The Effects of Full Day Versus Half Day Kindergarten: Review and Analysis of National and

Indiana Data

Jonathan A. Plucker, Ph.D. Jessica J. Eaton Kelly E. Rapp Woong Lim

Jeffrey Nowak, Ph.D. John A. Hansen Amy Bartleson

Center for Evaluation and Education Policy (formerly the Indiana Education Policy Center)

January 9, 2004

Prepared for the Indiana Association of Public School Superintendents Information and Research Commission One North Capitol Avenue, Suite 1215

Indianapolis, IN 46204

INDIANA ASSOCIATIONOF

PUBLIC SCHOOL SUPERINTENDENTS

INFORMATION AND RESEARCH COMMISSION MEMBERS

Superintendent Jeff H. Abbott Superintendent Thomas J. LittleEast Allen County Schools Kokomo- Center Township Cons. Sch. Corp.New Haven, Indiana Kokomo, Indiana

Superintendent Neyland Clark Superintendent Vickie MarkavitchSouth Harrison Comm. School Corporation Penn-Harris-Madison School CorporationCorydon, Indiana Mishawaka, Indiana

Superintendent Marlin B. Creasy Superintendent Stephen J. PatzMuncie Community Schools Rising Sun-Ohio Co. Comm. School Corp.Muncie, Indiana Rising Sun, Indiana

Superintendent R. Stephen Gookins Superintendent Donald R. YeomanDelaware Community School Corporation Tri-Creek School CorporationMuncie, Indiana Lowell, Indiana

Superintendent William KirbyHuntington County Comm. Sch. Corp.Huntington, Indiana

FOR FURTHER INFORMATION CONTACT:

Dr. Roger W. Thornton, Executive DirectorIndiana Association of Public School Superintendents

ONE NORTH CAPITOL, SUITE 1215Indianapolis, IN 46204

317-639-0336

INDIANA ASSOCIATION OF PUBLIC SCHOOL SUPERINTENDENTSOne North Capitol, Suite 1215 - Indianapolis, Indiana 46204

2003-2004 Roster of IAPSS Officers and Committee Members

EXECUTIVE COMMITTEE

DONALD R. YEOMAN PRESIDENT

Tri-Creek School Corporation

THOMAS E. MCKAIG PAST-PRESIDENT

Peru Community School Corporation

WILLIAM J. CARNES PRESIDENT-ELECT

Whitley County Consolidated Schools

JIM FREELAND VICE-PRESIDENT

Batesville Community School Corporation

TERESA A. EINEMAN SECRETARY

Fayette County School Corporation

STEVE WITTENAUER TREASURER

Benton Community School Corporation

PHYLLIS AMICK LEGISLATIVE COMMITTEE CHAIRPERSON

Richmond Community Schools

DONALD STINSON POLICIES AND RESOLUTIONS

M.S.D. of Decatur Township COMMITTEE CHAIRPERSON

ALLEN BOURFF NORTHWEST DISTRICT I CHAIR

Knox Community School Corporation

ALAN MIDDLETON NORTHEAST DISTRICT II CHAIR

Garrett-Keyser-Butler Community School District

JOHN E. WILLIAMS NORTH CENTRAL DISTRICT III CHAIR

Delphi Community School Corporation

EARL WILLIAMS WEST CENTRAL DISTRICT IV CHAIR

South Putnam Community School Corporation

JOHN G. ELLIS CENTRAL DISTRICT V CHAIR

Noblesville Community Schools

MARLIN CREASY EAST CENTRAL DISTRICT VI CHAIR

Muncie Community Schools

DOUGLAS ROSE SOUTHWEST DISTRICT VII CHAIR

Vincennes Community Schools

THOMAS BOOK SOUTHEAST DISTRICT VIII CHAIR

South Dearborn Community Sch. Corp.

i

The Indiana Association of Public School Superintendents recently contracted withthe Center for Evaluation and Education Policy (formerly the Indiana Education PolicyCenter) to conduct a review of research on full day kindergarten. The goal of the report isto provide useful information to Indiana policymakers as they debate the merits of full versushalf day programs.

This report sought to answer three questions: What does the national research sayabout the effectiveness of full day kindergarten? What does the Indiana data say about fullday kindergarten? And how is time used within full day kindergarten programs? Finally, thereport concludes with a series of recommendations regarding Indiana policy on full daykindergarten.

What Does the National Research Say About the Effectiveness ofFull Day Kindergarten?

Center staff conducted an exhaustive review of the literature regarding the effects offull day kindergarten programs. Primary sources were acquired from the on-line databasesERIC and PsycInfo. Additionally, lead authors of full day kindergarten projects werecontacted. Only those reports that directly compare the experiences of studentsparticipating in full versus half day programs were included in the review. Research onalternate day programs, in which students attend an entire day of kindergarten on alternatingdays, is not included due to the lack of evidence supporting the effectiveness of theseprograms. Evidence was gathered supporting the following areas: attendance; academicachievement, including grade retention and special education referral; social and behavioraleffects; and effects on disadvantaged students.

On the issue of attendance, findings are inconclusive. The studies involving academicachievement, grade retention, special education referrals, and social and behavioral effectsgenerally support the effectiveness of full day over half day programs. Disadvantagedstudents in full day kindergarten were also found to experience greater academic benefitsthan students in half day programs, although the magnitude of these greater benefits is againinconclusive. The national research suggests that there are no negative outcomes commonlyassociated with full day kindergarten.

What Does the Indiana Data Say About the Effectiveness of Full DayKindergarten?

Researchers have conducted several evaluations of full day kindergarten in Indiana.These studies followed students in the following Indiana school districts: Evansville-Vanderburgh School Corporation, Lawrence Township, Perry Township, MuncieCommunity Schools, and Indianapolis Public Schools. Each set of studies is reviewed in thissection. Data were also collected from and analyzed from two additional Indiana districts: alarge, urban district and a rural district.

The Effects of Full Day Versus Half Day Kindergarten:Review and Analysis of National and Indiana Data

Executive Summary

ii

Several criteria were used to identify research for this section of the report: First,data had to be available for full day kindergarten students and a control group of students(usually half day students in the same district or school). Second, full day programs neededto be every day programs, not full day–alternate day programs. Third, extended dayprograms could not be included unless substantive instruction occurred during the extendedday part of the program (i.e., the program could not be half day kindergarten plus afternoonchild care).

Results from the eight Indiana data sets reflect the results of the national research onfull versus half day kindergarten. As was the case with the national data, the Indianaresearch suggests that there are no negative outcomes commonly associated with full daykindergarten, and that – at worst – full day kindergarten and half day kindergarten havesimilar effects. Significant results in support of the benefits of full day over half daykindergarten were found in many of the comparisons within these studies. When analyzedon the major dimensions of academic achievement, grade level retention, special educationreferrals, and social and behavioral effects, the benefits of full day kindergarten programsare apparent.

How is Time Used in Full Day Programs?

To many critics, full day programs have the potential to be nothing more than halfday kindergarten with an extra half day of play time. We identified only two recent studiesthat provide detailed analyses of how time is spent in full and half day programs: onenational study and one study of students in Wisconsin. Given this paucity of research, weconducted a two phase study of instructional activities in Indiana kindergarten classrooms:In phase one, we collected full and half day kindergarten schedules from Indiana schoolsand compared the scheduled activities. In phase two, we conducted several site visits to fullday programs to establish the validity of submitted schedules.

The research literature and data collected for this report provide evidence that timein full day kindergarten programs is different both quantitatively and qualitatively from howtime is used in half day programs. Across all of the schools in the Indiana sample, theproportion of instructional time is similar across program types, resulting in much greaterinstructional time in full day programs, representing approximately 40-50% more instructionin full day programs than half day programs. The Wisconsin study and the Indiana site visitssuggest that in individual classrooms, the additional time leads to greater use of child-initiated activities. In the site visit schools, these activities were almost universallyinstructional in nature and did not involve play. The published research also providesconvincing evidence that certain types of reading skills and grouping strategies are moreprevalent in full day programs, including reading aloud, peer tutoring, and mixed-abilitygrouping. During the Indiana site visits, researchers saw evidence of these activities, but halfday programs were not observed and therefore comparisons cannot be made acrossprogram types.

iii

Table of Contents

Executive Summary i

Introduction1

Section I: Review of Research on Full Day Kindergarten Effects 3Attendance 3Academic Achievement 3Grade Retention and Special Education Referrals 5Social and Behavioral Effects 5Disadvantaged Students 6Summary 7

Section II: Data from Indiana Districts on Full Day Kindergarten Effects 9Introduction 9Evansville Study 10Lawrence Township Study 14Perry Township Report 16Muncie Community Schools Study 17Indianapolis Public Schools Study 17Analysis of Data from a Large, Urban School District 19Analysis of Data from a School in a Rural District 21Summary

Section III: Analysis of Daily Activities in Full Versus Half DayKindergarten Programs 23Introduction 23Published Studies on How Time is Spent in Full and Half Day Classrooms 23Curriculum/Schedule Analysis 24Site Visits 26Summary 27

Conclusions and Recommendations 29

References 31

Appendix A: Summary of Select Kindergarten Policies for All 50 States 35

iv

1

School districts, both in Indiana and nationally, have begun to focus attention on the impor-tance of early childhood education. These efforts are based on the belief that preparing allstudents to be “ready to learn” by first grade will help shrink the achievement gap betweensubgroups of students. A common strategy among these early childhood initiatives is toextend traditional half day kindergarten programs to full day programs.

