Dropout Prevention_what Corks Clearing House

7/29/2019 Dropout Prevention_what Corks Clearing House

1/72

Dropout PreventionDropout Prevention

IES PRACTICE GUIDE

NCEE 2008-4025

U.S. DEPARTMENT OF EDUCATION

WHAT WORKS CLEARINGHOUSE


2/72

The Institute o Education Sciences (IES) publishes practice guides in educationto bring the best available evidence and expertise to bear on the types o systemicchallenges that cannot currently be addressed by single interventions or programs.Authors o practice guides seldom conduct the types o systematic literature searchesthat are the backbone o a meta-analysis, although they take advantage o such work

when it is already published. Instead, authors use their expertise to identiy themost important research with respect to their recommendations, augmented by asearch o recent publications to ensure that research citations are up-to-date.

Unique to IES-sponsored practice guides is that they are subjected to rigorous exter-nal peer review through the same oce that is responsible or independent reviewo other IES publications. A critical task or peer reviewers o a practice guide is todetermine whether the evidence cited in support o particular recommendationsis up-to-date and that studies o similar or better quality that point in a dierentdirection have not been ignored. Because practice guides depend on the expertiseo their authors and their group decision-making, the content o a practice guide is

not and should not be viewed as a set o recommendations that in every case de-pends on and ows inevitably rom scientifc research.

The goal o this practice guide is to ormulate specifc and coherent evidence-basedrecommendations or use by educators addressing the challenge o reducing drop-ping out, a challenge that lacks developed or evaluated packaged approaches. Theguide provides practical, clear inormation on critical topics related to dropout pre-vention and is based on the best available evidence as judged by the review panel.Recommendations presented in this guide should not be construed to imply thatno urther research is warranted on the eectiveness o particular strategies ordropout prevention.


3/72

IES PRACTICE GUIDE

Dropout

Prevention

September 2008

PanelMark Dynarski (Chair)

MATHEMATICA POLICY RESEARCH

Linda Clarke

CITYOF HOUSTON

Brian Cobb

COLORADO STATE UNIVERSITY

Jeremy Finn

STATE UNIVERSITYOF NEW YORKBUFFALO

Russell Rumberger

UNIVERSITYOF CALIFORNIASANTA BARBARA

Jay Smink

NATIONAL DROPOUT PREVENTION CENTER/NETWORK

StafKristin Hallgren


Brian Gill


NCEE 2008-4025

U.S. DEPARTMENT OF EDUCATION

f


4/72

This report was prepared or the National Center or Education Evaluation and RegionalAssistance, Institute o Education Sciences under Contract ED-02-CO-0022.

DisclaimerThe opinions and positions expressed in this practice guide are the authors and do not

necessarily represent the opinions and positions o the Institute o Education Sciencesor the U.S. Department o Education. This practice guide should be reviewed and ap-plied according to the specifc needs o the educators and education agency using it,and with ull realization that it represents the judgments o the review panel regard-ing what constitutes sensible practice, based on the research that was available at thetime o publication. This practice guide should be used as a tool to assist in decision-making rather than as a cookbook. Any reerences within the document to specifceducation products are illustrative and do not imply endorsement o these products tothe exclusion o other products that are not reerenced.

U.S. Department o EducationMargaret SpellingsSecretary

Institute o Education SciencesGrover J. WhitehurstDirector

National Center or Education Evaluation and Regional AssistancePhoebe CottinghamCommissioner

September 2008

This report is in the public domain. While permission to reprint this publication is notnecessary, the citation should be:

Dynarski, M., Clarke, L., Cobb, B., Finn, J., Rumberger, R., and Smink, J. (2008). DropoutPrevention: A Practice Guide (NCEE 20084025). Washington, DC: National Center orEducation Evaluation and Regional Assistance, Institute o Education Sciences, U.S.Department o Education. Retrieved rom http://ies.ed.gov/ncee/wwc.

This report is available on the IES Web site at http://ies.ed.gov/ncee/wwc.

Alternative FormatsOn request, this publication can be made available in alternative ormats, such as Braille,large print, audiotape, or computer diskette. For more inormation, call the AlternativeFormat Center at 2022058113.


5/72

( iii )

Dropout Prevention

Contents

Introduction 1Te Wat Wors Cearingouse standards and teir reevance to tis guide 3

Dropout Prevention 4

Overview 4

Scope o te Practice Guide 8

Cecist or carrying out te recommendations 10

Recommendation 1. Utiize data systems tat support a reaistic diagnosis o tenumber o students wo drop out and tat ep identiy individua students at igris o dropping out (diagnostic). 12

Recommendation 2. Assign adut advocates to students at ris o dropping out(targeted intervention). 17

Recommendation 3. Provide academic support and enricment to improveacademic perormance (targeted intervention). 22

Recommendation 4. Impement programs to improve students cassroom beavior

and socia sis (targeted intervention). 26

Recommendation 5. Personaize te earning environment and instructiona process(scoowide intervention). 30

Recommendation 6. Provide rigorous and reevant instruction to better engagestudents in earning and provide te sis needed to graduate and to serve temater tey eave scoo (scoowide intervention). 34

Concusion 39

Appendix A. Postscript rom te Institute o Education Sciences 40

Appendix B. About te autors 43

Appendix C. Discosure o potentia conicts o interest 45

Appendix D. Tecnica inormation on te studies 46

Reerences 59


6/72

( iv )

DROPOUT PREVENTION

List o tables

1. Institute o Education Sciences eves o evidence or practice guides 2

2. Recommendations and corresponding eves o evidence to support eac 6

D1. Intensity o components o interventions reated to recommendations 47

D2. Summary: Recommendation 2 53






7/72

( 1 )

Introduction

This guide is intended to be useul to edu-

cators in high schools and middle schools,

to superintendents and school boards, and

to state policymakers in planning and ex-ecuting dropout prevention strategies. The

target audience includes school adminis-

trators as well as district-level adminis-

trators. This guide seeks to help them de-

velop practice and policy alternatives or

implementation. The guide includes spe-

cifc recommendations and indicates the

quality o the evidence that supports these

recommendations. In addition, we have

provided a description o some ways each

recommendation could be carried out. Our

examples should not be construed as the

best or most eective ways to carry out

each recommendation. Rather, the exam-

ples illustrate practices that were noted by

previously implemented dropout preven-

tion programs as having had an impact on

staying in school, progressing in school,

or completing school. Readers need to

note that the specifc ways in which the

practices were implemented varied widely

based on each schools context.

