NO CHILD LEFT BEHIND: WHAT WE KNOW AND WHAT WE NEED TO KNOW · 2011-05-22 · reform aims to hold...

NO CHILD LEFT BEHIND:WHAT WE KNOW AND WHAT WE NEED TO KNOW

GLENN MALEYKO, ED.S.Wayne State University

Department of Educational Leadership and Policy Studies

MARYrzA A. GAWLIK, PH.D.Wayne State University

Department of Educational Leadership and Policy Studies

Drawing from empirically based and theoretical literature in thefield, this review examines Adequate Yearly Progress (AYP) andthe accountability provisions found in Title One of the NCLBlegislation. States have the ability to statistically manipulatetheir AYP implementation, which may give a false impression tothe public that AYP is a consistent measure of school effective-ness across the country. The first section identifies themeasurement concerns with the implementation of AYP. Thesecond section of the review deals with the benefits of AYP andthe unintended consequences. The review also analyzes the com-plexities involved with establishing school accountability andthe effective and ineffective provisions of the NCLB reform.Finally, this paper concludes with recommended areas ofresearch for policymakers and educators alike who are interest-ed in sustainable reform.

IntroductionThe No Child Left Behind (NCLB)

reform aims to hold educational agenciesand states accountable for improving thequality of education for all students. Itseeks to identify and transform low-per-forming schools that have failed to providea high quality education to their studentsinto successful schools. Furthermore, theaccountability provisions in NCLB intendto close the achievement gap between highand low achieving students and especial-ly the achievement gaps between minorityand non-minority students along with theadvantaged and disadvantaged students.The reform seeks to accomplish this goalusing state assessment systems that aredesigned to ensure that students are meet-ing state academic and grade level contentexpectations (NCLB, 2002, section 101).

The implementation of the NCLB goalscalls for high-level standards that are mea-surable for all students. There is no doubtthat NCLB has provided for an increasedfocus on student populations that have tra-ditionally performed at low levels(Borowski & Sneed, 2006; Guilfoyle,2006; Haycock, 2006; Hess, 2006; Hess &Petrelli, 2006; Kane, et al. 2002; Lewis,2006) however, there are some faults withthe Adequate Yearly Progress (AYP) as towhether or not AYP will be able to pro-vide an accurate measurement of the goalsthat are stated in the Title One purposestatement of the NCLB legislation. Someof those faults include states being allowedto develop their own standards, test scoreproficiency levels, and statistical mea-surement formulas under AYP (Harris,2007; Olson & Jacobson, 2006; Popham,

600

No Child Left Behind... / 601

2005a; Porter, Linn, & Trimble, 2005;Wiley, Mathis, & Garcia, 2005). Cronin,Dahlin, Adkins, and Kingsbury (2007)found that fifty different educational mea-surement standards are implemented acrossthe United States.

With almost a decade into the reform,NCLB is a large and complex piece of leg-islation that elicits a focus on public schooleducation. Drawing from empirically basedand theoretical literature in the field, thisreview examines AYP and the account-ability provisions found in Title One of theNCLB legislation. States have the abilityto statistically manipulate their AYP imple-mentation, which may give a falseimpression to the public that AYP is a con-sistent measure of school effectivenessacross the country. The first section iden-tifies the measurement concems with theimplementation of AYP (Harris, 2007;Olson & Jacobson, 2006; Popham, 2005a;Porter, et al., 2005; Wiley, et al., 2005).The second section deals with the benefitsof AYP and the unintended consequences(Borowski & Sneed, 2006; Guilfoyle,2006; Haycock, 2006; Hess, 2006; Hess& Petrelli, 2006; Kane, et al. 2002; Lewis,2006). The review also analyzes the com-plexities involved with establishing schoolaccountability and the effective and inef-fective provisions of the NCLB reform.Finally, this paper concludes with recom-mendations for research for policymakersand educators alike who are interested insustainable reform.

Measurement Concerns with the Use ofAYP to Evaluate School Effectiveness

Statistical Issues, Manipulationsof the AYP data, and Lack ofConsistency across the Country

There are a number of problems asso-ciated with the usage of the current AYPformula in order to measure and evaluateschool effectiveness. The first problem areaconcerns the manner in which the statisti-cal manipulations are occurring with theimplementation of AYP across the coun-try. Borowski and Sneed (2006) concludethat the manner in which AYP is deter-mined is arbitrary in nature. States havethe ability to lower standards and manip-ulate statistical measures of AYP that mayresult in the lack of improvement in instruc-tion.

Porter et al. (2005) found that statesexercise flexibility in implementingachievement tests and this consequential-ly impacts whether or not schools or schooldistricts make AYP. They focus on threespecific areas of measurement that have amajor impact on AYP. Those areas include:the line of trajectory that states establishen route to a 100% proficiency rate by theyear 2014, the minimum number of stu-dents that are necessary in order for thereto be a subgroup population that will countfor AYP purposes, and whether or not thestate uses a confidence interval along withhow wide the confidence interval is indetermining if schools or districts reachthe proficiency targets that are requiredunder AYP. The implementation of NCLBby the United States Department of Edu-cation (USDOE) provides states with some

602 / Education Vol. 131 No. 3

Figure 1. Confidence Interval Example

Note. This is a hypothetical example based on the confidence interval used by the Michigan Department of Education

(2008b) for the calculation of AYP.

leeway in determining how they measureAYPin the three above mentioned areas ofmeasurement.

Confidence Intervals

Porter et al. (2005) provide empiricaldata on the number of states that are notusing confidence interval vs. those statesthat are using confidence intervals and thesize of the confidence intervals that theyare using. When examining the confidenceinterval decisions by states, they identi-fied 2 states that use a confidence intervallower than 95 percent, 14 that use a 95 per-cent with a 3 one tailed approach, 1 usesa 98 percent trajectory with a one tailedapproach, 16 use a 99 percent two tailedconfidence interval approach, 6 are unspec-ified, and 11 do not use a confidenceinterval.

Popham (2005a) explains that it is notstatistically sound to implement the use ofa confidence interval for AYP calculations.The manipulation of the confidence inter-

vals has a dramatic impact on the numberof students and schools that pass profi-ciency tests and it results in an impact onthe AYP status of schools. Figure 1 pro-vides an illustration of the confidenceinterval, which is hypothetical for illus-tration purposes but not based on an actualstate example. In Figure 1, the hypothet-ical cut score on the assessment is 65percent with a confidence interval of + 5percent and the students who scored 65percent or above are deemed to be profi-cient. The students who scored from 60percent to 65 percent did not meet the pro-ficiency cut score, but they fell within theconfidence interval and they are deemedto be proficient due to the possibility ofmeasurement error. When a state imple-ments the use of the confidence intervalstudents who have score within the confi-dence interval are calculated as proficientunder the AYP formula.

A single measurement with a cut score of 65% in order toattain proficiency with a confidence interval of + 5%

Test scoreRange

0% 5% 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% % % % % % % % % % % % % % % % % % %

Test score percentage achieved

JoNotpmficlent u Confidence hter%d(ProAsionalyPPmflcient) o Proflcient(aboaecut ne)


Figure 2. AYP proficient schools in Michigan.