A number of perceived benefits are associated with full day programs, including betterpreparation for elementary school, especially for students at risk; continuation of preschoolprograms, many of which involve full day experience; greater access to support services forstudents with special needs; simplified child care and transportation responsibilities forparents; and reduced midday transportation responsibilities for schools. At the same time,critics of full day programs suggest several limitations, including cost of these programsrelated to need for increased space and instructional staff; full day programs becoming aproxy for childcare; young children’s ability to adapt and cope with full day experiences; anduncertainty about the long-term benefits of such programs.

As Indiana policymakers debate the merits of full versus half day programs, a review ofpertinent research will provide useful information. For this reason, the Indiana Association ofPublic School Superintendents contracted with the Indiana Education Policy Center toconduct a review of research on full day kindergarten. This report has three sections: InSection I, we review research conducted nationally and in other states on full day kindergar-ten. Section II reviews published and unpublished analyses of data from Indiana students;this section also contains data from three districts that were analyzed specifically for thisreport. Section III provides analysis of half and full day kindergarten curricula and sched-ules. Finally, the report concludes with a series of recommendations regarding state policyon kindergarten activities for the state of Indiana.

Participation in Full Day Kindergarten

State policies vary widely with respect to full day kindergarten (see Appendix A). As of thebeginning of the 2003-2004 school year, 40 states required public school districts to offerkindergarten programs, of which 10 required districts to offer full day programs.1 However,participation in kindergarten is mandatory in only 14 states, with full day participationmandatory in only 2 of those states (Louisiana and West Virginia). Few states mandate thatdistricts offer half day programs if full day programs are being offered.

According to the U.S. Department of Education’s Early Childhood Longitudinal Survey(ECLS; Walston & West, 2002), enrollment in full day kindergarten also varies widelyacross states, communities, and schools. Nationwide, enrollment ranges from 83% in thesouthern states to 23% in the western states.2 Demographic variables impacting enrollmentin full day kindergarten include: location of the child’s home, race, ethnicity, poverty, andtype of school, i.e., public or private (Denton, West, & Walston, 2003; Walston & West,2002). Students from rural and urban districts are more likely to attend full day kindergartenprograms than their suburban counterparts. Ethnic groups participating in full day

Introduction

2

1 Alabama, Arkansas, Georgia, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, SouthCarolina, West Virginia (ECS, 2003)2 Northeast 41%, Midwest 45%

kindergarten are: African American (79%), white (49%), Hispanic (46%), Asian (40%).Attendance in full day kindergarten is greater for children of poverty level families(62%) than for children of families with non-poverty level status (51%). Students fromnon-English speaking homes attend full day programs at a lower rate than children fromEnglish speaking homes (45% vs. 55%). Of schools studied in ECLS, 70% of privateinstitutions offer full day programs compared with 54% of public schools.

Currently, only 12% of Indiana kindergarten students attend full day programs (LAS,1999). This low participation rate is primarily due to the very limited funding for full daykindergarten programs provided by the state (i.e., $8.5 million for the 2003-2004school year vs. roughly $100 million estimated to be needed to offer universal full daykindergarten for the 2001 fiscal year; Goodpaster, 1999). However, several Indianaschool districts have created these programs with a combination of local, Title I, andother state and federal funds. As a result, access to full day kindergarten has beenuneven across the state. To date, a statewide evaluation of the funding mechanisms forand effectiveness of these programs has yet to be completed.

3

Section I: Review of Research on Full Day Kindergarten Effects

Indiana Education Policy Center (IEPC)staff conducted an exhaustive review ofresearch literature regarding the effectsof full day kindergarten programs.Primary sources were acquired from theon-line databases ERIC and PsycInfo.Additionally, lead authors of full daykindergarten projects were contacted.Only those reports that directlycompare the experiences of studentsparticipating in full versus half dayprograms were included in the review.Research on alternate day programs, inwhich students attend an entire day ofkindergarten on alternating days, is notincluded due to the lack of evidencesupporting the effectiveness of theseprograms.1 Evidence was gatheredsupporting the following areas:attendance; academic achievement,including grade retention and specialeducation referral; social and behavioraleffects; and effects on disadvantagedstudents.

Attendance

Findings relative to kindergartenprogram type and attendance aremixed. Goodwin (1989) found greaterdaily attendance in full day kindergartenprograms than in half day programs.However, in the Madison MetropolitanStudy (1985) and a study by Humphrey(1980), no difference was reported, andEvans and Marken (1983) found thatfull day kindergarteners missed anaverage of 3.7 more days of school thanstudents enrolled in half day programs.The mixed results are difficult tointerpret: Full day students may needmore absences for doctor’sappointments and other necessary out-of-school activities, all of which may

have occurred outside of kindergartenduring half day programs. In addition, thesample of students probably hasconsiderable influence on attendanceresults.

Academic Achievement

Academic achievement outcome dataassociated with full and half daykindergartens have been reported in threetypes of studies: meta-analyses; large,national-scale studies; and programevaluations of specific full and half dayprograms. Findings from meta-analysis andlarge-scale studies appear to support fullday kindergarten with respect to short- andlong-term academic achievement. Forexample, in a meta-analysis, McConnelland Tesch (1986) compared the findingsfrom nine studies regarding full and half daykindergarten programs. These studiesbased their conclusions on analysis using 64test instruments. The aggregated data revealthat 40 of the 64 comparisons (63%) favorfull day kindergarten with respect toacademic achievement gains with nocomparisons favoring half day programs.Further, Fusaro’s (1997) meta-analysis ofachievement test results from 21 studiesfound a large effect size of .77 favoring fullday kindergarten, explaining roughly 60%of the variance in the achievement testoutcomes.

In an ongoing, national study performed aspart of the Early Childhood LongitudinalStudy conducted by the U.S. Departmentof Education’s National Center for Educa-tion Statistics (NCES), NCES staff arefollowing a nationally representative sampleof children from kindergarten through thefifth grade, beginning with the kindergartenclass of 1998-1999. Initial analyses of fall-

4

spring assessment results from thekindergarten year found the readinggains of full day kindergarten studentsto be 0.12 standard deviations abovethose of students in half day programsafter adjusting for external variables(Walston, West, & Rathbun, 2002).Results on the mathematics assess-ments suggest a similar benefit in favorof full day programs, again after adjust-ing for other child, family, and class-room characteristics.

The related literature with respect tosmall-scale research or programevaluations, in general, points tofindings that support full daykindergarten as a contributing factor ofgreater academic achievement (e.g.,Coladarci & Ervin, 2000; Cryan et al.,1992; Elicker & Mathur, 1997; Hills,1985; Hough & Bryde, 1996;Koopmans, 1991; Lore, 1992). Forexample, according to an evaluation inPasco School District 1 in Washington(1987), full day kindergarten studentsperformed statistically significantlyhigher than half day students onnationally standardized tests measuringskills in reading, spelling, andhandwriting, as well as on testsmeasuring individual skills (e.g., color,shape, numbers, letters, quantities). Thefull day kindergarten students alsoshowed statistically significant meangains on the English, mathematics, andhandwriting subtests of thesestandardized achievement tests. Similarresults have been found in studies onother achievement outcomes, includingprerequisite reading skills (da Costa &Bell, 2000) and oral languageassessments (Wang & Johnstone,1999). The benefits of full dayprograms appear to extend beyond theend of kindergarten, with evidence thatfull day students have higher academic

achievement in third (Mueller, 1977)and eighth grades (Nieman & Gastright,1981a; Pasco School District, 1987) inboth reading and mathematics.

Within our review, some small-scalestudies, however, did not findstatistically significant differences inacademic achievement based onkindergarten program type (Cryan etal., 1992; Holmes & McConnell, 1990;Johnson, 1974; McClinton & Topping,1981; Stofflet, 1998). In one example,the scores of the Gates-MacGinitieReading Test administered bySergesketter and Gilman (1988)indicated no difference in readingachievement between kindergartenprograms. Subsequent re-analysishowever, suggests that full daykindergarteners did in fact havesignificantly higher scores (Fusaro,1997). These results may be due to thesmall sample size of the studies whichreduces the power of the significancetests, the lack of appropriate controlgroups, or the lack of pre-interventiondata, all of which have been identifiedas weaknesses in early childhoodresearch (Coladarci & Ervin, 2000;Fusaro, 1997; Puleo, 1988). Whenthese threats to internal validity areaddressed, small scale studies generallysupport full day relative to half dayprograms.

5

Table 1. Eighth grade follow-up on special education referrals and grade retention

Fullday

Halfday

Special EducationReferral

5% 11%

Grade Retention 9% 12%

Grade Retention and Special Educa-tion Referrals

Grade retention and special educationreferrals are indirect indicators ofstudent achievement, and the impact offull versus half day programs on thesefactors is addressed in several studies.Cryan et al. (1992) uncovered evidencethat full day programs resulted in 17%-55% fewer grade retentions but foundno relationship between program typeand special education provisions. In afollow-up of the Anchorage SchoolDistrict full day kindergarten study,Stofflet (1998) found that first graderetention was less likely for studentswho had attended full day kindergarten.Other research has been mixed withregard to findings of change in retentionrate or referral rate. In a longitudinalstudy, Evans and Marken (1983) foundno relationship between kindergartenprogram type and “the number ofchildren placed in categoricalprograms.”

An eight-year longitudinal studyconducted by Nieman and Gastright(1981a, 1981b) favors full daykindergarten. The researchers foundthat both special education referrals andgrade retention levels were greater withrespect to students that had attendedthe half day kindergarten program. A

summary of these results is presented inTable 1.