We, the authors, are a small group with

expertise in various dimensions o this

topic. Several o us are also experts in

research methodology. The evidence we

considered in developing this document

ranges rom experimental evaluations o

dropout prevention programs to expert

analyses o dropout prevention practices.

For questions about what works best, high-

quality experimental and quasi-experi-

mental studiessuch as those meeting thecriteria o the What Works Clearinghouse

(http://www.whatworks.ed.gov)have a

privileged position. In all cases, we pay

particular attention to patterns o fndings

that are replicated across studies.

The process or deriving the recommenda-

tions began by collecting and examining

research studies that have evaluated the im-

pacts o dropout prevention programs. This

review relied heavily, but not exclusively, on

the existing reviews o dropout prevention

programs that meet the evidence standards

o the What Works Clearinghouse (WWC).

Dropout prevention interventions almostalways include multiple components. This

bundling o components presents chal-

lenges when reviewing levels o evidence

or each recommendation because evi-

dence o the impact o specifc interven-

tion components on dropping out cannot

ormally be attributed to one component

o an intervention. Identifcation o key

components o each intervention thereore

necessarily relied, to a signifcant degree,

on the panels expert judgment.

Following the identifcation o key com-

ponents o individual interventions, the

interventions and key components were

placed in a working matrix that acilitated

the identiication o eatures that were

common to multiple interventions and

thereore logical candidates as generally

successul practices.

The panel determined the level o evidence

or each recommendation by consideringthe eects o the intervention as deter-

mined by the WWC (table 1), the intensity

o each component toward the impacts

ound in the evaluation, and the number

o evaluations conducted or interventions

that included the component.

Strongreers to consistent and generaliz-

able evidence that a dropout prevention

programs causes better outcomes.1

Moderate reers either to evidence rom

studies that allow strong causal conclu-

sions but cannot be generalized with assur-

ance to the population on which a recom-

mendation is ocused (perhaps because the

1. Following WWC guidelines, we consider a

positive, statistically signifcant eect or large

(greater than 0.25) eect size as an indicator

o positive eects.


8/72

INTRODUCTION

( 2 )

Table 1. Institute o Education Sciences levels o evidence or practice guides

Strong

In general, characterization o the evidence or a recommendation as strong requires both studies with

high internal validity (i.e., studies whose designs can support causal conclusions) and studies with high

external validity (i.e., studies that in total include enough o the range o participants and settings on

which the recommendation is ocused to support the conclusion that the results can be generalized tothose participants and settings). Strong evidence or this practice guide is operationalized as:

A systematic review o research that generally meets the standards o the What Works Clearing-

house (WWC) (see http://ies.ed.gov/ncee/wwc/) and supports the eectiveness o a program, prac-

tice, or approach, with no contradictory evidence o similar quality; OR

Several well-designed, randomized controlled trials or well-designed quasi-experiments that gen-

erally meet the standards o the WWC and support the eectiveness o a program, practice, or

approach, with no contradictory evidence o similar quality; OR

One large, well-designed, randomized controlled, multisite trial that meets the WWC standards

and supports the eectiveness o a program, practice, or approach, with no contradictory evi-

dence o similar quality; OR

For assessments, evidence o reliability and validity that meets the Standards or Educational and

Psychological Testing.

a

Moderate

In general, characterization o the evidence or a recommendation as moderate requires studies with

high internal validity but moderate external validity, or studies with high external validity but mod-

erate internal validity. In other words, moderate evidence is derived rom studies that support strong

causal conclusions but where generalization is uncertain, or studies that support the generality o a

relationship but where the causality is uncertain. Moderate evidence or this practice guide is opera-

tionalized as:

Experiments or quasi-experiments generally meeting the WWC standards and supporting the e-

ectiveness o a program, practice, or approach with small sample sizes and/or other conditions

o implementation or analysis that limit generalizability and no contrary evidence; OR

Comparison group studies that do not demonstrate equivalence o groups at pretest and there-

ore do not meet the WWC standards but that (a) consistently show enhanced outcomes or par-

ticipants experiencing a particular program, practice, or approach and (b) have no major awsrelated to internal validity other than lack o demonstrated equivalence at pretest (e.g., only one

teacher or one class per condition, unequal amounts o instructional time, highly biased outcome

measures); OR

Correlational research with strong statistical controls or selection bias and or discerning inu-

ence o endogenous actors and no contrary evidence; OR

For assessments, evidence o reliability that meets the Standards or Educational and Psychological

Testingb but with evidence o validity rom samples not adequately representative o the popula-

tion on which the recommendation is ocused.

Low

In general, characterization o the evidence or a recommendation as low means that the recom-

mendation is based on expert opinion derived rom strong fndings or theories in related areas

and/or expert opinion buttressed by direct evidence that does not rise to the moderate or strong

levels. Low evidence is operationalized as evidence not meeting the standards or the moderateor high level.

a. American Educational Research Association, American Psychological Association, and National Council on Measure-

ment in Education (1999).

b. Ibid.


9/72

INTRODUCTION

( 3 )

fndings have not been widely replicated) or

to evidence rom studies that are generaliz-

able but have more causal ambiguity than

oered by experimental designs (e.g., statis-

tical models o correlational data or group

comparison designs or which equivalenceo the groups at pretest is uncertain).