Note. The data was-taken from the Center for Education Policy Studies (2007) report along with AYPdata from the Michigan Department of Education,

The end result is the perception thatschools are successful according to AYPwhen in fact it might have been the imple-mentation of the confidence interval thatallowed them to meet AYP standards. Forexample, starting in the 2004-05 schoolyear, Michigan began to use a two levelof standard error above and below studentproficient test scores (Center on Educa-tion Policy[CEP], 2007; MichiganDepartment of Education [MDE], 2008a).Thus, students who fall within the confi-dence interval qualify as proficient even ifthey do not meet the specific cut score onthe state assessment. The Michigan Con-solidated State Application for theimplementation of AYP (2005) that wasapproved by the USDOE with the use of

the confidence interval resulted in moreschools making AYP in Michigan (CEP,2007). For example, in the year 2004-05without the usage of a confidence intervalin Michigan, 79 percent of elementaryschools make AYP and 74 percent of mid-dle schools make AYP. With the use of theconfidence interval, 89 percent of ele-mentary schools in Michigan make AYPand 82 percent of middle schools makeAYP (CEP, 2007). Figure 2 illustrates thatthe use of a confidence interval can havean impact on the number of schools in astate that make AYP.

Subgroup Size (N)Some studies (Meir, Kohn, Darling-

Hammond, Sizer, & Wood, 2004; Olson

Pen�ent of AYP p�ficient Schoots in Ivichigan with andPercent of AYP proficient Sclioolls in Michigan with and

without the use of confidence intervals 2004-05

100%,

800/% 4 %... . . ..

70/*- -

03 ro confidence interwd501/6 --

400/6- m confidenc internd

209/6 -_ _ _

1090 - -00/0_--

riddle school elermntay school


& Jacobson, 2006; Porter, et al., 2005)found that there are significant inconsis-tencies from state to state with the size (N)of subgroups that count for AYP purpos-es. For example, Olson and Jacobson(2006) found that the minimum N size fora subgroup in Louisiana is 10 in contrastto California where it is 50 and it must bean equivalent of 15 percent of the schoolpopulation until the number reaches 100.The study includes data taken from theAssociated Press from the 2003-04 schoolyear enrollment figures. The study ana-lyzed students in grades 3 through 8 andgrade 10 since those are the grades that aretested each year under NCLB. The studyanalyzed five major subgroups that arecounted under the NCLB AYP provision,which includes: White, Black, Asian,Native American, and Hispanic. They didnot look at the other subgroups such asELL students or special education studentsbecause there would have been the possi-bility that those students could have beencounted more than once since they mightqualify into more than one category.

Based on the AYP data, Olson andJacobson (2006) found that over 1.9 mil-lion students across the United States arenot being counted for AYP purposes dueto the minimal subgroup requirements.This includes over 400,000 minority kidsin California that are not being counted.The number of students that were notcounted in the five subgroups that wereexamined includes fewer than 2 percent ofWhite students, 10 percent of Black andHispanic, 33 percent of Asian students,and 50 percent of American Indians. Theyalso found that over 93 percent of the stu-

dents are included in both an overall schoolpopulation and in a racial or ethnic cate-gory.

The findings by Olson and Jacobson(2006) with the statistical manipulation ofthe subgroup size are consistent with thefindings by Porter, et al. (2005) which con-cluded that different (N) size minimalsubgroup calculations were creating incon-sistencies from state to state in regards tothe number of schools that are making AYP.Porter, et al. (2005) found that the sub-group sizes that states are using for AYPcalculations range from 0 to 100 across theUnited States. There are nine states thatuse a minimal subgroup size that is greaterthan 42, fourteen states use a minimal sub-groups size of 40 and twenty-seven statesdefine a minimal subgroup of 34 or less.Figure 3 provides a sample of state mini-mal subgroup levels and it illustrates thedisparities with the minimal subgroup sizeimplementation among the states.


Figure 3. Subgroup minimal size in a sample of states.

10-a StbgoM rininmd size in a sarVe

d •StatM N

40

2D0

60

caTifiNna Raiodaa

Loustana Vchgan North Texas b.Caroina

0* In CaliFornia the subgroup must equal at least ISpercent of the population otherwise the minimal could rise is

to

100.

t In Florida if the subgroup can be as low as 50 students as long as they represent 10percent of the school

population.

Note. The data is available in the article by Olson and Jacobson (2006) and on the state department of education

websites via their consolidated school AYP applications that were submitted to the United States Department of

Education.

These studies (Olson & Jacobson, 2006;Porter el al. 2005) demonstrate that thereare inconsistent measurement standardsacross the county, which produces mis-leading information to the public when itcomes to AYP. This misleading informa-tion includes the success and failures ofschools across the country under the AYPaccountability provisions in NCLB. Theresearch provides professionals and thepublic with important information and itgives the perception that statistical manip-ulation is being used so that students arenot represented in specific subgroups. Thestudies uncover one of the problems withthe AYP calculation and the inconsisten-

cy from state to state.

Lines of TrajectoryPast research (Popham, 2005a; Porter,

et al. 2005; Wiley, et al., 2005) shows thatthe manipulation of the AYP line of tra-jectory has a major impact on the numberof schools that make AYP. Porter, et al. pro-vide an analysis of the number of statesthat are adhering to different measurementdesigns. They identify three specific formsor target lines of trajectory that states usewith the evaluation of school effectivenessaccording to AYP. The researchers pro-vide statistical data on the number of statesthat are using straight line achievement


trajectories, back end loaded trajectories,front end load trajectories, and straight linetrajectories with plateaus. They identified24 states that use the back loaded trajec-tory, 19 that use the straight trajectory withplateaus, 4 that use the straight trajectory,no states use the front loaded trajectory,and three states had an unidentified tra-jectory. This further exemplifies theinconsistent measures that are being imple-mented across the country when it comes

to determining the AYP status of schools.The following figures provide examplesof the line of trajectory that states mightchoose to use en route to a 100 percentproficiency level by the year 2014. Fig-ure 4 provides an illustration thestraightline trajectory, Figure 5 illustratesthe straight line trajectory with plateaus,and Figure 6 illustrates the back end loadedtrajectory.

Figure 4. Straight line trajectory.

Note. This figure is based on the der•-iption presented by Porter ct of. (2005) and Wiley et al. (2005) which Is aligncd with

the example presented by the Minnesota Department of7rducation in their Consolidated State Application.

FigureS. Straight line trajectory with plateaus.

Note. This figure is based on the description presented by Porter. ct al. (2005) and Wiley et al. (2005) which is aligned wiNh

the Illinois example (br their initial and final years ofAYP implentenatilon in their Consolidated State Application

100-So-

s'o70__

o40

•o i L Tn30201

0 -

Yaw

100

855

70

soOtstscore0 SO

40 -- le

20 eas

1010

YOW


Figure 6. Back end loaded trajectory.

Note. T'in ligure is based on the description presented by Porter ct al. (2005) and Wiley ct tit. (2005) which is alig,ned with

data prescnted by the Michigan Depotnment of rUducation (2008) in their Consolidated State Application.

Mathis (2004) calls it a fallacy tobelieve that schools have the ability tomake steady progress on achievement tests.According to Mathis, the implementationof the line of trajectory by NCLB is notfounded in any research. He argues thatthe implementation of the balloon effect(or back loaded trajectory as it is definedby Porter, et al. 2005) where there is a mod-erate amount of progress made initiallyand then a high level of progress in thelater years, is only delaying the inevitablethat all schools will eventually fail.Changes through the reauthorization ofNCLB might be the only solution to theproblem that eventually all schools willfail by 2014 (Mathis, 2004).