Social and Behavioral Effects

Several research studies investigatingpro-social and behavioral developmentfavor full day over half day programs(Cryan et al., 1992; Humphrey, 1980;Wang & Johnstone, 1999). Cryan et al.(1992) investigated facets such asoriginality, individual learning,involvement in classroom activities,productivity with peers, intellectualdependency, failure, anxiety,withdrawal, blaming, and approach toteachers. They found that full dayprograms favored the development ofthese pro-social and behavior attributeswith the exception of blaming. Full dayprograms have also been shown tofoster greater independence (Puleo,1988) as well as a greater degree ofactive engagement (Elicker & Mathur,1997). In a study by Hoffman andDaniels (1986), data supported half dayprograms regarding personal and socialdevelopment. No significantly differentgains in maturity level were found instudies by Jones, Pollock, andMarockie (1988) or Puleo (1988).

A major concern related to full daykindergarten is whether young studentscan handle a full day of instructionalactivity. Opinions are mixed on this

6

issue, but the research generallysupports the conclusion thatkindergarten students adjust to thelonger days without major difficulties(e.g., Hough, 1996; Koopmans, 1991;McConnell & Tesch, 1986).

Disadvantaged Students

Several studies indicate that full daykindergarten has the greatest effect onat-risk children and children fromeducationally disadvantaged homes(Clark, 2001; Clark & Kirk, 2000; daCosta & Bell, 2000; Fusaro, 1997;Jones, Pollock, & Marockie, 1988;Karweit, 1992; Koopmans, 1991;Ohio State Legislative Office ofEducation Oversight, 1997; PascoSchool District 1, 1987; Puleo, 1988;Rothenberg, 1984). Students from lowsocioeconomic backgrounds alsobenefit considerably from full dayprograms (Jones, Pollock, &Marockie, 1988). Further, students ator below the poverty level enrolled infull day kindergarten scored statisticallysignificantly higher in math and readingthan their half day counterparts. Poorchildren enrolled in full daykindergarten programs testedstatistically significantly above half daypupils on reading, spatial and verbalskills, naming colors and letters, andidentifying numerals (Pasco SchoolDistrict 1, 1987). Out of all groupsstudied, students living at the povertylevel who were enrolled in full daykindergarten programs also had thegreatest improvement in Englishvocabulary.

Perhaps the most convincing evidencecomes from a recent analysis of theEarly Childhood Longitudinal Study bythe U.S. Department of Education(Walston et al., 2002). As noted

above, the ECLS data provideevidence that full day kindergartenprograms are associated with greaterreading achievement gains during thekindergarten year than half dayprograms. Similar benefits wereobserved when student race andpoverty status were examined, althoughthe results are complex. For example,full day versus half day differences arenot consistent for children fromhouseholds below the poverty thresholdas compared to those from householdsat or above the poverty line. The largergain in math scores for full daycompared to half day kindergartners ismore pronounced for children at orabove the poverty threshold (8.7 vs.7.3; effect size = 0.28) compared tochildren living in households below thepoverty threshold (7.7 vs. 7.2; effectsize = 0.10). Interestingly, the presenceof an aide is associated with greaterreading gains for Black children in fullday kindergartens (9.5 mean gain withan aide vs. 7.7 mean gain without anaide; SD = 6.0, effect size = 0.30) andfor Hispanic children in full day dayprograms (11.6 with an aide vs. 10.1without an aide; SD = 6.27, effect size= 0.24). For these minority children,gains associated with full daykindergarten and the presence of anaide are absent in the correspondinghalf day programs. However, thispattern is not observed for mathperformance. The authors conclude that“Providing full day kindergarten haslong been considered an effectiveapproach for improving minority readingachievement; this finding suggests thatthis approach, coupled with thepresence of an instructional classroomaide, may improve the achievement ofminority children in kindergarten”(Walston et al., 2002, p. 18).

7

Summary

Many studies have compared theeffects of full day kindergarten to thoseof half day kindergarten on a number ofdimensions. In this review of the existingresearch, project staff analyzed theresearch on several categories ofstudent outcomes. On the issue ofattendance, findings are inconclusive.The studies involving academicachievement, including grade retentionand special education referrals,generally support the effectiveness offull day over half day programs.However, the magnitude of this positiveeffect varies considerably from study tostudy, with many researchers reportinglarge effects, many others reportingsmall effects, and a minority reportingnegligible effects. Findings on social andbehavioral effects are mostly in favor offull day programs. Disadvantagedstudents in full day kindergarten werealso found to experience greateracademic benefits than students in halfday programs, although the magnitudeof these greater benefits is againvariable. The national research suggeststhat there are no negative outcomescommonly associated with full daykindergarten.

1 Research consistently favors every day fullday kindergarten over alternate day full daykindergarten programs with respect tostudent achievement (e.g., Elicker, 2000;Pasco School District, 1987). Fusaro (1997)found that full day kindergarten was superiorto both half day and alternate day programs,and half day kindergarten was better thanalternate-day kindergarten regarding studentachievement, while McConnel and Tesch(1986) found no significant differencesbetween alternate and half day programs,which is not surprising given the twoprograms usually involve similar amounts ofinstructional time.

8

9

Section II: Data from Indiana Districts on Full Day KindergartenEffects

Researchers have conducted severalevaluations of full day kindergarten inIndiana. These studies followed studentsin the following Indiana school districts:Evansville-Vanderburgh SchoolCorporation, Lawrence Township,Perry Township, Muncie CommunitySchools, and Indianapolis PublicSchools. Each set of studies is reviewedin this section. The Evansville study is byfar the most exhaustive and,consequently, is discussed in the mostdetail. Data were also collected fromand analyzed from two additionalIndiana districts: a large, urban districtand a rural district.

Several criteria were used to identifyresearch for this section of the report:First, data had to be available for fullday kindergarten students and a controlgroup of students (usually half daystudents in the same district or school).Second, full day programs needed to beevery day programs, not full day –alternate day programs. Third, extendedday programs could not be includedunless substantive instruction occurredduring the extended day part of theprogram (i.e., the program could not behalf day kindergarten plus afternoonchild care). Table 2 summarizes theresults of the Indiana studies.

Table 2. Summary of the Results of the Indiana StudiesSchool District Summary

Evansville Results from matched comparisons of full to half day students on testscores, report cards, and questionnaires were in favor of full daykindergarten on every dimension (e.g. academics, social behavior)studied.

LawrenceTownship

Gain scores on tests of Letter Identification and Concepts About Print werenot statistically significantly different for full versus half day kindergartenstudents.

Perry Township Full day kindergarten students in this district are considered “at risk”, butthe program seems to be effective due to the large percentage (83%) ofstudents scoring in the normal range on the Conners’ Teacher Rating Scale-Revised: Long Version.

MuncieCommunity

Students enrolled in full day kindergarten experience an increase in scoreson the TerraNova and ISTEP+ tests by the end of first grade, but the scoreswere shown to decrease over time in this study that spanned from 1997 to2002.

Indianapolis PublicSchools

Full Day Lottery Kindergarten and Magnet Full Day Kindergartenprograms had the highest achievement on the Signposts Early LiteracyBattery in 2002, followed by Half Day Kindergartens and then Title I FullDay programs.

Large, Urban ISTEP+ mathematics and language arts scores were analyzed based onkindergarten program type as well as race and free lunch status, but therewere no statistically significant findings to indicate a greater effect of fullday programs.

Rural Language arts passing rates on the 2002 and 2003 ISTEP+ were notstatistically significantly different for full versus half day kindergartenstudents, but a moderate difference was found for mathematics scores,regardless of inclusion of special education students in the sample.

10

The Evansville-Vanderburgh Longi-tudinal Study

An extensive study comparing full dayand half day kindergarten programswas conducted by the Evansville-Vanderburgh School Corporation. Thislongitudinal study began in 1978 withfour schools initiating full daykindergarten programs in Evansville,Indiana. Matched comparisons wereconducted with control groups fromfour district schools offering half dayprograms. The first cohort of studentsattended kindergarten in the 1978-79school year, with a second cohortfollowing in 1979-80. Matching wasbased on SES backgrounds at theschool-level with a participant-levelrandom sample populating the controlgroup of half day kindergarten students.The purpose of this study was toexamine possible long-term benefits offull day kindergarten participation. Datacollected consisted of standardizedtests such as the Boehm Test of BasicConcepts and the CaliforniaAchievement Tests; report cards;school records; teacher, parent, andstudent questionnaires; and interviews.

Attendance

The Evansville researchers examinedattendance patterns during the partici-pants’ kindergarten years and laterduring their middle school years. Forkindergarteners in full and half dayprograms, researchers found absenteerates of 8.5% and 10.8% respectively.For participants followed through sixth,seventh, and eighth grade, the meandifference between days absent was notstatistically significant.

Academic Achievement

Academic achievement was one of themost salient aspects of the Evansville-Vanderburgh study. Data addressingscholastic progress as well as earlycognitive development were collectedfrom scores on several tests, reportcard grades, and the TeacherOpinionnaire.

The Boehm Test of Basic Concepts wasadministered to the full day kindergartengroup before and after the 1979-1980school year in order to gauge academicachievement. Scores from the full daygroup were 15% higher than averagemid-level socioeconomic norm. Theseresults suggest that the full daykindergarten participants had greateracademic progress throughout the yearwhen compared with the norm.Additionally, the California AchievementTests produced useful data forexamining achievement differencesbetween full day and half daykindergarteners. All but one of thesubtest scores and the combined scoreswere significantly higher for full daykindergarteners.

In the spring of 1980 the Gates-MacGinitie Reading Tests were given tothe 1978-1979 cohort. Results showedthat full day students had better readingability in first grade than did half dayparticipants. This test was administereda second time two years later to boththe 1978-1979 and 1979-1980cohorts, and results showed that bothfull day groups had significantly higherability in comprehension and vocabularyskills than did students in the half dayprograms.