Lowreers to expert opinion based on rea-

sonable extrapolations rom research and

theory on other topics and evidence rom

studies that do not meet the standards or

moderate or strong evidence.

The What Works Clearinghousestandards and their relevance tothis guide

The panel relied on WWC evidence stan-

dards to assess the quality o evidence

supporting educational programs and

practices. The WWC addresses evidence

or the causal validity o instructional

programs and practices according to WWC

standards. Inormation about these stan-

dards is available at http://ies.ed.gov/

ncee/wwc. The technical quality o each

study is rated and placed into one o three

categories:

Meets Evidence Standardsor random-

ized controlled trials and regression

discontinuity studies that provide the

strongest evidence o causal validity.

Meets Evidence Standards with Res-

ervationsor all quasi-experimental

studies with no design aws and ran-

domized controlled trials that have

problems with randomization, attri-

tion, or disruption.

Does Not Meet Evidence Screensor

studies that do not provide strong evi-

dence o causal validity.

Following the recommendations and sug-

gestions or carrying out the recommen-

dations, appendix D presents more in-

ormation on the research evidence rom

the WWC-rated evaluations to support the

recommendation.

We appreciate the eorts o Kristin Hall-

gren and Brian Gill, MPR sta members

who participated in the panel meetings,

characterized the research fndings, and

drated the guide. Kristin Hallgren had pri-

mary responsibility or drating the guide

and revising it. We also thank Duncan

Chaplin or helpul eedback and reviews

o earlier versions o this guide.

Mark DynarskiBrian Cobb

Linda Clarke

Jeremy Finn

Russell Rumberger

Jay Smink


10/72

( 4 )

Dropout Prevention

Overview

Each year more than hal a million young

people drop out o high school, and therate at which they drop out has remained

about the same or the last 30 years, even

as spending on education has increased

signifcantly.1

For society as a whole, helping young peo-

ple stay in and complete high school is a

worthwhile objective. Dropouts typically

earn less than graduates: the average earn-

ings dierence is estimated to be $9,000

a year and $260,000 over the course o a

lietime.2 The economic consequences o

dropping out may continue to worsen as

jobs or low-skilled workers dry up.3 Drop-

outs contribute only about hal as much in

taxes as do high school graduates.4 They

draw larger government subsidies in the

orm o ood stamps, housing assistance,

and welare payments.5 They have a dra-

matically increased chance o landing in

prison, and they have worse health out-

comes and lower lie expectancies.6

This guide presents a series o six recom-

mendations or reducing dropout rates

(table 2). The recommendations are di-

vided into three categories: (i) diagnostic

processes or identiying student-level and

schoolwide dropout problems; (ii) targeted

interventions or a subset o middle and

high school students who are identifed as

at risk o dropping out; and (iii) schoolwide

reorms designed to enhance engagement

1. Heckman and LaFontaine (2007); Warren

and Halpern-Manners (2007).

2. U.S. Bureau o the Census (2006); Rouse

(2007).

3. Carnevale and Desrochers (2003).

4. Rouse (2005).

5. Waldogel, Garfnkel, and Kelly (2007).

6. Moretti (2007); Muennig (2007).

or all students and prevent dropout more

generally (see table 2).

The frst recommendation advises schools

and districts to utilize data systems that

support a realistic diagnosis o the num-ber o students who drop out and that help

identiy individual students at high risk o

dropping out. This recommendation is es-

sential or diagnosing the extent to which

schools will need to implement strategies

to address dropping out. In addition, the

implementation o any o the subsequent

recommendations will involve continually

returning to the individual student data to

monitor the success o the strategy and to

adjust approaches as needed.

The panel viewed increasing student en-

gagement as critical to preventing drop-

ping out. Engagement involves active par-

ticipation in learning and schoolwork as

well as in the social lie o school. While

dropping out typically occurs during high

school, the disengagement process may

begin much earlier and include academic,

social, and behavioral components.7 The

trajectory o a young person progressing

in school begins in elementary grades,where students establish an interest in

school and the academic and behavioral

skills necessary to successully proceed.

During the middle school years, students

interest in school and academic skills may

begin to lag, so that by the time students

transition to high school, students who are

at risk o dropping out may need intensive

individual support or other supports to

re-engage them in the purpose o educa-

tion. Educators and policymakers needto consider how to implement intermedi-

ate strategies aimed at increasing student

engagement.

Engagement includes both behavioral

and psychological components. Atten-

dance, class participation, eort in doing

7. National Research Council (2004); Finn

(1989).


11/72

OVERVIEW

( 5 )

schoolwork, and avoidance o disciplin-

ary actions (notably suspensions) are be-

havioral indicators o engagement, while

interest and enthusiasm, a sense o be-

longing, and identifcation with the school

constitutes psychological engagement.8

Both aspects o engagement have been

associated with dropping out o school.9

Attendance in school activities and eeling

a sense o belonging in the school commu-

nity are both critical components o school

engagement and should be addressed as

part o dropout prevention or intervention

strategies.

Recommendations two, three, and our

suggest targeting students who are the

most at risk o dropping out by intensively

intervening in their academic, social, and

personal lives. Many students do not re-

quire special attention to prevent them

rom dropping out. Successul identifca-

tion o the students who are in act at risk

can permit the implementation o inten-

sive targeted interventions. The three tar-

geted recommendations are complemen-

tary to each other, and the panel suggests

employing them together.

Recommendations ive and six suggest

comprehensive, schoolwide reorm strat-

egies aimed at increasing engagement o

all students in school. These might be

adopted in schools with unusually high

dropout rates, where a large proportion

o the student population is at risk. These

recommendations recognize the act that

dropping out is not always or entirely a

unction o the attitudes, behaviors, and

external environment o the studentsthat dysunctional schools can encourage

dropping out. When the school is part o

the problem, recommendations fve and

six propose ambitious eorts to change

8. Christenson (2002); Fredericks, Blumeneld,

and Paris (2004).