Cut Score ManipulationResearchers in the field (Darling-Ham-

mond, 2007a; Darling-Hammond, 2007b;Guilfoyle, 2006; Harris, 2007; Sunderman,Kim, & Orfield, 2005) call into question

the use of a single measurement cut scorewith the analysis of school effectiveness.Harris (2007) examined the use of cutscores in the NCLB accountability mea-sures using data sets from Michigan andFlorida to illustrate how the different lev-els of cut scores can affect the number ofstudents that are deemed high performingaccording to their poverty levels. Harrisshows that when states use different cutscores like those in Michigan and Florida,the level that is used can directly impactthe number of students that will be deemedproficient. Harris found that if a state usesa cut score that is halfway between themean achievement of high minority stu-dents and low minority studentpopulations, the result can be a dramaticshift in the numbers of students who donot make the proficiency level in the lowminority schools. Harris discovered thatstates can manipulate the numbers ofschools and students that are proficient by

100

cut Score 60

40

. -_ oftTri•--tw01

Year


simply changing the cut scores for theachievement of standards on their stateassessments. The result is a dramaticimpact on the proficiency averages ofschools that have high percentages ofpoverty and minority students. The cutscore manipulation that states might imple-ment can provide the appearance that gainshave occurred without any improvement inthe conditions of school programs or class-room instruction. The findings by Harrisidentify another method in which statescan use to modify the percentage of schoolsthat meet AYP standards that is not a resultof improvements in school quality. Themanipulation of the statistical data can havea dramatic impact on AYP scores.

Lack of Consistency withAYP Across the Country

The implementation of the AYPaccountability provisions in NCLB pro-vide for a low level of consistency amongthe 50 states due to the wide variety of sta-tistical methods that states use to calculateAYP. Porter, et al. (2005) analyzed howthe state of Kentucky was implementingtheir AYP data according to subgroup size,confidence intervals, and the line of tra-jectory. They found that the statemanipulation of those statistical measuresin combination can have a dramatic impacton the number of schools that made AYPin Kentucky. For example, Kentucky holdsschools accountable for subgroup achieve-ment when they have a minimal of 60 inthe subgroup, they use a two-tailed 99 per-cent confidence interval, and they have aback loaded trajectory. With the imple-mentation of those measurementsprocedures, in 2003 the state of Kentucky

had 90 percent of their schools meet theAYP requirements. In 2004, 94 percent ofthe schools met the AYP requirementsunder the same measurement procedures.When the researchers eliminated the use ofthe confidence interval, they determinedthat only 61 percent of the schools in Ken-tucky would have made AYP in 2003 and72 percent would have made it in 2004.When they further created a straight linetrajectory and eliminated the confidenceinterval, the researchers determined thatonly 45 percent of schools would have metAYP requirements in 2003 and 59 percentwould have met AYP requirements in 2004.When they modified the minimum num-ber of students for subgroups to bedisaggregated from 60 to 30, used thestraight line of trajectory, and eliminatedthe confidence interval, they determinedthat 31 percent of schools would have metAYP in 2003 and 44 percent in 2004. Fig-ure 7 illustrates the data findings by Porteret al. with significant changes that occuras they changed the statistical methodol-ogy for the achievement of AYP in the stateof Kentucky for the 2003 and 2004 schoolyears.


Figure 7. AYP Proficient schools in Kentucky with statistical manipulation

Includes the use of the confidence interval, minimal subgroup size of 60, and a back end loaded trajectory.

Includes the use of the minimal subgroup of 60 and the back end loaded trajectory with no confidence interval.

Includes the use of the minimal subgroup size of 60, with a straight line trajectory and no confidence interval.

d Includes a minimal subgroup size of 30, with a straight line trajectory and no confidence interval.

Note: This figure is based on data that was presented in the study by Porter et al. (2005).

A number of implications have surfacedas a result of the inconsistencies in policy(Porter et aL., 2005). The first implicationis that there are different standards that arebeing implemented by different states inorder to measure the AYP accountabilityprovisions in NCLB. The result is littleconsistency between the states when itcomes to setting AYP requirements. Anoth-er implication is that the different choicesstates are making with regards to how theydesign and measure AYP has a dramaticimpact on the number of schools that meetAYP requirements. This includes makingchanges in the confidence interval mea-surement, the number of students in aspecific subgroup, and the type of trajec-tory that a state employs. This means that

it is virtually impossible to perform a faircomparison of the AYP data from state tostate without examining the specific deci-sions that a state has taken while evaluatingtheir schools according to AYP. It is clearthat the statistical designs that states use tomeasure AYP have an enormous impacton the number of schools in a state thatmeet AYP requirements. Since failure tomeet AYP comes with a cost through theimplementation of sanctions that arerequired by NCLB for Title I schools, it isprobable that states would continue tochoose less stringent designs when mea-suring schools according to AYP.

%That Met AYP FiO31~


Reliability Issues with a Single MeasureAccountability System

There is a consensus among researchersin regards to the reliability issues that areassociated with the use of a single mea-surement to evaluate school effectiveness.Several studies (Darling-Hammond,2007a: Darling-Hammond, 2007b; Elmore,2002; Harris, 2007; Hess & Petrilli, 2006;Kane, Douglas, & Geppert, 2002; Linn &Haug, 2002; Sunderman, et al. 2005) callinto question the reliability of the single testmethod in evaluating school effectiveness.Elmore (2002) argues that there are manytechnical difficulties with the reliance ona single test to evaluate school effective-ness and the achievement of statecurriculum standards. Elmore (2002) fur-ther claims that the creation of anaccountability system at the state level wasessentially a political act. He believes thatthe push for state accountability at the locallevel is a risky business for state politi-cians. Since most politicians do not haveexpertise when it comes to high stakesassessments, they are unaware of the reli-ability issues that are involved in using asingle measurement to assess an entireschool or district. Furthermore, Kane et al.(2002) point out a weakness with using asingle year test to measure school effectivesbecause different cohorts of students aremeasured against each other in order todetermine a schools' AYP status. The studyby Kane et al. emphasizes there are dif-ferent variables at play within the cohortgroups which ultimately have an impacton standardized test achievement. Theresult is unreliable data when it comes tomeasuring the effectiveness of schools.

High Flying Schools andthe One Year Phenomena

Past research (Harris, 2007; Kane et al.2002) found there can be fluctuations inannual test scores by schools in a singleyear that have nothing to do with gains inschool performance. The study by Kane etal. (2002) found that the size of the schoolhas statistical significance with the varianceof test scores from year to year. The larg-er schools are more stable and have lessvariance over time. Thus, measuring stu-dents accordingly by only using year toyear gains based on cut scores is statisti-cally unreliable. "With test scores beingso volatile, school personnel are at a sub-stantial risk of being punished or rewardedfor results that are beyond their control"(Kane et al., 2002, p. 60). Linn and Haug(2002) found that many schools that arebeing recognized as outstanding andothers that are labeled as needing improve-ment which has to do with randomfluctuations and not from changes in thequality of education in school performance.

Harris (2007) provides a direct rebut-tal to research reports by the EducationTrust and Heritage Foundation that haveattempted to identify high-flying schools.High-flying schools are defined as thoseschools that help students reach high lev-els of achievement despite high povertylevels or high levels of minority students.Harris evaluated the data that was used bythe Education Trust study to define schoolsthat are labeled as high-flying. He ana-lyzed the high flying definition that wasused by the Education Trust at 1-1 -1. Thismeans that the school needs to achieve ata high level in either year of the study, in


any of two subjects, and in any of twogrades that were measured. He then ana-lyzed this data in a graphic form by makingmore stringent requirements with the useof the classifications 2-1-1, 2-2-1, and the2-2-2. For example, the last definition 2-2-2 means that schools needed to performat a high level for two years in two subjectsand in two grades.