11

Finally, the Comprehensive Test ofBasic Skills was administered in thespring of 1982 when the 1978-1979cohort was in third grade. Students whohad attended full day kindergartenscored significantly higher than the halfday participants in 10 of the 14 areas.This test was administered again whenthe students were in fifth grade, and afinal time in seventh grade. Both fifth-grade and seventh-grade scores for thefull day kindergarten group were higherthan the half day group in all 14 areas.

The only instrument to favor the half daykindergarten programs was theEvaluation Scale—Cursive, ahandwriting test that was administeredto both the 1978-1979 and 1979-1980cohorts in the fall of 1982. Resultsshowed that the half daykindergarteners scored higher than fullday students.

Data were also collected from reportcards in the first, second, and thirdgrade years for both cohorts. Over 23comparisons, full day participants werefound to have higher percentages ofsatisfactory marks and lowerpercentages of less-than-satisfactorymarks than half day participants.

The Teacher Opinionnaire was given toall of the primary teachers involved inthe study after the cohort students hadfinished primary school. When askedabout student work habits, 60% of theteachers believed that full daykindergarten participants had betterwork habits than half day students.Sixty-four percent of teachers thoughtthat students from the full daykindergarten group functioned moreindependently than the half day group.Further, the majority of teachersresponded that former full day

kindergarten participants had better-developed fine motor skills, gross motorcoordination, and handwriting. Sixty-four percent of teachers also believedthat students from the full day groupwere better at following directions.However, when asked aboutdifferences in academic ability, resultswere mixed as 38% of teachers agreedthere were no apparent differencesbetween groups but 38% believed therewere.

Grade-Level Retention and SpecialEducation Referrals

This study was also concerned withdifferences in grade retention andspecial education referrals between thefull day and half day kindergartengroups. Grade retention rates for allstudents participating in the study werecompared in the summer of 1982. Non-promotions due to special educationplacement were included in the totals.

Half day kindergarten students hadmore grade retentions than full daystudents in both the 1978-1979 and1979-1980 cohorts. Nineteen percentof the former half day kindergartenstudents from the 1978-1979 groupwere not promoted to the next gradelevel at the end of either kindergarten,first, second, or third grade whereasonly 9% of full day students were notpromoted. The half day students in the1979-1980 cohort also had a higherrate of grade retentions, with 17% ofthe half day students being retained andonly 4% of the full day students beingretained.

Comparisons of special educationplacements between the full day andhalf day kindergarten groups were alsoobtained, with results strongly favoring

12

neither full nor half day programs. Agreater percentage of students in the1978-1979 full day group was placedin special education (6% compared to2% of half day participants). Thesepercentages were not statisticallysignificant, however, and thus could beattributed to sample variation. The1979-1980 group comparison showedthat 4% of full day students and 3% ofhalf day students had special educationplacements.

Social and Behavioral Effects

Mixed methods were used to assesspossible differences in social andbehavioral development in relation tokindergarten program type. Instrumentand survey responses as well asconduct marks on report cards wereexamined.

The Survey of School Attitudes wasdistributed in 1982 when the 1978-1979 cohort was in third grade and the1979-1980 cohort was in secondgrade. The findings of this survey arecontradictory. Mean scores showedschool attitudes among the 1978-1979full day group to be higher than those ofhalf day students in all four subjectareas: math, reading, science, andsocial studies. In contrast, students inthe 1979-1980 group who hadattended half day kindergarten hadhigher scores than full day programparticipants. The overall results indicateno significant difference in schoolattitudes between students who wereenrolled in half day kindergarten andthose who attended full day programs.

The Teacher Opinionnaire wascompleted by the 25 primary schoolteachers in first through third gradeswho were involved with the study. The

opinionnaire covered a broad range ofquestions comparing effects of full dayversus half day kindergarteninvolvement. The results revealed thatthe majority of teachers believed fullday kindergarten participation hadbeneficial socialization and behavioraloutcomes. Forty-two percent of theteachers indicated that children whohad been enrolled in full daykindergarten appeared to be better atsocializing with peers than half daykindergarteners. Also, 52% of teachersbelieved the attention spans of studentswho had attended full daykindergarteners were longer than thoseof former half day kindergartenattendants.

The Piers-Harris Children’s Self-Concept Scale was administered in1982 when the 1978-1979 cohort wasin third grade and the 1979-1980cohort was in second grade. Testresults show that for both cohorts,students who had attended full daykindergarten had better self-conceptscores than those who attended halfday programs.

Conduct marks on student report cardswere looked at for first, second, andthird grade students in the 1978-1979cohort, and the first and second gradereport cards were used for the 1979-1980 group. Possible scores in the 12conduct areas (such as “shows self-confidence,” “works well with others,”or “uses self-control”) includedSatisfactory Progress, ImprovementShown, or Needs Improvement.Results of a statistical test indicated thatin first grade, the full day kindergartengroup and half day group had an equalpercentage of Satisfactory marks, butthe full day group had a lowerpercentage of less-than-satisfactory

13

marks. Reports from the second gradeindicated that former full daykindergarten attendants had a greaterpercentage of Satisfactory marks andfewer less-than-satisfactory marks thanthe half day group. Remarks about thethird graders who previously attendedfull day kindergarten were Satisfactorymore often than those of the half daygroup. Results for the 1978-1979cohort are given in Table 3. Overall,students who formerly attended full dayprograms had a higher percentage ofSatisfactory conduct marks andreceived far fewer less-than-satisfactorymarks than their half day counterparts.

The results of the Survey of SchoolAttitudes, the Teacher Opinionnaire, thePiers-Harris Children’s Self-ConceptScale, and the report card conductmarks indicate that participation in fullday kindergarten programs may lead togreater social and behavioral gains thanhalf day participation. Although theSurvey of School Attitudes showed nosignificant difference betweenexperimental and control groups, theother three methods favored full daykindergarten.

Table 3. Second Grade Academic Marks for 1978-1979 Kindergarten Students

Subject Group NSatis-factoryn %

Im-provementShownn %

NeedsIm-provementn %

Chi-Square p

Mathematics

Reading

Hand-writing

Spelling

Full dayHalf day

Full dayHalf day

Full dayHalf day

Full dayHalf day

7697

7697

7697

7697

69 9153 55

67 8860 62

56 7458 60

70 9264 66

3 416 16

4 57 7

5 78 8

2 312 12

4 528 29

5 730 31

15 2031 32

4 521 22

26.84

16.76

3.80

16.67

< .01

< .01

N.S.

< .01

14

Extracurricular Activities

Participation in extracurricular activitieswas also examined for the 1978-1979cohort as they reached grades six,seven, and eight. Former full daykindergarten attendants had higherparticipation rates during sixth andeighth grades than former half daystudents in The Academic Academy, anafter-school program offering variousperformance and scholastic activities. Acomparison of the total athleticparticipation among the sixth, seventh,and eighth graders included in the studyshowed that the full day group hadhigher participation rates than the halfday group. Additionally, students whohad attended full day kindergarten hadhigher participation rates in otheractivities such as band, cheerleading,and student council during all threegrades. Overall, extracurricular activityinvolvement was greater for studentswho had attended full day kindergarten.

Parent Responses

A Parent Questionnaire was mailed toall parents of kindergarteners involvedin the study at the end of theirkindergarten year. The questionnaireaddressed parental concerns aboutlearning and program preference aswell as reasons for program selection.In response to the question of howmuch their children had learned inkindergarten, parents representing bothfull and half day students believed theirchildren had learned a great deal(82%). Parents of full day studentsindicated that they felt their childrenexperienced greater levels of cognitive,psychomotor, affective, and linguisticgrowth than did the parents of half daystudents. If given a choice, 92% of the130 full day kindergarten parents who

returned the questionnaire stated thatthey would choose full day over halfday kindergarten. Fifty-two percent ofthe parents of half day studentsindicated that they would have chosenfull day kindergarten. Free responsecomments from parents were varied,expressing support or criticism for bothtypes of programs.

A second questionnaire was mailed toparents of the full day kindergartengroup when the 1978-1979 cohort wasin fourth grade and the 1979-1980cohort was in third grade. Of the 92respondents, 95% believed that theirchild learned more in full daykindergarten than he or she would havelearned in a half day program. Ninety-five percent also indicated that theirchildren were better prepared for firstgrade because of full day kindergarten.Further, these parents also favored fullday kindergarten over half dayprograms with respect to gains in theirchild’s self-control. Overall, full daykindergarten parents were happy withtheir child’s full day kindergartenparticipation and highly favored full dayprograms over half day kindergarten.

Lawrence Township Study

Renbarger (2003) examined literacygains of kindergarteners enrolled inthree types of programs: full day, halfday, and alternate-day. This studyevaluated pre and post-test scores forLetter Identification and ConceptsAbout Print (CAP). The main objectiveof this study was to determine whetheror not program type was related tosignificant differential improvements inliteracy achievement. The researchersampled 1530 students enrolled in boththe 2000-2001 and 2001-2002 school

15

years who consisted of the followingdemographics: 58% Caucasian, 32%African American, 8% Latino, and 2%Asian. Twenty-seven percent of thesample qualified for free or reducedlunch. The study took place in acentralized kindergarten programconsisting of 24 classrooms in a schooldistrict comprised of 16,000 students.Each classroom was designedspecifically for kindergarten programsand each contained a literacy center.

Parents were offered the option of fullday, half day or full day alternate dayprograms, with the full day programbeing only the second most popularselection, 250 students enrolled eachyear. Full day kindergarten participationcost $70.00 per week versus the fee-free half and alternate-day programs.The most popular program amongparents was the alternate-day programwith 500 participants, with the least

popular program being half day havingan enrollment of 167 students.Transportation was provided to andfrom school for the alternate-daystudents, but midday transportation wasnot available for half day attendants.