9. Jessor, Turbin, and Costa (1998); Newmann,

Wehlage, and Lamborn (1992); Rumberger and

Larson (1998).

the environment, curriculum, and cul-

ture o the school. Recommendation fve

provides strategies or personalizing the

school environment in an eort to address

the problem o anonymity and provide all

students with a sense o belonging. Rec-ommendation six builds on this suggestion

by encouraging schools to provide stu-

dents with meaningul learning through

a consistent emphasis on postsecondary

opportunities.

The panel believes that the greatest suc-

cess in reducing dropout rates will be

achieved where multiple approaches are

adopted as part o a comprehensive strat-

egy to increase student engagement. Al-

though some o these strategies may have

the capacity to improve graduation rates

at the margin when implemented individ-

ually, the panel strongly recommends a

strategic approach that integrates multiple

recommendations and has the potential to

make a bigger dierence.

It is important or the reader to remember

that the levels o evidence ratings delin-

eated in table 2 above are not a judgment

by the authors o this practice guide ohow eective each o these six recom-

mended practices will prove to be when

implemented in a school, nor are they even

a judgment by the authors o what prior

research has to say about their eective-

ness. As noted in appendix A, these levels

o evidence ratings reect judgments by

the authors o the qualityo the existing

research literature to support a causal

claim that when these recommended prac-

tices have been implemented in the past,positive eects on dropout reduction have

been observed. They do not reect judg-

ments by the authors about the relative

strength o these positive eects or the

relative importance o these individual

recommendations.

An example may help illustrate this distinc-

tion. Recommendation 1 has a low level

o evidence rating. This means that there


12/72

OVERVIEW

( 6 )

Table 2. Recommendations and corresponding levels o evidence tosupport each

Recommendation Level o evidence

Diagnostic

1. Utilize data systems that support a realistic diagnosis o the number o

students who drop out and that help identiy individual students at high

risk o dropping out. States, districts and schools should develop compre-

hensive, longitudinal, student level databases with unique IDs that, at a

minimum, include data on student absences, grade retention, and low aca-

demic achievement. Data should be reviewed regularly, with a particular

emphasis beore the transitions to middle school and high school.

Low

Targeted interventions

2. Assign adult advocates to students at risk o dropping out. Adult advo-

cates should have an appropriate background and low caseloads, and

be purposeully matched with students. Adequate training and support

should be provided or advocates.

Moderate

3. Provide academic support and enrichment to improve academic per-

ormance. Help students to improve academic perormance and re-

engage in school. This should be implemented in conjunction with

other recommendations.

Moderate

4. Implement programs to improve students classroom behavior and social

skills.Students should establish attainable academic and behavioral goals

and be recognized when they accomplish them. Schools can teach strate-

gies to strengthen problem-solving and decision-making skills, and part-

ner with community-based agencies to provide students with supports to

address external actors aecting social and behavioral interactions.

Low

Schoolwide interventions

5. Personalize the learning environment and instructional process. A per-

sonalized learning environment creates a sense o belonging and osters

a school climate where students and teachers get to know one another

and can provide academic, social, and behavioral encouragement.

Moderate

6. Provide rigorous and relevant instruction to better engage students in

learning and provide the skills needed to graduate and to serve them

ater they leave school. Engagement can be increased by providing

students with the necessary skills to complete high school and by in-

troducing students to postsecondary options.

Moderate

Source: Authors compilation based on analysis described in text.


13/72

OVERVIEW

( 7 )

are ew existing studies designed to test,

in a discrete and valid way, the connection

between utilizing diagnostic data systems

and dropout reduction. Nonetheless, the

authors o this practice guide, based on ex-

pert judgment and knowledge o practice,

consider utilizing diagnostic data systems

to be an important component o a well-

implemented dropout prevention interven-

tion. Hence, although its level o evidence

rating is considered low, it is included here

as one o our six recommended practices.


14/72

( 8 )

Scope o thePractice Guide

The purpose o this practice guide is to

provide evidence-based recommendationson preventing dropping out. These recom-

mendations are intended to promote stu-

dent engagement with school, suggesting

practical ways in which administrators

can structure eorts aimed at individual

students, schoolwide communities, or ide-

ally both.

This practice guide recommends steps or

educators, administrators, and policymak-

ers to reduce dropping out. The guide aims

to identiy eective practices contributing

to staying in school and completing high

school with a regular diploma. It does not

address recovery practices designed to

bring dropouts back to school or to help

them earn a General Educational Devel-

opment (GED) certifcate. Nor does it dis-

cuss rule-based eorts to keep students

in school longer, such as raising the age at

which students are permitted to drop out

or increasing the enorcement o truancy

laws. Although such eorts may in someinstances reduce dropout rates, this guide

is ocused on strategies that increase stu-

dents motivation to stay in school rather

than penalizing them or not staying in

school.

The suggestions in this guide are in-

tended to be school-based practices that

address students academic, behavioral,

and personal needs. While the panel

eels strongly that early interventionsin preschool and elementary grades can

establish a critical oundation or school

engagement, this guide addresses more

immediate precursors to dropping out

that can be implemented at the middle

or high school level. In the same vein,

while the panel acknowledges the im-

portance o eorts to address the mul-

tiple contexts in which students live,

the guide does not address practices

that involve community- or amily-based

interventions.

The promising results o some dropout

prevention programs and school reorm

initiatives suggest the possibility thatgraduation rates across the country might

be measurably improved by implementing

their strategies on a larger scale. The chal-

lenge or the panel, in creating this guide,

was to distill the evidence rom specifc

programs into a set o more general strat-

egies. The aim o the guide is not to en-

dorse specifc branded interventions, but

to identiy a set o strategies and practices

that are key components o interventions

that have demonstrated promise in reduc-

ing dropout rates.