Harris (2007) found the data that wasprovided in the Education Trust study,which defines schools as high-flying, onlytook into account a single year's worth ofdata. When multiple year data sets are usedthat take into account multiple tests at dif-ferent grade levels, Harris found that lowpoverty schools are 22 times more likelyto become high performing when com-pared with the high poverty schools. Healso found that low poverty and low minor-ity schools are 89 times more likely tobecome high performing vs. a high pover-ty and high minority school. The findingsby Harris point out another problem withAYP as there is the possibility of the sta-tistical chance where some schools will dowell on a single assessment for an indi-vidual school year. However, theprobability of achieving that same result formultiple years and over multiple tests isless likely to occur and the chance for sta-tistical error is reduced.

AYP Measures Social Capital,Minority Status, and Racial Subgroups

Literature in the field (Elmore, 2002;Harris, 2007; Wiley et al. 2006; Meier, etal., 2004; Mathis, 2004a; Reeves, 2006)has shown that poverty and minority sta-tus have a major impact on the achievement

levels of students in schools. Elmore (2002)points out that the implementation of AYPis doing nothing more than measuring thesocial capital of students in schools. Elmoreargues that standardized test scores showthe social capital of the students in schools.Most high-performing schools elicit high-er performance by relying on the socialcapital of their students and families ratherthan on the internal capacity of the schoolsthemselves. Most low-performing schoolscannot rely on the social capital of studentsand families and instead must rely on theirorganization capacity. Hence, with little orno investment in capacity, low-perform-ing schools get worse relative tohigh-performing schools.

The argument that Elmore (2002)frames leads to the conclusion that whilethe intent of NCLB is to improve schoolsfor underperforming students; it will actu-ally work in reverse as it broadens the gapbetween the schools with a high socialcapacity and a low social capacity. Thejustification for this argument is the pri-mary focus of the NCLB reform which isthe belief that school accountability mea-sures and sanctions will motivate studentsand schools. Elmore indicates that this phi-losophy is based upon a faulty premise.The accountability sanctions do not provideorganizational capital to the schools thatrequire this capital in order to achieve alevel of success with underperforming stu-dents.

The identification of what Elmore(2002) refers to as social capital is an inter-esting perspective because he argues inorder to improve schools, there is a needto augment the social and organization cap-


ital of the underperforming schools. Hemaintains that the social capital of the stu-dents and their families is what is beingmeasured by the standardized, single mea-surement accountability system. The resultis a need for extensive studies in order toanalyze his hypothesis along with an analy-sis of the possible positive and/or negativeimpact that the NCLB accountability pro-visions are having on public education inthe United States.

Researchers (Mathis, 2004b; Meir, etal. 2004; Wiley et al, 2005) point out thatschools cannot do everything to fix thesocietal problems. Mathis (2004a) arguesthat the schools cannot create other con-ditions in society where the students attendschool with high levels of readiness, moti-vation, a high level of academic ability,and support from the home. He states thatthe use of test scores to close the achieve-ment gap is more of an ideological beliefrather than a belief that is founded inresearch. Mathis (2004a) further statesthat not all students start at the same level.He argues that it is unrealistic to believethat all students will be able to achieve thesame proficiency levels. Mathis refers topersonal capital which is the readiness andability to achieve on tests. In conjunctionwith Elmore (2002), Mathis (2004a) arguesthat factors outside the school known associal capital have an impact on studentachievement. He defines social capital asa combination of the following; parents'education level, the values the familyplaces on education, socioeconomic sta-tus, the peer group influence, and similarassets and liabilities. Mathis (2004a) goeson to say there is a great gap in the edu-

cational equity in schools in the UnitedStates and he cites a study by the Organi-zation for the Economic Cooperation andDevelopment which ranked the UnitedStates twenty-second out of twenty-fiveindustrialized countries when it comes toeducational equality. As schools are labeledas failing, it is possible that the NCLBreform will help to distinguish the dispar-ities among the high poverty students andlow poverty students.

Harris (2007) found that African-Amer-ican kindergarteners are achieving at a rateof 34 percentile points below the levels ofwhite kindergarteners. He believes thatthe current political climate has evolvedin a manner in which school systems areblamed poor levels of student achievement.He states that the politicians are neglect-ing to look at other factors that come intoplay with regards to student achievement.Harris argues that a fault in NCLB is thefailure to account for learning growth bystudents which is in contrast to the solemeasure of cut score proficiency achieve-ment. He asserts that the AYP measurementsystem cannot be solely attributed to learn-ing in schools.

Harris (2007) illustrates his positionwith the following quote:

Consider the foot-race analogy madeby President Lyndon Johnson whenhe argued for affirmative action andcompensatory education. Johnsonsaid that undernourished studentswould lose the vast majority of therunning races, not because other stu-dents or track coach failed to try hardenough, but because the studentswere undernourished. (p. 373)


Figure 8. Schools in Michigan that are projected to make AYP in 2014.

Note. This figure is based on data that was presented in the Wiley et al. study (2005).

This analogy illustrates that the thereare other variables at play when studentscome to school with inadequate social cap-ital. The problem with AYP is that it doesnot take into account the amount of socialcapital or the starting point at which the stu-dents enter school.

Inevitability of Labeling Schools as FailingWiley et al. (2005) found that a large

number of schools will be labeled as fail-ing by the year 2014. Their research isbased upon an examination of six states inthe Great Lakes Regions; 1) Illinois, 2)Indiana, 3) Michigan, 4) Minnesota, 5)Ohio, and 6) Wisconsin. The authors foundthat due to the different requirements andparameters that are mandated by AYP ineach of the six states, each state exhibits adifferent but increasing pattern of havingmore schools not make AYP by the year2014. This study did not take into consid-eration the implications of the subgroupissues or the integration of the science testinto the AYP equation.

Even with the most conservative esti-mate, Wiley, et al. (2005) found that the

number of failing schools is projected tototal well over 50 percent. For example, inMichigan they found that in 2004, 25 per-cent of the schools failed to meet AYP.Michigan uses the stair step trajectorymethod until 2010 when it becomes a back-loaded trajectory. The minimum size fora subgroup is 30 and there was no confi-dence interval that was used by the statein 2005. According to the more conserv-ative growth estimates, Michigan ispredicted to have 3.2 percent of schoolsmake AYP in 2014, while 46.2 percent willbe safe harbor eligible, and 50.6 percentwill not make AYP and will not be safeharbor eligible. With a more aggressive orlow growth prediction, the averages in thesame order change to 0.2 percent, 13.2 per-cent, and 86.6 percent. The sharp declinein schools that will not make AYP is pre-dicted to occur in 2010 in Michigan due tothe line of trajectory. Figure 8 illustratesthe percentage of schools that are project-ed to make AYP and fail to make AYP inthe year 2014 in Michigan.