Two measures of literacy wereadministered pre and post from theObservation Survey of Early LiteracyAchievement: 1. Letter Identification, 2.Concepts about Print. Pre/postdifference scores across the threeclassrooms were compared using theBrown-Forsythe test statistic. This teststatistic is a complement to ANOVAused for unbalanced sample sizes. Astatistically significant difference at thep < 0.05 level was found to exist acrossclassrooms when comparing the gainscores for both Letter Identification andConcepts about Print. A comparison ofthese differences is reported in Table 4and Table 5.

Table 4. Pre/Post Gains for Letter IdentificationProgram Comparison Corresponding

MeansSignificance*

Full day – Half day 14.09, 16.00 nsFull day – Alternate-day 14.09, 13.71 nsHalf day – Alternate-day 16.00, 13.71 p < 0.05

*Test statistic based on Dunnett’s C

Table 5. Pre/Post Gains for Concepts about PrintProgram Comparison Corresponding

MeansSignificance*

Full day – Half day 7.52, 7.01 nsFull day – Alternate-day 7.52, 6.88 p < 0.05Half day – Alternate-day 7.01, 6.88 ns

*Test statistic based on Dunnett’s C

16

Renbarger’s literacy study suggeststhat in finding significant differencesbetween half and alternate-daykindergarten programs for LetterIdentification, children may bedisadvantaged in an alternate-dayclassroom with respect to frequency ofrehearsal and exposure to concepts.Here, students with daily exposure tolearning activities indeed measuredhigher gains. Regarding the significantdifference between full and alternate-day programs with respect to conceptsabout print, Renbarger’s work againsuggests that frequency of exposure tocurricular learning activities results inhigher achievement gains.

Perry Township Study

Full day kindergarten in PerryTownship is a part of the On TrackProgram, an initiative designed toimprove education and academicsuccess for at-risk children. Aftersecond grade, the On Track studentsenter the regular classroom, but it ishypothesized that by the end of thekindergarten year, the students shouldbe “on track” developmentallyregarding their behavior, motivation,and academics (Zielke, 2002).

On Track participants are chosenthrough a three-step process: 1) aparent/caregiver interview; 2) testingon a nationally normed kindergartendevelopmental and basic skills screen;and 3) three weeks of classroomobservations. If all three measuresyield results that are interpreted toshow that a child was at-risk, he or sheis invited to experience the benefits ofthe On Track Program (Zielke, 2002).

In order to determine if the On TrackProgram is successful in helping at-riskstudents to function within a nationallystandardized range of behavior, a self-regulation, cognitive and behaviorassessment instrument wasadministered in three consecutive years(Spring 2000, Spring 2001, and Spring2002). In the spring of 2002,kindergarten teachers completed theConners’ Teacher Rating Scale-Revised: Long Version (CTRS-R:L) foreach of the On Track students in thestudy as well as their matched non-at-risk controls (Zielke, 2002).

Eighty-three percent (83%) of thethirteen subtest scores for each OnTrack student were in the normal range,as defined by the CTRS-R:L, while1.2% of the scores fell in the significantconcern range. Although a greaternumber of the non-at-risk matchedcontrols’ scores were in the normalrange (99.1%) and none wereconsidered to be of significant concern,the On Track Program appears to beeffective due to the large percentage ofat-risk students functioning within age-appropriate limits after one year in theprogram (Zielke, 2002).

When individuals’ scores wereaveraged to create 13 mean subtestscores for each group (the On Trackstudents and the non-at-risk controlgroup), both groups fell within thenormal range of functioning. Thecontrol group had lower mean scoreson the subtests, however, indicatingbetter functioning, with the largestdiscrepancy between groups being onthe “anxious/shy behavior” subtest.Each student was assigned an overallmean score on the CTRS-R:L, andthese scores were compared for eachmatched pair. Sixteen of the nineteenOn Track students fell within thenormal functioning range, and eight ofthe subjects either met or surpassed

17

their matched control’s score, indicatingthe performance of an at-risk student ator above the level of a non-at-risk peer(Zielke, 2002).

Muncie Community Schools Study

Based on the positive reaction to a pilotprogram which offered full daykindergarten to a portion of Title Istudents, the Muncie CommunitySchools in 1996 decided to beginimplementing full day kindergartens inall of their Title I schools. By 1998, theExtended Day Kindergarten programwas in place in all seven Title I schoolsand was providing additional qualitytime on task and attempting to level theplaying field for at-risk students(VanFleet, 2002).

In 2002, the Director of ElementaryInstruction for the Muncie CommunitySchools produced a report evaluatingthe status of the full day kindergartenprograms (VanFleet, 2002). Inpreparation for this report, VanFleetmet frequently with Title I principals andteachers and obtained TerraNova andISTEP+ test data from 1997 through2002. VanFleet analyzed the full dayprogram’s progress on the areas ofstandards, curricular uniformity,communication, and sustainability.

Test data included in the report werethe within-schools comparison of thescores of each cohort as theyprogressed through grade levels as wellas a comparison by school of the meanscores for each grade from years 1997to 2001. In general, after an initial jumpin scores between the kindergarten andfirst grade years, as the full daykindergarten students advanced fartherin school, their total scores on the testsdeclined. VanFleet (2002) addresses

this trend in the sustainability section ofhis report, and he suggests increasedcommunication among teachers as thestudents pass from grade to grade. Forthe most part, from year to year eachgrade level’s mean scores haveincreased. However, to truly be astandardized comparison, moreinformation on the test scores, such aspercentile rank, is needed to assess anyactual improvement.

In addition to providing the test data,VanFleet also offers somerecommendations for improvement ofthe Muncie full day kindergartenprogram. From his study, he observedthat there is little standardization ofcurricular strategies or materials amongschools or teachers within the sameschools. As previously mentioned,VanFleet (2002) also writes of theimportance of communication pertainingto teaching strategies and studentprogress among teachers at differentgrade levels. With these improvements,the full day Title I program can developinto an even more effective means ofleveling the playing field foreconomically disadvantaged students.

Indianapolis Public Schools Study

In 1997, the Indianapolis PublicSchools designed and implemented tenpilot full day kindergartens to test theimpact of extended learningopportunities on students. Thekindergartens were placed in highschools and elementary schools andwere compared on a measure ofvocabulary (the Peabody PictureVocabulary Test-Revised) to fivegeneral education half day classes andthirteen Title I full day classes. Resultswere convincingly supportive of thebenefits of full day kindergarten,

18

regardless of location, and the researchteam strongly recommended that theIndianapolis Public Schools fund fullday programs (Tatum, 1998).

Four years later, the 2001-2002Indianapolis Public SchoolsKindergarten Programs Annual Report(Beatty, 2002) describes three differentfull day kindergarten programs: Title IFull Day Kindergarten, Magnet FullDay Kindergarten, and Full DayLottery Kindergarten. Participation inthe Title I programs is determinedduring the first week of school, and thechildren most at risk for academicfailure are placed in this extendedschool day. There are 47 Title I FullDay programs in the district, eachaveraging a class size of 16. Thetwelve Magnet Full Day Kindergartensare populated through a lotterydrawing, the traditional magnet process.The Full Day Lottery Kindergartenprograms select their 230 students (23per class) through a lottery as well andare located in elementary, middle, andhigh schools (Beatty, 2002).

In order to reach 100% studentachievement by 2014, as outlined in theNo Child Left Behind Act of 2001(NCLB), the district has set short termgoals that focus on building educationalfoundations in kindergarten. To assessthe progress of the kindergartenprograms toward reaching these goals,during the 2001-2002 school yearstudents were given the Signposts EarlyLiteracy Battery, an instrument whichmeasures language arts and readingskills. Although designed to determinestudent achievement levels in relation tothe NCLB standards, this assessmentalso yielded interesting results about theperformance of students in full daykindergarten programs compared to

those in half day programs (Beatty,2002).

Excluded from this summary are thescores of Half Day Plus Daycareprograms, as they made up only 6% ofthe sample. Full Day LotteryKindergartens had the highestpercentage (73%) of students scoringat or above the 50th percentile on theSignposts Early Literacy Battery,followed by Magnet Full DayKindergartens (64%). The Title Iprograms had the least amount ofachievement (50%), just behind HalfDay programs at 53% (Beatty, 2002).

Magnet Full Day Kindergartens hadthe highest percentage of students atthe advanced reading level (8%), whileFull Day Lottery and Half DayKindergartens tied at 3% and Title IFull Day Kindergartens had 2%. Atthe middle reading level, Full DayLottery Kindergartens had the highestpercentage of students (79%), whileHalf Day Kindergartens had 64% andMagnet Full Day Kindergartens had63%. Title I Full Day Kindergartenshad the lowest percentage of students(60%) performing at the middle level.Conversely, the Title I programs hadthe highest percentage of students withlow reading abilities (38%). Half dayprograms had 33% of their students atthe lowest reading level while Magnetand Lottery Full Day Kindergartenshad 29% and 18%, respectively(Beatty, 2002).

Data comparing the 2001 to 2002scores was not available for theMagnet Full Day Kindergartens, butthe Title I and Lottery Full DayKindergartens showed improvementson various measures. The Title I FullDay programs increased the

19

percentage of students scoring at orabove the 50th percentile from 36% to50%, and they decreased thepercentage of students at the lowestreading level from 40% to 30%. FullDay Lottery Kindergartens also noticedan improvement in meeting the 50th

percentile standard, from 61% ofstudents in 2001 to 73% in 2002, andthe percentage of students reading atthe lowest level decreased from 25% to18% (Beatty, 2002).