Dropout prevention interventions almost

always include multiple components, and

the eects o specifc intervention compo-

nents on dropping out cannot be causally

attributed to one component o an inter-

vention. To assess the importance o spe-

cifc components and strategies, the panel

reviewed the implementation reports o

interventions that have been rigorously

evaluated to document the components oeach intervention. It then grouped inter-

ventions that included similar components

to derive the six recommendations, and

reerred to implementation reports to pro-

vide accurate suggestions or how schools

might successully carry out each recom-

mendation. The panel considered the ex-

tent to which various components were

describedin implementation reports or

by developersas signifcant aspects o

the intervention (see appendix D).

Schools and districts may have challenges

in implementing the panels recommenda-

tions in a way that is as eective as the

model programs that were reviewed or

this guide. Initiation o targeted, school-

wide, and long-term strategies should

include a plan to ensure high-quality

implementation o the strategies. High-

quality implementation o the strategies,


15/72

SCOPE Of ThE PRACTICE GUIDE

( 9 )

individually and as part o a larger, com-

prehensive plan, will require investments

in proessional development or sta, not

only to promote sta skills but also, where

necessary, to change sta behaviors and

attitudes.

Administrators, sta, and policymakers

may not see immediate beneits o the

ruits o their dropout prevention eorts;

schoolwide reorms and eorts with mid-

dle school students, or example, may take

years to produce measurable improve-

ments, even i implemented eectively.

Nevertheless, the recommendations in

this guide derive rom the characteristics

o dropout prevention programs, school

reorms, and policy interventions thathave shown promise to reduce dropping

out. While these programs vary in their

specifc components, they have eatures

in common that suggest general strategies

or educators and policymakers trying to

reduce dropout rates.


16/72

( 10 )

Checklist or carrying out therecommendations

Recommendation 1.Utilize data systems that support a

realistic diagnosis o the number ostudents who drop out and that helpidentiy individual students at highrisk o dropping out

Use ongitudina, student-eve data to

get an accurate read o graduation and drop-

out rates.

Use data to identiy incoming students

wit istories o academic probems, tru-

ancy, beaviora probems, and retentions.

Monitor te academic and socia peror-

mance o a students continuay.

Review student-eve data to identiy

students at ris o dropping out beore ey

academic transitions.

Monitor students sense o engagement

and beonging in scoo.

Coect and document accurate inorma-tion on student witdrawas.

Recommendation 2.Assign adult advocates to students atrisk o dropping out

Coose aduts wo are committed to

investing in te students persona and aca-

demic success, eep caseoads ow, and

purposeuy matc students wit adut

advocates.

Estabis a reguar time in te scoo day

or wee or students to meet wit te adut.

Communicate wit adut advocates

about te various obstaces students may

encounterand provide adut advocates

wit guidance and training about ow to

wor wit students, parents, or scoo sta

to address te probems.

Recommendation 3.Provide academic support andenrichment to improve academicperormance

Provide individua or sma group sup-port in test-taing sis, study sis, or tar-

geted subject areas suc as reading, writing,

or mat.

Provide extra study time and opportu-

nities or credit recovery and accumuation

troug ater scoo, Saturday scoo, or

summer enricment programs.

Recommendation 4.Implement programs to improvestudents classroom behavior andsocial skills

Use adut advocates or oter engaged

aduts to ep students estabis attainabe

academic and beaviora goas wit specifc

bencmars.

Recognize student accompisments.

Teac strategies to strengten probem-

soving and decision-maing sis.

Estabis partnersips wit community-

based program providers and oter agen-

cies suc as socia services, weare, menta

eat, and aw enorcement.

Recommendation 5.Personalize the learning environmentand instructional process

Estabis sma earning communities.

Estabis team teacing.

Create smaer casses.

Create extended time in cassroom

troug canges to te scoo scedue.

Encourage student participation in

extracurricuar activities.


17/72

ChECklIST fOR CARRyING OUT ThE RECOMMENDATIONS

( 11 )

Recommendation 6.Provide rigorous and relevantinstruction to better engage studentsin learning and provide the skillsneeded to graduate and to serve them

ater they leave school

Provide teacers wit ongoing ways to

expand teir nowedge and improve teir

sis.

Integrate academic content wit career

and si-based temes troug career acad-

emies or mutipe patways modes.

host career days and oer opportuni-

ties or wor-reated experiences and visits

to postsecondary campuses.

Provide students wit extra assistanceand inormation about te demands o

coege.

Partner wit oca businesses to provide

opportunities or wor-reated experience

suc as internsips, simuated job inter-

views, or ong-term empoyment.


18/72

( 12 )

Recommendation 1.Utilize data systemsthat support a realisticdiagnosis o thenumber o studentswho drop out andthat help identiyindividual students athigh risk o droppingout (diagnostic).

Reguary anayzing student data is te

critica frst step bot or determiningte scope o te dropout probem and

or identiying te specifc students

wo are at ris o dropping out and

soud be considered or extra services

or supports. Te eectiveness o

programs to reduce dropping out

depends on weter tey are provided

to te students wo are most in need,

and weter tey are designed to meet

student needs. Programs designed to

target students at ris o dropping out

need a way to identiy te popuation

tey wis to serve.

In addition, in some ig scoos,

arge proportions o students ai

to graduate. Scoos were data

indicate tat very arge proportions o

students are at ris o dropping out

soud consider adopting systemic,

scoowide canges aongside targeted

programs or individua students wo

are at te igest ris o dropping

out. An initia diagnostic assessment

can ep determine weter te scope

o te probem merits scoowide

interventions aongside interventions

targeted to students at particuar ris o

dropping out. Even ten, compreensive

scoo reorm modes wi sti need to

identiy students wo need extra ep

troug targeted programs.