The findings by Wiley et al. (2005) havemajor implications for the current state of

100.00%

90.0"/A

"o 70.00%

70.00%

60.00%

50cosethe growth 3.2011a 46.20% 50.g0%

*g gresh got h 0.20}% 13.20% 86.600/aGrowth Category


the NCLB accountability system. As theproficiency bar rises, a large majority ofschools are predicted to be labeled as fail-ing by 2014 under the NCLB reform. Whenlooking at the AYP trajectory that leads to2014, a reasonable assumption could bethat the number of schools that fail to makeAYP will increase similar to the predic-tions in this study. Looking at the numberof schools that have currently failed tomake AYP and aligning them with the tra-jectory increase by each state until the year2014 may lead to this conclusion. The find-ings in this study show the only thing thatmay change that projection is revisions inCongress through the reauthorizationprocess. Researchers in the field (CEP,2005; Darling-Hammond, 2007b; Hess &Petrelli, 2006; Lewis, 2006; Linn, 2003;Mathis, 2004a) state that the target of hav-ing 100 percent of all students achieve highstandards is statistically impossible andschools are set up for ultimate failure underNCLB. Mathis (2004a) references the factthat in 2003,87 percent of schools in Flori-da and 57 percent in Delaware were labeledas failing schools. He concludes that anincreasing number of schools will be pun-ished as failing.

Summary of Measurement Concerns with theUse of AYP to Evaluate School Effectiveness

An analysis of the empirical researchand literature in the field has provided aconsensus that there are multiple problemswith the NCLB accountability provisions.Every state has the autonomy to set up theirown AYP proficiency standards and stateadministered tests in reading, mathemat-ics, and science. The result is difficult to

use AYP to accurately compare the quali-ty of schools across the United States. Thestatistical manipulations of the AYP mea-surement systems at the state level providethe public with a false impression of a con-sistent measure of school effectiveness.The research leads to a conclusion thatmodifications to the accountability provi-sions in NCLB will be necessary in orderfor the legislation to effectively improvethe conditions of school programs andclassroom instruction.

AYP Benefits andUnintended Consequences

Philosophical intentThere is a consensus in the field

(Borowski & Sneed, 2006; Darling-Ham-mond, 2007b; Guilfoyle, 2006; Haycock,2006; Hess, 2006; Hess & Petrelli, 2005;Kane, et al. 2002; Lewis, 2006), that themost beneficial part of the NCLB reformis the philosophical premise that all chil-dren will count and no child will be leftbehind. Borowski and Sneed state the fol-lowing:

In our view, NCLB's most impor-tant potential benefits lie in itsrecognition of the fundamentalpremise that all children can learnand have a right to be taught, andthe corollary principal that educa-tors, parents and students shouldreceive periodic assessments of howstudents are progressing toward theattainment of high academic stan-dards. (p. 504)Borowski and Sneed's (2006) opinions

have been molded through the litigation


of desegregation plans for schools in theUnited States. They often found that dis-tricts could not provide disaggregated databy race, students with disabilities, or thenumber of students in honors programs onstate assessments. They assert thataccountability systems have the potentialto have a positive impact on public edu-cation as parents are informed about howschools are achieving on assessments andwithin particular subgroups. Borowski andSneed (2006) state the following:

NCLB's principal strength is itsfocus on improving students' acad-emic performance, and doings so forall children of all backgrounds. Theact requires that all states establishstandards in reading, mathematics,and science and develop assessmentsto determine whether students areproficient in meeting the standards.NCLB requires that all students meetthese standards by the 2013-2014school year. (p.50 6)The philosophy found in NCLB is a

focus on high academic standards and suc-cess for all students in all areas andsubgroups which according to Borowskiand Sneed (2006) have a great potential toprovide educational equity. This potentialis powerful because it mandates improve-ment not just in the aggregate, but also forall subgroups through disaggregated data.Borowski and Sneed argue that the intentof the legislation must go beyond havingthe data disaggregated in order to informclassroom instructional practices with thegoal to make improvements for all studentsin schools.

Consequences of Sanctionsand Accountability

Impact on Subgroup PopulationsHaycock (2006) found that at Centen-

nial Place Elementary School in Atlanta,administrators, and teachers are using dataand information to assess the needs of stu-dents. The school disaggregated the datadue to NCLB and they were able to pin-point a weakness with special educationstudents. She posits that the result wasgrowth in mathematics and reading thatproduced a 30 point gain. She also refer-ences Woodville Elementary in Richmond,Virginia where there has been a 20 pointincrease in students meeting the standardson the state assessment. In order to sup-port her conclusions from the schools thatshe references, Haycock refers to the Hall& Kennedy (2006) study which found thatin elementary reading, 27 of 31 statesshowed improvement and 22 of 29 nar-rowed the gap between African Americanand White students while 24 of 29 nar-rowed gaps between Latino and Whitestudents. In elementary math, 29 of 32states showed improvement, 26 of 30 nar-rowed the gap between African Americanstudents and White students, while 22 of30 narrowed the gap between Latino andwhite students.

Haycock (2006) claims that a 30 pointgrowth was made by a school in Atlanta asa result of the positive impact that NCLBis having on the school. However, thereare numerous variables that could have hadan impact on the achievement of thatschool. Some of the previous literaturehas illustrated the manner in which the


states manipulate their variables like sub-group (N) totals, confidence intervals, andthe proficient cut scores on assessments,which could also have an impact on theannual gains that are referenced.

Poverty and racial diversityResearch in the field (Mathis, 2004a;

Meir, et al. 2004; Sunderman, et al. 2005)shows that the poor and ethnically diverseschools will be punished first through thesanction provisions. Those schools withmore subgroups and higher percentages ofracial groups of students will have a greaterchance of failing to meet AYP standardswhich will result in the diversion of fund-ing from those schools to pay for themandated sanctions. Porter et al. (2005)posit schools with a 25 percent or moreblack student population in Kentucky weremuch more likely to fail AYP at a 25-30point increase. They also discoveredschools with higher percentages of Eco-nomically Disadvantaged (ED) populationswere more likely to fail to meet the AYPrequirements. When they developed ademanding AYP design by eliminating theconfidence interval and using a straight-line trajectory, there were an even greaternumber of schools (73 percent) with a highED population that failed to make AYP.This was in contrast to only 30 percent ofschools failing to meet AYP requirementswith lower ED populations. A study bySpringer (2008) found that schools withlarger Hispanic populations were 8 per-cent more likely to fail AYP. Schools withlarger ED populations were more likely tofail to meet AYP by 9 percent. However,it is difficult to reliably conclude that these

results are transferable to other states sincethere were lower numbers of racial sub-groups and lower numbers of ED studentsin the state that Springer studied when com-pared to other states across the country.

Educational TriageEducational triage involves focusing on

the students who are just below the profi-ciency level in order to accelerate theirachievement so that they meet the statestandards. Those students are often labeledas the "bubble kids". Guilfoyle (2006)argues that due to the NCLB accountabil-ity provisions, schools are focusing on thestudents who are achieving just below thecut line in order to improve their highstakes test scores. Springer (2008) sug-gests that if educational triage is occurringin schools, then the students in the low cat-egories that have little chance of passingalong with those in the high categories thathave already achieved proficiency mightreceive less attention and resources. The"bubble kids" benefit by receivingresources that might have otherwise beenused for the highest or lowest achievingstudents.