Overall, the full day programs aremaking progress toward the district’sshort-terms achievement goals. TheMagnet and Lottery Full DayKindergarten students are alsooutperforming the Half DayKindergarten students, further evidenceof the effectiveness of this type ofeducation.

Analysis of Data from a Large,Urban School District

Detailed ISTEP+ data were obtainedfrom a very large, urban school districtin Indiana. After removing records withmissing data, the sample size for theanalyses was 1,886. All students werethird graders during the 1998-1999

school year and attended either half orfull day kindergarten in the samedistrict. Pre-kindergarten measures ofability or achievement were notavailable, but full day kindergarten inthis district was targeted solely in Title Ischools during these years. The districtimplemented full day kindergarten,using a combination of Title I anddistrict funds, primarily to lessen theachievement gap between students.

A multivariate analysis of variance wasused to analyze student standardizedscores on the ISTEP+, withmathematics and language artsstandardized scores as dependentvariables and program type (half or fullday), free lunch status (pay for lunch orfree/reduced lunch), and race (blackand Hispanic or white) as independentvariables. Results provide evidence thatthe differences between full and halfday students are negligible, and theinteractions between program type,student race, and lunch status also lacksignificance (Table 6). However,although differences related to thethree-way interaction are very small,we noticed an interesting pattern amongthe data (Figures 1-4).

Table 5. Multivariate Analysis of Variance of ISTEP Language Arts and Mathematics Scores

Source of Variance Wilks’ Lambda df1 df2Multivariate

Fp η2

Race .936 2 1839 62.72 < .001 .064

Lunch .963 2 1839 34.95 < .001 .037

Program Type .997 2 1839 2.77 .06 .003

Race x Program .999 2 1839 .57 .56 .001

Lunch x Program 1.000 2 1839 .43 .65 .000

Race x Lunch x Program .997 2 1839 2.36 .10 .003

20

Figure 1. ISTEP+ Language Arts Scale Scores for Black and Hispanic Students by Lunch Status

Figure 3. ISTEP+ Language Arts Scale Scores forBlack and Hispanic Students by Lunch Status

400

420

440

460

480

500

520

540

Pay Lunch Reduced + Free Lunch

AM + PM Program Full-day Program

Figure 2. ISTEP+ Language Arts Scale Scores for White Students by Lunch Status

400

420

440

460

480

500

520

540

Pay Lunch Reduced + Free Lunch

AM + PM Program Full-day Program

Figure 3. ISTEP+ Mathematics Scale Scores for Black and Hispanic Students by Lunch Status

400

420

440

460

480

500

520

540

Pay Lunch Reduced + Free Lunch

AM + PM Program Full-day Program

21

Interpretation of these results is difficultin the absence of pre-kindergartenability and achievement data, but theresults suggest that the district’s goal forimplementing full day kindergarten –shrink achievement gaps between highperforming, middle class, white studentsand low performing, poorer, racialminority students – may have beenachieved (i.e., full day students, whopresumably started at a lower level thanmost half day students, perform as wellas half day students by third grade, andminority students and non-reduced/freelunch students appear to benefit fromfull day programs more than otherstudents). However, the district still hasa problem with economicallydisadvantaged and minority studentsperforming significantly lower than otherstudents, although the effect sizeestimate provides evidence that thedifference is small.

Analysis of Data from a School in aRural District

A concern among educators is thedegree to which kindergarten instructionshould be led by certified teachers (i.e.,Can aides run extended day programs,therefore saving money?). Project staffobtained data from a rural Indianadistrict to investigate this issue. The finalsample was 162 students, 70 of whomattended half day kindergarten with anextended day program in 1999-2000and took the 2002 ISTEP in third grade,92 of whom attended a full day programin 2000-2001 and took the 2003 ISTEPin third grade.1 Chi square tests ofstatistical significance were used toanalyze the pass rates on language artsand mathematics subtests with Cramer’sV calculated as an effect size estimate.District personnel were not aware of anyother instructional or structural changesin those schools that would account forany differences between the 2002 and2003 cohorts.

Figure 4. ISTEP+ Mathematics Scale Scores forWhite Students by Lunch Status

400

420

440

460

480

500

520

540

Pay Lunch Reduced + Free Lunch

AM + PM Program Full-day Program

22

Two sets of analyses were conducted,with all students included in the first setand all students with the exception ofstudents receiving special educationservices included in the second set. Forall students, statistically significantdifferences do not appear to existbetween the two cohorts on ISTEP+language arts scores (÷2 (2) = 2.60, p =.27, V = .13), but small to moderatedifferences were found on mathematicsscores (÷2 (2) = 9.72, p = .008, V =.25). These results suggest that studentswho experienced full day kindergartenpassed the ISTEP+ mathematics examsat higher rates than half day/extendedday students (70% vs. 89%,respectively). Language arts passingrates were not significantly different(71% full day vs. 77% half day). Resultswere similar when students receivingspecial education services wereremoved from the sample. These resultsprovide evidence that students whoparticipated in full day kindergartenwere more likely than half day/extendedday students to pass the mathematicsportion of the ISTEP+ exam in thirdgrade.

Summary

Results from the eight Indiana datasets reflect the results of the nationalresearch on full versus half daykindergarten. As was the case with thenational data, the Indiana researchsuggests that there are no negativeoutcomes commonly associated withfull day kindergarten, and that – atworst – full day kindergarten and halfday kindergarten have similar effects.Significant results in support of thebenefits of full day over half daykindergarten were found in many ofthe comparisons within these studies.When analyzed on the majordimensions of academic achievement,retention and special educationreferrals and social and behavioraleffects, the benefits of full daykindergarten programs are apparent.

1 The extended day program was taught byan aide and was not coordinated withteachers in the half day kindergartenprogram.

23

Section III: Analysis of Daily Activities in Full Versus Half DayKindergarten Programs

As Center staff conducted the researchreported in the first two sections of thisreport, one specific criticism of full daykindergarten emerged as anunexamined issue. To many critics, fullday programs have the potential to benothing more than half day kindergartenwith an extra half day of play time. Weidentified only two recent studies thatprovide detailed analyses of how timeis spent in full and half day programs:one national study and one study ofstudents in Wisconsin.

Given this paucity of research, weconducted a two phase study ofinstructional activities in Indianakindergarten classrooms: In phase one,we collected full and half daykindergarten schedules from Indianaschools and compared the scheduledactivities. In phase two, we conductedseveral site visits to full day programsto establish the validity of submittedschedules. In this section of the report,we review the two published studies,describe the analysis of Indianakindergarten schedules, and summarizethe results of the site visits. Weconclude this section with a briefsummary.

Published Studies on How Time isSpent in Full and Half DayClassrooms

Research from the 1970s and early1980s suggested that extending halfday programs to a full school daywould have little positive effect if theextra time was used merely aschildcare or babysitting (Harding &Safer, 1988). These same authorsconclude that research from that period

suggests that full day kindergarten timeshould be spent “providing a variety ofeducation activities related to the …needs of kindergarten children,” thatthese activities should bedevelopmentally appropriate, and timeshould be set aside for both structuredand unstructured play (p. 61).

Elicker and Mathur (1997) conducted acomprehensive, multi-year evaluation offull and half day programs in anunidentified community in Wisconsin.They found that students in full dayclassrooms spent significantly more timein small-group teaching, one-on-oneteacher-student interactions, and self-initiated learning activities. These self-initiated activities, which accounted forapproximately 85 minutes per day,included play. Although full day studentsalso spent more time in large-groupteaching contexts, the percentage oftime spent in these activities wasconsiderably less than that experiencedby half day students.

Denton et al. (2003) report the resultsof a recent survey of classroomactivities from the Early ChildhoodLongitudinal Study. The data, whichwere teacher-reported, indicated thatpercents of time spent on variousinstructional approaches was similaracross program types. However, resultsregarding play time were not reported(Figure 5).

24

At the same time, ECLS data provideevidence that full day programs aremore likely than half day programs touse mixed-level grouping, achievementlevel grouping, and peer tutoring. Inaddition, full day programs were morelikely than half day programs to spendtime on certain skills each day,including “letter recognition, letter-sound match, conventions of print,vocabulary, making predictions basedon text, using context clues forcomprehension, rhyming words,reading aloud, reading multi-syllablewords, and alphabetizing” (p. 12).

Curriculum/Schedule Analyses

With the assistance of the IndianaAssociation of Public SchoolSuperintendents, each public schooldistrict in the state was asked to submittypical daily schedules for their full andhalf day programs. Extended day oralternate day-full day programs were

not included in the analysis. Afterremoving those schedules from thedata, schedules from 131 programtypes were received: If a districtprovided more than one half or full dayschedule, all schedules of similarprogram type were averaged together.Data from nine schools were excludedfrom the analyses due to missing data.

Minutes per day were calculated foreach of the following categories:language arts, including English, writing,reading, and related activities;mathematics; other instruction; andstructured play time. Averages andstandard deviations for each categoryof activity are included in Table 7. T-tests were conducted to determinewhether the differences betweenprogram types were statisticallysignificant, and Cohen’s d wascalculated as an effect size estimate. Aswas expected, full day kindergartenschedules included much more total

Figure 5. Average Percent of Class Time that Public Kindergarten Classes Used VariousInstructional Approaches, Spring 1999

Note. Data from Denton et al. (2003).