Level o evidence: Low

The panel judged the level o evidence orthis recommendation as lowbecause there

have been no studies that directly evalu-

ate the eect o using data on staying in

school, progressing in school, or complet-

ing school. The panel believes, nonethe-

less, that this recommendation is a criti-

cal component in identiying students or

whom the subsequent recommendations

o this practice guide are targeted. The

eectiveness o the targeted and school-

wide interventions in the recommenda-

tions that ollow will depend on the extent

to which they are based on an accurate

assessment o the dropout problem. In

addition, it is critical that updated, real-

time data be used to evaluate the quality

o implementation o any o the recom-

mended practices.

Brie summary o evidence tosupport this recommendation

The critical frst step or preventing drop-ping out is understanding who is at risk

o dropping out.1 Implementing a generic

program without assessing the extent o

the problem and accurately identiying

the students who need it is ill advised.

Dropout interventions should be matched

to the characteristics, climate, and prac-

tices o the school and its students who

are at risk o dropping out.2 Schools need

to identiy accurately the speciic stu-

dents who need intervention, and chooseinterventions that align with an accurate

assessment o the problem. Respond-

ing to symptoms may be ineective i

the source o the problem is not under-

stood. For example, schools with chronic

1. Kronick and Hargis (1998); Morton (1998);

Skromme, Van Allen, and Bensen (1998).

2. Duttweiler (1995); Wehlage et al. (1989).


19/72

1. UTIlIzE DATA SySTEMS

( 13 )

attendance problems may be tempted to

implement stronger attendance monitor-

ing, Attendance monitoring may be neces-

sary, but the schools also need to collect

inormation on why students are not at-

tending i they are to eectively addressthe problem.3

The development o comprehensive, lon-

gitudinal, student-level databases that in-

clude unique student IDs has permitted

researchers to identiy actors associated

with dropping out. Such databases now

permit school personnel to better identiy

the individual students at risk o dropping

out, and to identiy them earlier.4 Research-

ers agree that student absences, grade re-

tention, and low academic achievement are

indicators or dropping out,5 and research

shows that critical transition points such

as the move rom middle school to high

school are dicult or already struggling

students.6 Low socio-economic status and

behavioral problems are also known risk

actors or dropping out.7

How to carry out thisrecommendation

Districts and schools should use student

data to answer our questions: (1) What

3. Dynarski and Gleason (1998).

4. Farmer and Payne (1992); Kronick and Har-

gis (1998); Roderick (1993); Suh, Suh, and Hous-

ton (in press); Vallerand, Fortier, and Guay

(1997); Wehlage (1989).

5. Battin-Pearson et al. (2000); Barrington and

Hendricks (1989); Garnier, Stein, and Jacobs

(1997); Ensminger and Slusarick (1992); Jimer-son, Anderson, and Whipple (2002); Alexander,

Entwisle, and Horsey (1997); Finn and Rock

(1997); Morris, Ehren, and Lenz (1991); Rum-

berger (1995); Allensworth and Easton (2005).

6. Allensworth and Easton (2007); Roderick

and Camburn (1999).

7. Goldschmidt and Wang (1999); Rumberger

and Larson (1998); Ekstrom et al. (1986); Phelan

(1992); Rumberger (1987); Suh, Suh, and Hous-

ton (in press).

is the scope o the dropout problem? (2)

Which students are at high risk o drop-

ping out? (3) Why do individual students

drop out? (4) When are students are at risk

o dropping out? Schools should desig-

nate a sta member or team to regularlymonitor data on incoming students, exist-

ing students, and students who recently

let school. This task should not just be

a review o data. It should include regu-

lar monitoring and ollowing up with stu-

dents when needed, which could be done

through student advisories or adult advo-

cates (see recommendation 2).

1. Use ongitudina, student-eve data to get

an accurate read o graduation and drop-

out rates. historicay, states ave oten

overestimated graduation rates and under-

estimated dropout rates by estabising

generous defnitions tat do not provide a

compete picture o te number o students

wo drop out.8 Defnitions o dropout rates

are sometimes not ceary connected to def-

nitions o graduation rates, wic can ead

to contradictory impressions depending on

weter graduation rates or dropout rates

are examined. To respond to te dropout

probem, states, districts, and scoos frstneed an accurate understanding o its scope.

Tis requires, ideay, te use o ongitudina

student databases wit unique statewide

identifers or individua students, tat o-

ow tem rom ig scoo entry to gradu-

ation or dropout and tat incude a pubic,

carter, and private scoos to account or

scoo or district transers. Suc ongitudina

databases aow poicymaers to measure

8. See Engberg and Gill (2006) or an exampleo how this works in one state. Dropout rates

in Pennsylvania have been calculated by divid-

ing the total number o students who ocially

dropped out in a 12-month period by the total

enrollment in grades 712 during that period.

This produces very low dropout estimates or

two reasons. First, many dropouts are not o-

fcially reported as dropouts. Second, the de-

nominator includes large numbers o students

who are not old enough to drop out (such as

those in 7th and 8th grade).


20/72


( 14 )

graduation and dropout rates using sensibe

defnitions: Graduation rates can be defned

as te percentage o students wo graduate

witin our, fve, or six years o entering ig

scoo, and dropout rates can simiary be

defned as te percentage o students woeave scoo witout graduating ater our,

fve, or six years since entering ig scoo.

Te Nationa Governors Association (NGA;

2005) recenty endorsed te deveopment

o consistent and accurate measures tat

use student-eve data ongitudinay to ac-

curatey measure graduation and dropout

rates. Te U.S. Department o Education

(2008) aso recenty endorsed te NGA ap-

proac to be used nationwide.9

2. Use data to identiy incoming students

wit istories o academic probems, tru-

ancy, beaviora probems, and retentions.

Student absences, grade retention, ow aca-

demic acievement, and beavior probems

are strong indicators o dropout ris.10 At a

minimum, scoos soud review incoming

students attendance records, grade reten-

tion, discipinary records, and academic as-

sessments. Scoos soud review additiona

inormation about students rom previous

teacers about eve o motivation, academicpotentia, socia sis, or difcuty to teac.