The findings by Springer (2008) indi-cate that educational triage is not occurringin schools. Springer analyzed the testscores from over 300,000 students from awestern state in the United States. He usedthe academic achievement data from thefall and spring state test scores in the areaof mathematics. The state that he used forhis analysis has two assessments that aregiven to their students in one academicyear as opposed to the one assessment thatis mandated by NCLB. Springer could not


find any evidence that schools wereinvolved in educational triage with a real-location of their resources to the bubblekids. Some of his findings conclude that thelowest students gained more than thosenear the proficiency line for the schoolsthat failed to make AYP during the previ-ous school year. There was a 0.2 standarddeviation gain vs. the students who werenear the proficiency line. The students inthe highest categories did not have a neg-ative consequence as their achievementwas also in line with the "bubble kids".With the schools that made AYP the pre-vious year, Springer found that the studentsin the lowest category the previous yearshowed the highest gains. He also foundthat proficient students in failing schoolsgained more than proficient students innon-failing schools. These results suggestthat the higher achieving students andlower achieving students are not losing outon educational resources or interventionsfor the benefit of the "bubble kids" so thatthe schools can make AYP. However, sincethere are very small numbers or racial sub-groups in the state that was studied, itwould be difficult to conclude that theseresults were prevalent in other states withlarger populations of racial or ethnic diver-sity. One weakness with the study is thatSpringer only used mathematics test dataand not language arts or reading test scoreswhich is an equally important part of theAYP accountability formula. As a result,the findings can only be attributed to math-ematics instruction in the sample state.

Narrowing the CurriculumScholars posit that NCLB is narrowing

the curriculum as it discourages the usageof instructionally useful forms of assess-ment that involve extensive writing andanalysis (Darling-Hammond, 2007a; Rav-itch, 2010). Some states have been forcedto eliminate the usage of sophisticated per-formance-based assessment systems thatresemble those used in other nations aroundthe world that score well on internationalassessments. Guilfoyle (2006) argues thattesting and measurement is a major empha-sis in NCLB and since mathematics andreading are being tested, there is anunprecedented focus on those two areas.The areas of the curriculum that are notbeing measured by NCLB are suffering asschools start to preclude or reduce theirfocus in those areas. This includes histo-ry, art, civics, music, physical education,health, and other cultural areas (Ravitch,2010). Guilfoyle (2006) argues thatalthough these areas are not being mea-sured, they do have a tremendous impacton the education of students and they areextremely important. Since those areas arenot being tested, students and teachers arenot placing much emphasis on those con-tent areas. Guilfoyle (2006) also indicatesthat standardized assessments provide onlya limited amount of information. The test-ing provisions in NCLB are unable toprovide information about prevalentinstructional improvement. But, empiricaldata was not presented in order to deter-mine if there actually is a narrowing of thecurriculum. However, since NCLB mea-sures growth in mathematics and reading,it is probable that schools would place more


emphasis on those subjects while reducingthe other areas of the curricula. Sunder-man et al. (2005) collected qualitativesurvey data from teachers in two schooldistricts Fresno, California and RichmondCounty, Virginia. They found that teach-ers acknowledged that there was anarrowing of the curriculum as they indi-cated that more focus was given to the testareas in contrast to those areas that werenot tested and could be labeled as a formof curricular triage.

Focus on Basic Skills vs. Higher LevelThinking and Innovation

A strategy that educators could use toimprove tests scores on the NCLB account-ability measurements is to teach to the testwhere the sole focus is on basic skills andtest taking methods. However, teaching tothe test will not improve the schools over-all educational program in order to preparestudents for society. Schoen and Fusarel-li (2008) along with Guilfoyle (2006) arguethat there is a need to create a system thatrewards educators for innovations inschools instead of the current system thatplaces the utmost importance on masteryof core content and basic skills. Schoenand Fusarelli (2008) believe that the con-structivist approach that supports theacquisition of the skills that are needed forstudents to become successful in the 21stcentury are in competition with theaccountability provisions in NCLB thatemphasize content based learning and theneed to do well on a single measure.Schoen and Fusarelli (2008) along withDarling-Hammond (2007a) claim thatthere is a need to change the current assess-

ment system that is supported by NCLB.They believe that new assessments must bedesigned through the reauthorization ofNCLB. Those assessments should mea-sure higher order critical thinking skills,problem solving, and the ability to processinformation in a relevant manner. TheNCLB legislation has narrowed the cur-riculum and essentially made it difficultfor schools to promote innovation andmove towards the ideals of the 21st cen-tury schools movement (Ravitch, 2010;Darling-Hammond, 2007a; Darling-Ham-mond, 2007b; Meir et al., 2004; Schoen &Fusarelli, 2008).

Recommendations for AYP reform

The Need for Empirical Evidencethat Supports the Use of the AYPAccountability System

The intent of the NCLB reform is toimprove classroom instruction and the edu-cational programs for disadvantagedstudents. The consensus found in the lit-erature suggests that policies aligned withthe AYP provisions in NCLB are not com-pletely founded in empirical research giventhe goals of the reform. They were creat-ed on a political or ideological basis insteadof a scientific research basis. For exam-ple, many of the findings and conclusionsthat have been presented in this review(Borowski & Sneed, 2006; Elmore, 2002;Harris, 2007; Kane, et al. 2002; Lewis,2006; Sunderman, et al. 2005) indicate thatusing a single once a year high stakes testfor an accountability measure is unreliablein evaluating school effectiveness. Sun-derman et al. (2005) found there is a lack


of empirical data, which shows that theschool of choice provisions or supple-mental service provisions will produceimprovements for students. They point tothe fact that the sanctions might actuallyweaken the conditions of students inschools as funds are diverted.

Elmore (2002) states that if there is alack of support given to the internal capac-ity of schools to make improvements thatthe sanction provisions will have littleimpact on improving classroom instruc-tion as schools are labeled as failingaccording to AYP. Schoen and Fusarelli(2008) interviewed educators and theyfound that the accountability sanctions alsocreated pressure on principals who con-sidered the possibility of leaving schoolsdue to sanction provisions. Marzano,Waters, and McNulty (2005) found thatschool leadership has a substantial impacton student achievement. If high qualityprincipals leave schools due to the pres-sures of probation, then it is probable thatthose schools will have less of a chance ofmaking improvements.

Borowski and Sneed (2006) believe thatthe accountability provisions and sanctionsin the NCLB legislation are unproven andlack scientifically backed data that sup-ports the implementation of thoseprovisions. In their estimation, the legis-lation is a bandage job that attempts tobalance between implementing federalintervention and allowing for local andstate control of education. Thus, theyobserve the potential for grave harm asstates have at times lowered standards ormanipulated their statistical data in orderto avoid the costly penalties and sanctions.

Lewis (2006) argues there are major prob-lems with the current model ofaccountability known as AYP and believesit needs to be eliminated and a new modelshould be grounded in accountabilityresearch that will help to improve the con-ditions of schools.

The Use of Growth Data for AYPA large number of empirical research

studies and articles (Elmore, 2002; Hess &Petrilli, 2006; Koretz, 2008; Peterson &West, 2006; Popham, 2005b; Schoen &Fusarelli, 2008; Sunderman et al. 2005;Wiley et al. 2005) recommend the use ofgrowth data in the AYP formula. Researchby Sunderman et al. (2005) found that thestate accountability systems that weredeveloped in each of the states were gen-erally in contrast to the accountabilityprovisions in NCLB. Most of the statesused a form of growth data in order to eval-uate the schools in their states. They didnot solely rely on the absolute cut scoremethod that is prevalent in NCLB. Thisoften creates confusion for school stake-holders. For example, the schools inChicago were subject to three account-ability provisions at the national, state, andlocal levels - all of which could create dif-ferent results from the measures that wereimplemented in each system.