38

27

1520

40

28

1418

05

101520253035404550

Whole class Small group Teacher-directedindividual

Child-directed

Full day

Half day

25

instructional time than half dayschedules (t = 23.52, p < .01, d =3.99): Full day mean = 284.10 minutes(SD = 43.75), half day mean = 141.47minutes (SD = 21.38). Whenextrapolated to a 180-day school year,this represents an additional 428 hoursof instruction per year. Assuming a sixhour school day, 428 hours representsroughly 71 additional days ofinstruction.1 When analyzed bycategory (Table 7), full daykindergarten schedules includedsignificantly more minutes for each typeof activity than half day schedules.

To determine the relative use of specificactivities in each type of kindergartenprogram, schedules were reanalyzed todetermine the length of the school dayin minutes. After subtracting minutes

Table 7. Statistical Test for Difference of Spending Time (minutes) between Half Day and Full DayKindergartena

Mean(SD)

Half day Full dayt p Effect Size a

LA 55.00(28.64) 107.45(44.71) 7.75 <.001 1.35

Math 22.64( 8.54) 35.21(15.05) 5.33 <.001 .96

OtherInstruction

71.48(32.24) 143.59(45.88) 10.06 <.001 1.78

Play 16.88( 8.03) 35.10(17.71) 6.95 <.001 1.21a N = 120

Note. Cohen’s d was used for the effect size. Pooled estimate of the population’s standard

deviation was calculated as follows;total

pooled df

dfdf )ˆ()ˆ(ˆ

222

211 σσσ ⋅+⋅

= .

devoted to lunch (but not recess), thistotal was used to determine the percentof time spent in each category of activity(Table 8). With respect to totalinstruction time, a slightly higherproportion of total time was devoted toinstruction in half day classrooms (t =2.45, p = .02, d = .45): Half day mean= 84.42% (SD = 10.65), full day mean= 78.50% (SD = 14.71). Whenanalyzed by category (Table 8), the onlysignificant difference was thatmathematics instruction accounted for ahigher proportion of instructional time inhalf day versus full day programs (note,however, that minutes of instruction inmathematics was higher in full dayprograms). Interestingly, the percent oftime devoted to play was similar acrossboth programs.

26

Site Visits

After collecting the schedule, Centerstaff became concerned that reportedtime may be different from how time isactually spent in kindergartenclassrooms. To investigate this issue,staff conducted site visits to 14classrooms in six schools that hadsubmitted full day kindergartenschedules. Schools were visited inurban, suburban, and rural districtsspread across Indiana. Observationslasted two to four hours and wereconducted at various points throughoutthe school day, and most teachers werenot given advanced notice of the exacttiming of the observations by the personorganizing the visits at each school. Atsome schools, teachers andadministrators were interviewed aboutfull day kindergarten after the classroomobservations.

Project staff observed instructionalactivities almost exclusively throughoutthe day, with the exception of recess

time following lunch. Child-initiatedactivities, which were categorized bythe published studies as play time, didnot involve play in the observedschools. Rather, these activities wereinstructional or provided opportunitiesfor review and application of materialthat was being covered that day inclass. For example, in one classroom,students were given 10 minutes to writeon a topic of their choosing after 10minutes of teacher-directed writinginstruction and practice. In severalschools, child-initiated activitiesoccurred through the use of learningcenters: In one classroom, studentsrotated among nine independentlearning centers for 45 minutes eachmorning, with the expectation that theywill visit four centers per day. Thecenters included activities in the areasof poetry, reading, listening, puzzles,language arts, computers, mathematics,art, and writing.

Table 8. Statistical Test for Difference of Portion of Spending Time (%) between Half Day and FullDay Kindergartena

Mean(SD)

Half day Full dayt p Effect Size

LA 33.06(17.48) 29.29(12.27) 1.37 .18 .26

Math 13.30( 4.97) 9.40( 3.96) 4.27 <.001 .90

OtherInstruction

42.91(18.88) 40.38(17.43) .75 .46 .14

Play 10.06( 5.21) 9.80( 5.97) .21 .84 .04a N = 122

27

Summary

The research literature and datacollected for this report provideevidence that time in full daykindergarten programs is different bothquantitatively and qualitatively from howtime is used in half day programs.Across all of the schools in the Indianasample, the proportion of instructionaltime is similar across program types,resulting in much greater instructionaltime in full day programs, representingapproximately 40-50% more instructionin full day programs than half dayprograms. The Wisconsin study and theIndiana site visits suggest that inindividual classrooms, the additionaltime leads to greater use of child-initiated activities. In the site visitschools, these activities were almostuniversally instructional in nature and didnot involve play. The published researchalso provides convincing evidence thatcertain types of reading skills andgrouping strategies are more prevalent infull day programs, including readingaloud, peer tutoring, and mixed-abilitygrouping. During the Indiana site visits,researchers saw evidence of theseactivities, but half day programs werenot observed and therefore comparisonscannot be made.

1 Assuming a five hour day, which is prob-ably more reasonable given that 100% of sixhours is not used for instruction, 428 hoursrepresents nearly 86 additional days ofinstruction.

28

29

1. Both the Indiana and national data collected and analyzed for this report provideevidence that, relative to half day programs, full day kindergarten is associatedwith a wide range of positive outcomes, including increased student achievementand social and behavioral development.

• In both our site visits and several of the published studies, teachers reportedthat the full day format allowed time to address state standards more effectivelyand address the diverse learning needs of students of differing abilities. Thiseffect cannot be assessed for a few years, but the impact on ISTEP+ scorescould be substantial if teacher perceptions are accurate.

• Any state-funded full day kindergarten program should include an evaluationcomponent to promote accountability. Although evaluation is critical to thesuccess of any educational program, evaluation is especially important insituations where programs should result in significant new expenditures and newsavings – a system should be put in place to ensure that savings related to, forexample, reduced special education referrals are being realized.

2. The positive outcomes associated with full day kindergarten appear to be largerfor disadvantaged students in both the national and Indiana research.

• Full day kindergarten appears to be effective in reducing achievement gaps. Iffunding for universal full day kindergarten is not available in the currenteconomic climate, funding could be focused on providing full day kindergartento schools with low achieving subgroups of students. National research suggeststhat minority students and students of lower socioeconomic means are morelikely to benefit from full day programs if the class size is fewer than 25 and anaide is available in the classroom.

3. Full day kindergarten, regardless of its organization and funding mechanism, isexpensive relative to half day programs. Costs include additional teachers,instructional aides, and classroom space (Harding, 1988; Rothenberg, 1984). InIndiana, the most widely cited current estimate for the costs of a full daykindergarten initiative is roughly $110 million.

• Schools, both nationally and in Indiana, use a range of strategies to pay for fullday kindergarten programs. The most common sources of funding are the stategeneral fund, existing Title I funds, and parent fees (often calculated on a slidingscale relative to family income).

Conclusions and Recommendations

30

• Savings resulting from full day kindergarten are difficult to determine.Substantial savings should be realized over the long-term due to reducedspecial education referrals and the need for less remediation, reduced need formidday transportation and crossing guards, and reduced need for half daychildcare programs. However, childcare costs will not be entirely eliminated(Elicker, 2000), as many families may still rely on childcare both before andafter students attend full day programs each day.

• A number of existing “full day” programs may actually be extended dayprograms, which are often staffed with aides. Any anticipated savings based onthe existence of current programs may prove to be smaller than anticipated.

• Alternate day full day programs are appealing due to the potential for reducedcosts, but this type of program is generally not associated with positiveoutcomes relative to every day full day or every day half day programs.

4. The literature contains many comments about the importance of quality versusquantity of kindergarten experience (i.e., it’s not full day, it’s what happens in fullday that counts). Although this perspective is valid, it oversimplifies the researchon instructional activities in full day classes. A better perspective is that the addedtime in a full day program fundamentally changes the nature of activities that occurin that program. Not only do teachers tend to do more in full day programs, theytend to do more of the instructional strategies that researchers recommend topromote young children’s learning.

• Although a few studies suggest that small class sizes are more effective than fullday kindergarten in raising student achievement, other studies provide evidencethat full day classes of moderate size (e.g., fewer than 25 students) are optimal.Indeed, Walston et al. (2002) found evidence that full day kindergarten doesnot necessarily mitigate the negative effects of large class sizes on studentachievement.

31

References

Beatty, N. (2002). Indianapolis PublicSchools kindergarten programsannual report 2001-2002 including atwo-year comparison of studentperformance. Indianapolis, IN:Indianapolis Public Schools Title IDepartment.

Clark, P. (2001). Recent research on all-daykindergarten. ERIC Digest.Champaign, IL: ERIC Clearinghouse onElementary and Early ChildhoodEducation.

Clark, P., & Kirk, E. (2000). All-daykindergarten. [Electronic version].Childhood Education, 76, 228-231.

Coladarci, T., & Ervin, R. (2000). Full dayversus half day kindergarten: Effectson standardized and local assessmentsof academic outcomes. Paper presentedat the annual meeting of the AmericanEducational Research Association, NewOrleans, LA.

Cryan, J.R., Sheehan, R., Wiechel, J., &Bandy-Hedden, I.G. (1992). Successoutcomes of full day kindergarten: Morepositive behavior and increasedachievement in the years after. EarlyChildhood Research Quarterly, 7, 187-203.

da Costa, J.L, & Bell, S. (2000, May). Fullday kindergarten at an inner cityelementary school: Perceived andactual effects. Paper presented atAnnual Conference of AmericanEducational Research Association, NewOrleans, LA. (ERIC ReproductionService No. ED 451 938).

Denton, K., West, J., & Walston, J. (2003).The condition of education specialanalysis: Reading—young children’sachievement and classroomexperiences. U.S. Department ofEducation Institute of EducationSciences NCES 2003-070. http://nces.ed.gov/pubs2003/2003070.pdf .Retrieved 10/17/2003.