Because eementary teacers interact regu-

ary wit te same group o students, tis

approac may be especiay useu in midde

scoos to assist students wit te transition

between eementary and midde scoo.

3. Monitor te academic and socia peror-

mance o a students continuay. Scoos

soud monitor student progress by reg-

uary reviewing student transcripts, test

9. U.S. Department o Education (2008).

10. Battin-Pearson et al. (2000); Barrington

and Hendricks (1989); Garnier, Stein, and Ja-

cobs (1997); Ensminger and Slusarick (1992);

Jimerson, Anderson, and Whipple (2002); Al-

exander, Entwisle, and Horsey (1997); Finn and

Rock (1997); Morris, Ehren, and Lenz (1991);

Rumberger (1995); Allensworth and Easton

(2005).

scores, and discipine reerras. Scoos can

use tis type o data to identiy students wo

recenty experienced a ie event, academic

caenges, or oter socia or beaviora

probems tat may indicate a iger ris o

dropping out. Te burden o data monitoringcan be reduced by using automated aerts in

te eectronic data systems to ca attention

to students wose beavior or progress sug-

gests increased ris o dropping out.11

4. Review student-eve data to identiy stu-

dents at ris o dropping out beore ey

academic transitions. Researc suggests

tat students are more iey to drop out o

scoo oowing a transition to ig scoo,

but tey can be agged eary or ris o

dropping out.12 Scoos soud pay particu-

ar attention to students wo ave aied

courses, encountered requent discipinary

probems, or been cronicay absent in

eary ig scoo, midde scoo, and even

eementary scoo.

5. Monitor students sense o engagement

and beonging in scoo. Data coection

and monitoring about scoo cimate or te

nature o teacer-student interactions can

ep scoos identiy areas or improve-ment. Scoos can survey students periodi-

cay or conduct purposeuy seected sma

group interviews to earn about student per-

ceptions o scoo cimate and teir sense

o beonging and engagement. Surveys or

ocus groups can cover topics suc as te

supportiveness o te scoo environment,

perceptions o saety, academic rigor, and

interactions wit aduts and oter students.13

Data coected can be used to monitor scoo

cimate and ep identiy were to ocus re-

11. Neild, Balanz, and Herzog (2007); Allens-

worth and Easton (2005).

12. Allensworth and Easton (2007); Roderick

and Camburn (1999).

13. For examples o school climate surveys in

practice, see Austin and Benard (2007), Chi-

cago Public Schools (2007), and Willms and

Flanagan (2008).


21/72


( 15 )

orm eorts and weter current eorts are

eective in improving scoo cimate and

engagement.

6. Coect and document accurate inorma-

tion on student witdrawas. Departurecodes oten disguise te rea reasons wy

students become disengaged or drop out

o scoo.14 I data on scoo eavers is not

specifc or accurate, administrators wi not

be abe to assess te rea probems and

possiby not pursue appropriate dropout

prevention practices. Administrative codes

suc as et scoo oer itte ep to admin-

istrators woring to understand ow many

students drop out and wy. At te same

time, student mobiity aso contributes to

te probem, but is not aways documented

accuratey. Codes indicating a transer to

anoter scoo soud be used ony wen

enroment in te oter scoo as been veri-

fed. Districts need unique IDs or students

and poicies or assigning tem at te state

eve so tat dropout rates are neiter over

nor undercounted.

Potential roadblocks andsuggested approaches

The sheer volume o data generated in

school systems is one roadblock. Increased

resources, such as sta or data entry or

systemwide technology updates, can help

surmount it, but fnding these resources

may be dicult given other budgeting pri-

orities. Nonetheless, the panel highlights

these roadblocks in an eort to underscore

the importance o data in preventing drop-

ping out.

1. Systems and procedures to update data

are sow and outdated. Districts and scoos

wit outdated data systems wi encoun-

ter probems consoidating and anayzing

student-eve data.

Suggested Approach: Integrated, real-time,

longitudinal data systems now exist that

14. See, or example, Engberg and Gill (2006).

can provide educators and administrators

with timely and comprehensive inorma-

tion on each student.

2. Bureaucratic or organizationa obstaces

may inder inormation saring betweente district and scoos, scoo to scoo,

or witin scoos.

Suggested Approach: Districts can central-

ize data collection and dissemination to

schools on a monthly or quarterly basis so

that schools can regularly monitor student

progress. Districts can also set up systems

to promote the sharing o inormation

across grade levels within eeder patterns

o schools, such as using unique student

IDs, to increase sta accountability or stu-

dents moving rom elementary to middle

school and rom middle to high school.

Schools can adopt policies that promote

regular communication about student

data. For example, attendance oce sta

can update teachers about students who

have had many absences, or early warn-

ing systems can use attendance data com-

bined with and inormation about course

ailures, which do not require waiting or

district or state data, to identiy studentsat risk o dropping out.

3. Data codes do not accuratey reect stu-

dent mobiity, in part because o stigmas

about scoos wit ig numbers o drop-

outs. District and scoo administrators

cannot adequatey design and impement

dropout prevention practices i te data un-

derestimate te extent o te probem.

Suggested Approach: Districts and schoolsneed to adjust data codes to include reasons

or leaving. Students should not be counted

as transerring to another school unless the

receiving school has ormally verifed the

students enrollment. Central oce sta can

occasionally conduct audits o withdrawal

data to veriy the accuracy o the data.

4. Sta may become burdened wit extra

data monitoring responsibiities.


22/72


( 16 )

Suggested Approach: Early warning sys-

tems should be automated to the extent

possible, to avoid additional burden on

sta. Electronic systems should include

programs that automatically ag students

showing a high risk o dropping out, asindicated by substantial attendance prob-

lems, course ailures, grade retention, and

behavioral problems. Schools can create a

team o individuals including attendance

sta, counselors, adult advocates (see rec-

ommendation 2), and other relevant sta

to monitor data rom dierent sources,

such as attendance and course ailuredata, in an eort to spread responsibility

and use the data thoughtully.