Peterson & West (2006) used the Flori-da Comprehensive Achievement Tests(FACT) for their study for the 2002-2003and 2003-2004 school years. They foundthat schools making AYP had a standarddeviation that was 9 percent above theamount gained by students at schools notmaking AYP. This standard deviation sta-


tistic was stated to be equivalent to 1/3 ofa school year of student growth. Theachievement in reading was a 7 percentstandard deviation higher in the schoolsthat made AYP when compared to thoseschools that did not make AYP. In schoolsthat made AYP, the standard deviation was11 percent higher for African Americansand 12 percent higher for Hispanic stu-dents when compared to the samesubgroups of students in schools that didnot make AYP. There was a 6 percent dif-ference of a standard deviation in the areaof reading for both of those ethnic groups.When they completed a statistical analy-sis to take into account for the ELL, specialeducation, socioeconomic status, ethnici-ty, and mobility rates, the average was 4percent of a standard deviation in mathe-matics and 2 percent of a standard deviationin reading for schools that met AYP vs.those schools that did not meet AYP.

Schools in Florida that made AYP out-performed the schools that did not makeAYP 71 percent of the time according tothe analysis of student growth data on themathematics assessments. This resulted ina 28 percent error rate. Peterson and West(2006) disaggregated the data to furtheranalyze the differences according to growthdata in relation to the Florida system thatevaluates the schools on a 5 point scale Athrough F. While the learning gains wereat a 7percent standard deviation higher inschools that scored A versus those thatscored F, the gains between the A schoolsand the F schools was 25 percent of a stan-dard deviation or the equivalent of oneschool year of growth. In 2004, only 47of the states 2649 schools were given an

F and 184 were given a grade of and E.This is in contrast to a total of 75 percentof the schools that did not make AYPincluding more than half of the schoolsthat received an A grade.

The findings by Peterson and West(2006) have important implications as thereauthorization of NCLB is being debat-ed in Congress. The study shows that theAYP data does not have the ability to mea-sure the growth that is occurring in schools.It brings into question the true goals inNCLB. For example, would a parent wantto send their child to a school that is hav-ing more success with student gains vs.those that appear to be making the cutscores in NCLB? The study points out thatthe accountability system in Florida andthe NCLB accountability system createsconfusion for Florida residents. Half ofthe schools that scored an A on the Flori-da accountability system are labeled asfailing under the NCLB accountability sys-tem. The data further suggests that AYPhas a large error rate when labeling schoolsas successful in accordance with the growthdata.

A Callfor Multiple MeasuresThe use of a single measure with the

evaluation of school effectiveness is not areliable source of measurement. Severalresearchers (Darling-Hammond, 2007a;Elmore, 2002; Guilfoyle, 2006; Kane etal. 2002; Scheon & Fusarelli, 2008; Wileyet al. 2005) support the use of multiplemeasures in order to evaluate school effec-tiveness. Kane et al. (2002) examined theimpact of the use of a single measure toevaluate school effectiveness in their study.


They used data from California to showthat schools with racially homogeneouspopulations have a better statistical chanceof winning performance awards under thesubgroup accountability provisions in thatstate. They point out that schools withlarger racial subgroup populations have agreater chance of failure and that the NCLBreform might provide an incentive forstates to lower standards so that more stu-dents meet the proficiency levels. Theauthors identify four different types ofaccountability systems that were usedacross the country prior to NCLB. Stateslike Arizona, North Carolina, and Ten-nessee used a growth model. States suchas Texas and Illinois reverted to a crosscohort comparison where different cohortsof students were compared from one yearto the next. Other states like Californiarated their schools based on change in testscores from one year to another. Somestates used accountability methods thatlooked at including a combination ofgrowth and proficiency cut scores fromyear to year.

Researchers (Darling-Hammond,2007a; Elmore, 2002; Guilfoyle, 2006;Scheon & Fusarelli, 2008) further arguethat authentic learning projects and port-folios should be emphasized at a higher orequal level in NCLB. An emphasis onthose types of assessments with multiplemeasures would promote higher levelthinking and critical problem solvingapproaches to education. Another form ofmeasurement that could be used for theevaluation of school effectiveness is aprocess that has been implemented by theAdvanced Education Organization, for-

merly known as North Central Accredita-tion (NCA). This process involves theevaluation of school improvement plans,the inclusion of growth data, the use ofmultiple forms of achievement data, andthe reliance on school visits to measureschool effectiveness to determine accred-itation status. While this measurementwould be costly, it would be beneficial toexplore this option as a part of the AYPmeasure with regards to school effective-ness.

National Curriculum and

the Use of the NAEPAYP provisions are not consistent from

state to state as different states have dif-ferent cut scores and their trajectories areoften much different (Sunderman et al.,2005). Research shows that even whenone state might have higher standards whencompared to another state based on theproficiency levels, the proficiency targetscould be deceiving because the quality ofthe assessments used might be much dif-ferent in each state. Borowski and Sneed(2006) argue that there is a need for nation-al standards that protect the local rightsand state rights for implementing educa-tional policies and that minimal nationalstandards are necessary. They emphasizethat the arbitrary accountability provisionsthroughout the AYP process must changein order to equalize the accountability stan-dards among the states in a fair manner.Hoxby (2005) believes there is need to cre-ate a national benchmark system in orderto effectively evaluate school progressacross the country. Since the legal author-ity to implement education falls to state


governments, it is predicted to be difficultin getting approval in Congress and con-sensus among the states for a nationalcurriculum standard that is used in allstates. It is also probable that there wouldbe court challenges to any form of nation-al curriculum that is mandated by thefederal government. However, as Ravitch(2010) points out, we should establish anational curriculum that "declares ourintention to educate all children in the fullrange of liberal arts and sciences, as wellas physical education" (p, 231).

What We Still Need to KnowThe current accountability provisions

in NCLB have not proven effective in eval-uating school performance. It is apparentfrom the literature that there are numerousproblems with the implementation of AYPin order to motivate effective schoolimprovement efforts. Reforms to the leg-islation are needed in order to improve thereliability of measuring school effective-ness. This includes the need to standardizethe AYP measure across the country whileimplementing forms of measurements,which include growth data. Since the keyto school improvement is through theenhancement of the internal capacity ofthe school while increasing the internalaccountability mechanisms, it is criticalthat research is conducted in this area.Research is needed regarding the impactthat AYP is having on student learning,classroom instruction, and school improve-ment. Future work should examine theimpact AYP is having through other mea-sures, including qualitative data analysisthat examines how teachers and adminis-

trators are responding to AYP. Moreover,research is needed to address the questionof whether a school will improve if thestaff is replaced? Will a school get betterif it is turned into a charter school? Will itimprove in measureable ways if the statetakes it over? What policies could be imple-mented to attract and retain high-qualityteachers in low performing urban schools?These and a good many other questionsremain to be answered within the contextof NCLB.

ReferencesAbdelmann, C. E., R., Even, J., Kenyon, S., &

Marshall, J. (1999). When accountabilityknocks, will anyone answer? (No. CPRE-RR-42). Philadelphia, PA: Consortium for PolicyResearch in Education.

Abedi, J., Hofstetter, C. H., & Lord, C. (2004).Assessment accommodations for english lan-guage learners: Implications for policy-basedempirical research. Review of EducationalResearch, 74(1), 1-28.

