Children were screened by School dis-trict personnel, prc tested with theindividual Stanfor.2-Binet Intelligence Scale and the PeabodyPicture. IfocaOu-lary Test, and randomly assigned toexperimental andcontrol groups in each district(DiLorenzo, i968).Overthree years later, the students *vhoremained in the statewere given an achievementtest, the New York Sta..PupilEvaluation Program (PEE').Reading scores on the 711) werethe comparison criteria forS-B gains.Table IPoptaatian___Subjects720756ThatS-9 (Pre-Test)S-B (Post-Test)VPS-B (Pre-Test)S-B (Post-Test)PER_______DateAdministeroct19b5196696-29 6.1967The sample consisted of the405 subjects on which itwas Possi.ble to findrecorded scores on all thr

6 ee tests.1411 of the subjects were contr
ee tests.1411 of the subjects were controls inthe original study. MethodolcqvThroe- of the major critic isms directed at the evaluationof compensatory education programs have been (1) the failureto inclpde a centre' eroup, (2) the failure to take intoacc:unt regression when colripItine- gain scores, and ()) thelack et' evidence showing a substantial relationship betweengain and later achievement.The orooedures emtloyed attack each of the abo.e criticieme.First, the original sample was randomly selected from a gener-al population of disadvantaged children, rather than the par-tect:,ento 'being selected on the basis of extreme scores. -Acontrol group, which did not attend the prekindergarten pro-gram, va3 randomly selected from the eamnle. (Diorenzo, 1969)The second criticism, involving the failure to accountfor regression in the computation of gain scores, has beennut forward by many, including Jensen (1969) and several ofthe ccntributors to the Herris (1963) text,The procedurefollowed in this study has taken into account regressionthrough the use of the following technique.(1)A coeffi-cient of stability was obtained on the control group for

7 theinstrument on which the gain scores w
theinstrument on which the gain scores were to be computed; inthis case the Stanford-Binet Intelligence Test (S-B).Theobtained coefficient of stability was used ar an estimatefor the entire%samole.(2)The sample mean for the pe-test was then computed and each score subtracted from thismean.())This difference score was then multiplied by thecoefficient of stability.(4)The product that was calcu- lated %Nas added to T.he pre-test score to produce anexpectedpr)z;t-tst score with recrression taken into aecoult.(5) Theexpecpost-test score was then subtracted frrim theactualpost-teFA score with the result being thecorrected gainscore.(6)Finally, the corrected gain scores were corre-lated with an independent measure; forthis study, readingscores on the New York StatePupil Evaluation Program.The third criticism, concerning the lackof empiricalevidence showing a relationship between gain scoresandlater achievement, was attacked by comparingthe correla-tion between absolute S-B gain(post-test minus pre-test)and later PEP scores with the correlationbetween correctedS-B aain and later PEP scores.ResultsThe coefficient of Stability for the preand post S-B

8 was .698 for the 144 controls.The result
was .698 for the 144 controls.The results of the correlations betweenabsolute S-.6gain with PEP and corrected S-B gainwith PEP are shown inTable TT.Table II7--1Correlations betweenand PEPTyre ot SGorrela'tionNumberAbsolute gain0.12_405Corrected rain0.42405iIt was also felt by the authors thatit would beworthwhile looking at correlations within aseries of smallel.score ranges.The subjects were placed in cellsbased on their rre :'tar ford-Binet score.it w-ls decided to make eachcell_ repreL;ent one half a standprd deviation of the norma-2,pooulrioniEach cell would then represent a rather homo-carrolo, the difference which would the:: be high-lic-hted would be the rain which cachacell would show.(Table III)Table IIIMatrix of Subdivided IQ Range Showing Number ofS's within each cellI0 RANGE[u9JFCTSI 68725176-83 04-9192-29_,_100-107_, 108-1152ToU41 falT:11A-JECTSi296oI77125157110358" 1.....The above cells were also subdivided into experimeata't-end control.All of the cells showed a positive correlationbetween amount of Fain and IQ except the lowest extreme,which is significantly different at the .0lz_,-;,21 from allthe other fTrouns,The entire br