Education Commission of the States (2003,August). How states fund full daykindergarten. Denver, CO: Author.(available at http://www.ecs.org/

Elicker, J., & Mathur, S. (1997). What do theydo all day? Comprehensive evaluation offull day kindergarten. Early ChildhoodResearch Quarterly, 12, 459-480.

Entwisle, D. R. & Alexander, K. L. (1998).Facilitating the transition to first grade:The nature of transition and research onfactors affecting it. Elementary SchoolJournal, 98, 351-364.

Evans, E.D., & Marken, D. (1983).Longitudinal follow-up comparison ofconventional and extended-day publicschool kindergarten programs. (ERICDocument Reproduction Service No. ED254 298).

Fusaro, J.A. (1997). The effect of full daykindergarten on student achievement: Ameta-analysis. [Electronic version]. ChildStudy Journal, 27, 269-280.

Goodpaster, M. (1999, January). Full daykindergarten and school readiness[Fiscal impact statement]. Indianapolis,IN: Legislative Services Agency, Officeof Fiscal and Management Analysis.(available at S:\Projects\EDPOL\Full DayKindergarten\readings\Fiscal ImpactStatement, Senate Bill 0069.htm)

Goodwin, J. (1989, January). An analysis ofattendance and achievement for full dayand half day kindergarten, 1987-1988(Report No. 8911). Philadelphia SchoolDistrict, Pennsylvania Office of Researchand Evaluation. (ERIC Document ServiceNo. ED 308 939).

Harding, C.G., & Safer, L.A. (1988). Keepingour priorities straight: The real issue inthe full day kindergarten question.Planning & Changing, 19, 58-62.

Hills, T.W. (1985, September). All-daykindergarten: Resources for decisionmaking. New Jersey State Department ofEducation, Trenton Division of GeneralAcademic Education. (ERIC DocumentService No. ED 287 603).

Hoffman, A.R., & Daniels, S.J. (1986).Reading and reading readinessinstruction: A comparison of all-dayand half day kindergarten practices.(ERIC Document Reproduction ServiceNo. ED 271 226).

32

Holmes, C.T., & McConnell, B.M. (1990,April). Full day versus half daykindergarten: An experimental study.Paper presented at the annual meetingof the American Educational ResearchAssociation, Boston. (ERIC DocumentReproduction Service No. ED 369 540).

Hough, D., & Bryde, S. (1996). The effects offull day kindergarten on studentachievement and affect. (ERICDocument Reproduction Service No. ED395 691).

Humphrey, J.W. (1980). A study of theeffectiveness of full day kindergarten.(ERIC Document Reproduction ServiceNo. ED 190 224).

Johnson, E.W. (1974). An experimentalstudy of the comparison of pupilachievement in the all daykindergarten and one half day controlgroup. (ERIC Document ReproductionService No. ED 115 361).

Jones, H.L, Pollock, B., & Marockie, H.(1988). Full day kindergarten as atreatment for at-risk students: Ohiocounty schools. ERS Spectrum, 6, 3-7.

Karweit, N. (1985). Should we lengthen theschool term? Educational Researcher,14, 9-16.

Karweit, N. (1992). The kindergartenexperience. Educational Leadership,49, 82-86.

Koopmans, M. (1991). A study of thelongitudinal effects of all-daykindergarten attendance onachievement. Newark Board ofEducation, NJ. Office of Research,Evaluation and Testing.

Lore, R. (1992). Language developmentcomponent: Full day kindergartenprogram. Columbus, OH: ColumbusPublic Schools.

Lysiak, F., & Evans, C.L. (1976).Kindergarten—fun and games orreadiness for first grade: A comparisonof seven kindergarten curricula. Paperpresented at the annual meeting of theAmerican Educational ResearchAssociation, San Francisco, CA. (ERICDocument Reproduction Service No. ED121 803.

Madison Metropolitan School District.(1985). Evaluation of the full daykindergarten pilot, 1984-85, Madison,WI. Unpublished Manuscript.

McClinton, S.L., & Topping, C. (1981).Extended day kindergartens: Are theeffects intangible? Journal ofEducational Research, 75, 39-40.

McConnell, B.B., & Tesch, S. (1986).Effectiveness of kindergartenscheduling. Educational Leadership,44, 48-51.

Mueller, E.J. (1977). Effects of entry level andpattern of Title I. (ERIC DocumentReproduction Service No. ED 151 419).

Naron, N.K. (1981). The need for full daykindergarten. Educational Leadership,38, 306-309.

Nieman, R.H., & Gastright, J.F. (1981a). Thelong-term effects of ESEA Title Ipreschool and all day kindergarten: Aneight year follow-up study. (ERICDocument Reproduction Service No. ED198 949).

Nieman, R.H., & Gastright, J.F. (1981b) Thelong-term effects of Title I preschool all-day kindergarten. Phi Delta Kappan, 63,184-185.

Ohio State Legislative Office of EducationOversight. (1997). An overview of fullday kindergarten. Columbus, OH:Author. (ERIC Reproduction Service No.ED 408 046).

Pasco School District 1, Washington. (1987).Effectiveness of preschool; andcomparing full day half day, andalternate-day kindergartens. (ERICDocument Reproduction Service No. ED294 648).

Puleo, V.T. (1986). Current researchperspectives on full day kindergarten.ERS Spectrum, 4, 32-38.

Puleo, V.T. (1988). A review and critique ofresearch on full day kindergarten. TheElementary School Journal, 88, 427-439.

Robinson-Lewis, G. (1991, August). Full daykindergarten programs: 1988-1989,1989-1990, Summative evaluation.Kansas City, MO: Kansas City SchoolDistrict. (ERIC Reproduction Service No.ED 346 964).

Rothenberg, D. (1984). Full day or half daykindergarten? ERIC Digest. Urbana, IL:ERIC Clearinghouse on Elementary andEarly Childhood Education. (ERICDocument Reproduction Service No. ED256 474).

Sergesketter, K. & Gilman, D. (1988). Theeffect of length of time in kindergartenon reading achievement. (ERICDocument Service No. ED 324 664).

33

Stofflet, F.P. (1998). Anchorage schooldistrict full day kindergarten study: Afollow-up of the kindergarten classes of1987-88, 1988-89, 1989-90. Anchorage,AK: Anchorage School District. (ERICDocument Reproduction Service No. ED426 790).

Tatum, D. S. (1998). A comparison ofkindergarten programs in theIndianapolis Public Schools. Chicago,IL: The University of Chicago, MESAPsychometric Laboratory.

Tatum, D.S. (1999). Is more better?Measuring the effects of full daykindergarten. Popular Measurement, 2,24-26.

VanFleet, W. (2002). All-day kindergartenreport. Muncie, IN: Muncie PublicSchools, Office of the Director ofElementary Instruction.

Walston, J., & West., J. (2002, April).Classroom organization and curriculumdifferences between full day and part-day kindergarten programs in thenation’s public schools. Paper presentedat the Annual Conference of theAmerican Educational ResearchAssociation April, 2002 New Orleans,LA.

Walston, J., West, J., & Rathbun, A. (2002,April). Kindergarten program featuresand gains in reading and mathematicsachievement of public school children:Class size,length of school day andclassroom aides. Paper presented at theAnnual Conference of the AmericanEducational Research Association April,2002 New Orleans, LA.

Wang, Y.L, & Johnstone, W.G. (1999).Evaluation of a full day kindergartenprogram. ERS Spectrum, 17, 27-32.

Wichita Public Schools, KS. Office ofProgram Evaluation. (1989). All daykindergarten: Evaluation update.(ERIC Document No. ED 317 603).

Wisconsin State Department of PublicInstruction. (1980). A comprehensivestudy and evaluation of three types ofkindergarten programs—final report.(ERIC Document Reproduction ServiceNo. ED 201 384).

34

35

Appendix A: Summary of Select Kindergarten Policies for All 50 States1

StateDistrict Offeringof Kindergarten

StudentAttendance inKindergarten

District Offeringof Full dayKindergarten

StudentAttendance inFDK

If FDK ismandatory, mustdistricts also offerHDK?

Alabama M* P M P Not specifiedAlaska P P P P N/AArizona M P P P N/AArkansas M M M P Not specifiedCalifornia M P P2 P N/AConnecticut M M P P N/ADelaware M M P P N/AFlorida M P P P N/AGeorgia M P M P Not specifiedHawaii M P P P N/AIdaho P P P P N/AIllinois M P P P YesIndiana M P P P N/AIowa M P P P N/AKansas M P P P N/AKentucky M P P P N/ALouisiana M M M M NoMaine M P P P N/AMaryland M M M P Not specifiedMassachusetts M P P P N/AMichigan P P P P N/AMinnesota M P P P N/AMississippi M P M P Not specifiedMissouri M P P P N/AMontana M P P P N/ANebraska M P P P N/ANevada M M P P N/ANew Hampshire P P P P N/ANew Jersey P P P P Not specifiedNew Mexico M M P P Not specifiedNew York P P P P N/ANorth Carolina M P M P NoNorth Dakota P P P P N/AOhio M M P P YesOklahoma M M P3 P Not specifiedOregon M P P P N/APennsylvania P P P P N/ARhode Island M M P P N/ASouth Carolina M M M P YesSouth Dakota P P P P N/ATennessee M M P P N/ATexas M P P P N/AUtah M P P P N/AVermont M P P P N/AVirginia M M P P N/AWashington M P P P N/AWest Virginia M M M M NoWisconsin M P P P N/AWyoming M P P P N/A*M=Mandatory, P=Permissive

1 Source data retrieved January 2, 2004, from http://mb2.ecs.org/reports/Report.aspx?id=8.2 Limited to Early Primary Program students.3 Oklahoma has since required districts to offer full day kindergarten programs.