Borowski, J. W., & Sneed, M. (2006). Will NCLBimprove or harm public education? HarvardEducational Review, 76(4), 457-460.

Brookover, W. B. (1985). Can we make schoolseffective for minority students? Journal ofNegro Education, 54(3), 257-268.

Center on Education Policy (2007). What now?:Lessons from Michigan about restructuringschools and next steps under NCLB. Washing-ton, D.C.: CEP.

Coleman, J. S., & et al. (1966). Equality of edu-cational opportunity.

Cronin, J., Dahlin, M., Adkins, D., & Kingsbury,G. G. (2007). The proficiency illusion: ThomasR. Fordham Institute and Northwest Evalua-tion Association.

Darling-Hammond, L. (2007a). Evaluating "Nochild left behind". The Nation(May 21, 2007).


Darling-Hammond, L. (2007b). Race, inequalityand educational accountability: The irony of"No child left behind". Race, and Education,10(3), 245-260.

Edmonds, R. (1979a). A discussion of the litera-ture and issues related to effective schooling.

Edmonds, R. (1979b). Effective schools for theurban poor. Educational Leadership, 37(1), 15-24.

Edmonds, R. (1982). Programs of effectiveschools:An overview: Universities, state agen-cies and school districts have establishedschool improvement programs based on effec-tive school research. Educational Leadership,40(3), 4-11.

Elmore, R. F. (2002). Unwarranted intrusion. Edu-cation Next, 2(1), 30-35.

Fullan, M., Hill, P., & Crevola, C. (2006). Break-through. Thousand Oaks, CA Corwin Press.

Guilfoyle, C. (2006). NCLB: Is there life beyondtesting? Educational Leadership, 64(3), 8-13.

Hakuta, K., Butler, Y. G., & Witt, D. (2000). Howlong does it take for english learners to attainproficiency? Santa Barbara, CA: CaliforniaUniversity.

Harris, D. N. (2007). High-flying schools, studentdisadvantage, and the logic of NCLB. Ameri-can Journal of Education, 113(3), 367-394.

Haycock, K. (2006). No more invisible kids. Edu-cational Leadership, 64(3), 38-42.

Hess, F. M. (2006). Accountability without angst?Public opinion and no child left behind. Har-vard Educational Review, 76(4), 587-610.

Hess, F. M., & Petrilli, M. J. (2006). No child leftbehind New York, NY: Peter Lang Publishing.

Hoxby, C. M. (2005). Inadequate yearly progress:Unlocking the secrets of NCLB. EducationNext, 2005(3), 46-5 1.

Kane, T. J., Douglas, 0. S., & Geppert, J. (2002).Randomly accountable. Education Next, 2(1),57-68.

Koretz, D. (2008). A measured approach: Value-added models are a promising improvement,but no one measure can evaluate teacher per-formance. American Educator, 32(3), 18-27 &39.

Lewis, A. C. (2006). Clean up the test mess. PhiDelta Kappan, 87(9), 643-645.

Lezotte, L. W. (1992). Learn from effectiveschools. Social Policy, 22(3), 34-36.

Lezotte, L. W., & Jacoby, B. C. (1991). Effectiveschools practices that work.

Linn, R. L., & Haug, C. (2002). Stability of school-building accountability scores and gains.Educational Evaluation and Policy Analysis,24(1), 29-36.

Linn, R. L. (2003). Accountability: Responsibilityand reasonable expectations. EducationalResearcher, 32(7), 3-13.

Marzano, R. J. (2003). What works in schools:Translating research into action. Alexandira,VA: Association for Supervision and Curricu-lum Development.

Marzano, R. J., Waters, T., & McNulty, B. A.(2005). School leadership that works: Fromresearch to results. Alexandria, VA: Associa-tion for Supervision and CurriculumDevelopment.

Mathis, W. J. (2004a). NCLB and high-stakesaccountability: A cure? Or a symptom of thedisease? NCLB: Failed schools-or failedlaw? Educational Horizons, 82(2), 143-152.

Mathis, W. J. (2004b). No child left behind andthe insomnia plague. Clearing House, 78(1),30.

Meier, D., Kohn, A., Darling-Hammond, L., &Sizer, T. R. (2004). Many children left behind:How the no child left behind act is damagingour children and our schools. Boston, MA:Beacon Press.

Michigan Department of Education (2008a).Michigan to pilot growth model program fordetermining adequate yearly progress (AYP).

Michigan Department of Education (2008b).What's new in the 2008 Michigan school reportcards. Lansing, MI.


Mintrop, H. (2004). Schools on probation. NewYork, NY: Teachers College Press.

Olson, L., & Jacobson, L. (2006). Analysis findsminority NCLB scores widely excluded. Edu-cation Week, 25(33), 5.

Peterson, P. E., & West, M. R. (2006). Is yourchild's school effective?: Don't rely on NCLBto tell you. Education Next, 2006(4), 76-80.

Popham, W. J. (2005a). AYP wriggle roomrunning out. Educational Leadership, 63(1),85-87.

Popham, W. J. (2005b). Can growth ever be besidethe point? Educational Leadership, 63(3), 83-84.

Porter, A. C., Linn, L., & Trimble, C. S. (2005).The effects of state decisions about NCLB ade-quate yearly progress targets. EducationalMeasurement: Issues and Practice, 24(4), 32-39.

Pray, L. C., & MacSwan, J. (2002). Different ques-tion, same answer: How long does it take forenglish learning to acquire proficiency? .Paper presented at the American EducationalResearch Association.

Ravitch, D. (2010). The Death and Life of theGreat American School System: How Testingand Choice Are Undermining Education. NewYork, NY: Basic Books.

Reeves, D. B. (2004). Accountability for learning:How teachers and school leaders can takecharge. Alexandria, VA: Association forSupervison and Curriculum Development.

Springer, M. G. (2008). Accountability incentives:Do schools practice educational triage? Edu-cation Next, 8(1), 74-79.

Sunderman, G. L., Kim, J. S., & Orfield, G. (2005).NCLB meets school realities: Lessons from thefield. Thousand Oaks, CA: Corwin Press.

Sunderman, G. L., Tracey, C. A., Kim, J., &Orfield, 0. (2004). Listening to teachers:Classroom realities and no child left behind.Cambridge, MA: Harvard University.

U.S. Department of Education, N. C. f. E. S.(2003). NAEP validity studies: An agenda forNAEP validity research. Washington, D. C.

United States Department of Education (2002). NoChild Left Behind Act of 2001, P.L. 107-110.

Wiley, E. W., Mathis, W. J., & Garcia, D. R.(2005). The impact of the adequate yearlyprogress requirement of the federal "No childleft behind" Act on schools in the great lakesregion. Tempe, Arizona: Arizona State Univer-sity.

COPYRIGHT INFORMATION

Author:

Title:

Source:

ISSN:

Publisher:

Maleyko, Glenn; Gawlik, Marytza A.

No Child Left Behind: what we know and what we need to know

Education 131 no3 Spr 2011 p. 600-24

0013-1172

Project Innovation of Mobile

Box 8508, Spring Hill Station, Mobile, AL 36608

The magazine publisher is the copyright holder of this article and it is reproducedwith permission. Further reproduction of this article in violation of the copyright isprohibited.

This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub- licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Date post:	21-Jul-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

NO CHILD LEFT BEHIND: WHAT WE KNOW AND WHAT WE NEED TO KNOW · 2011-05-22 · reform aims to hold...

Documents