9 eakdown can bi,-2en inTable IV.1Any cell
eakdown can bi,-2en inTable IV.1Any cell containing less than 10 S's was not consideredmeaninrful and, therefore, a correlation was not computed. -7--Table IVMatrix of Correlations Between S's Pre-test Stanford-Binetand Pupil Evaluation Program Score Subdivided by IQ Range,Experimental, Control, and Overall..1111 nas___1615:75__Experimental76-83 184 -9192-99.29N=87:24N=38.28=125_1100-10..30N=.2N=22.39N=571108-1.0.N= 1.02N=21.17N=10.22N=15-.34Isk-14.49.18N=39N=49:a.44N=21114=28.5.2N=60iN=77INControlOverall-.08N=22ConclusionsIt is evident that the relationship between gains onstandardized intelligence tests and later achievement testsis not apparent from an examination of the raw data.Theeffects of regression on any change score must be taken intoaccount before any meaningful statements can be made concern-ing these scores.Once regression has been corrected withinthe gain scores, vastly different correlations appear betweenthese gains and later achievement (0.42 vs. 0.13).If suchcorrelations are consistently found to exist between gainscore and later achievement, the argument that prekindergar-ten programs attempting to raise IQ's are not educ

10 ationallysound, is not a valid one.The s
ationallysound, is not a valid one.The substantial correlationsbetween gain scores and achievement give support for cogni-tive programs aimed at raising IQ's.The importance of this study-is not as a defense foreducational programs designed to raise a child's IQ, butrather the fact that a correlation does exist between gainin IQ and later achievement test scores.For too long -8-educators have lived with the misrepresented knowledge thatno such correintion exists.Additional studies, uqinrt different achievement testmeasures and different intelligence tests on differentpopulations, need to be undertaken.This study shows thatsuch relationships do exist; the generalizability of theserelationships is yet to be shown. BIBLIOGRAPHY1.Alpern, Gerald D."The failure of anursery schoolenrichment program for culturally disadvantagedchildren." Paper read at annualmeeting ofAmerican OrthopsyChiatric Association,San Francisco.1966.2.Bereiter, Carl, & Engelman, Siegfried.Teachingdisadvantaged children in the preschool.New Jersey:Engelwood Cliffs.Prentice-Hall.1966.3.Birch, L. G."The Improvement of Reading Ability,"jritish Journal of Educational Psychology,XXP

11 art 2 (June, 1950), 73-76.4.Diederick, P
art 2 (June, 1950), 73-76.4.Diederick, Paul."Pitfalls in the Measurement ofGains in Achievement." The School Review.February, 1956.59-64.5.DiLorenzo, Louis T."Prekindergarten Programs forEducationally Disadvantaged Children."New YorkState Education Department.Albany, New York..December, 1969.6.Gray, Susan W. & Klaus, Rupert A.Early trainingprojects:interim report, Nevember 1963.Murfreesboro,City Schools and George Peabody CollegeforTeachers.Nashville, Tenn.7..The early training project for disadvantagedchildren;a rsport after five years.GeorgePeabody College for Teachers.Nashville, Tenn.September 1967.8.."An experimental preschoolprogram for culturally-deprived children." Project Report.GeorgePeabody College.Nashville, Tenn.1965.9.Harris, Cheater.(ed.)"Problems in Measuring Change."University of Wisconsin Press.1963.10.Jensen, Arthur R."How much can we boost IQ and AcademicAchievement." Harvard Education Review.79:1-123.Winter, 1969.11.Karnes, Merle B."A longitudinal study of disadvantagedchildren who participated in three differentpreschoolprograms."Institute for Research on ExceptionalChildren.. Univ. df Illinois, Urbana,Unpublished report