23/72

( 17 )

Recommendation 2.Assign adult advocatesto students at risk odropping out (targetedintervention).

Persona and academic needs can be

addressed troug a meaningu and

sustained persona reationsip wit

a trained adut. Te adut soud be

responsibe or addressing academic

and socia needs, communicating wit

te amiies, and advocating or te

student. Te adut and student soud

ave time to meet reguary. Trainingor adut advocates is essentia.

Level o evidence: Moderate

The panel judged the level o evidence or

this recommendation as moderate. The

panel examined fve experimental stud-

ies o our dropout interventions that in-

cluded an adult advocacy component.15

While the design quality o several o these

studies was o sucient quality to allow ahigh level o evidence rating, two impor-

tant eatures o this set o fve studies, in

the aggregate, suggested a moderate rat-

ing instead. First, and most important,

while three interventions included adult

advocates as a key component, none o

these fve studies assessed the discrete

eect o using adult advocates on drop-

out prevention outcomes. In all fve cases

the use o adult advocates was bundled

with other intervention components withno independent assessment o individual

component eects.

Second, across all fve studies, there was

a airly wide variation in observed eects

15. Larson and Rumberger (1995); Sinclair et

al. (1998); Sinclair, Christensen, and Thurlow

(2005); Shirm, Stuart, and McKie (2006); Quint

et al. (2005).

on dropout prevention and staying in

school. For example, o the three stud-

ies that included adult advocates as a key

component, only two interventions dem-

onstrated positive or potentially positive

eects on staying in school or progress-ing in school.16 The third intervention

showed no discernible eect on staying

in school.17 The generalizability o these

fndings is somewhat limited because the

extent o evidence or each eectiveness

rating was small or all three interven-

tions. However, these three interventions

demonstrated eectiveness in a variety

o settings, including middle schools and

high schools across several states. The

panel also considered a ourth interven-

tion that included a case worker or par-

ticipating students, but the role o this

adult was less substantial than the in-

tensive role played by the adult advocate

in the other three interventions.18 This

ourth intervention showed no discern-

ible eects on progressing in school and

completing school.

Brie summary o evidence tosupport the recommendation

Students at risk o dropping out oten

have signifcant personal, amily, and so-

cial barriers that interere with the ability

to go to school and do well.19 Research

suggests that students who have ongoing

relationships with adults eel a greater

sense o school membership, attachment,

and involvement.20 Additional beneits

o adult-student relationships include re-

duced risky behaviors, reduced absentee

rates, improved grades, and improved


al. (1998).

17. Quint et al. (2005).

18. Shirm et al. (2006).

19. Dynarski and Gleason (1998); Rumberger

(2004).

20. Wehlage (1989); Wehlage et al. (1989).


24/72

2. ASSIGN ADUlT ADVOCATES TO STUDENTS AT RISk Of DROPPING OUT

( 18 )

communication and social skills.21 The

adult advocate helps students overcome

these barriers by assisting the student in

addressing academic, personal, and emo-

tional needs. The advocate can model

positive and respectul behavior and oerguidance, stability, and assistance in mak-

ing intelligent choices.

In at least three interventions that have

been rigorously evaluated, adult advo-

cates played a key role in ostering school

engagement by providing students with

opportunities to develop a sense o be-

longing at school and by providing ac-

countability or academic or behavioral

progress.22 In each intervention, students

participating in treatment groups that in-

cluded intensive meetings with an adult

advocate assigned to the student showed

promising improvement in outcomes re-

lated to dropout prevention.

In one intervention adult advocates

worked intensively with high-risk and

disabled Latino students in one junior

high school.23 The adult-student rela-

tionship rested on our key principles:

accountability or student progress, ac-cepting students as they are, attending

to the complex needs o students at risk o

dropping out, and oering exibility and

individualization to the student. The aim

o the intervention was to build a trust-

ing relationship in which students elt

a sense o belonging and identifcation

with another adult and ultimately with

the school. The study ound that students

who participated in the treatment group

earned more credits toward graduation,demonstrated reduced absenteeism, and

21. Pringle et al. (1993); Cragar (1994); Sipe

(1996); McPartland and Nettles (1991); Gross-

man and Garry (1997).


al. (1998); Quint et al. (2005).

23. Larson and Rumberger (1995).

improved their grades relative to those in

the control group.

A second intervention also demonstrated

the promising eects that adult advocates

can have on outcomes related to dropoutprevention.24 Implemented in middle and

high schools, the intervention includes

an adult monitor who works intensively

with students to provide academic sup-

port, conict resolution skills, and rec-

reational and community service explo-

ration. Students in the treatment group

earned more credits toward high school

completion than students in the control

group and were less likely to have dropped

out o school at the end o the frst ollow-

up year.

Other interventions also incorporate

adult mentors, but with diering levels

o intensity o the relationship between

the adult and the student.25 The evalu-

ations o these other interventions did

not examine the primary outcomes re-

lated to dropping out: staying in school,

progressing in school, or completing

school.


Assign an adult advocate to work individu-

ally with students who are at a high risk

o dropping out. The adult advocate acts

as a case manager who interacts with the

student daily, and could be a resource

teacher, community member, or a social

worker. The adult advocate should oer

guidance on matters inside and outsideo school, model positive behavior and

decision-making skills, and be an encour-

aging and trusted person in the students

lie. The adult helps address obstacles

that prevent students rom progressing

24. Sinclair et al. (1998).

25. See LoSciuto et al. (1996); Harrell, Ca-

vanaugh, and Sridharan (1998).


25/72


( 19 )

in school. Here the panel provides spe-

cifc suggestions regarding who adult ad-

vocates should be, how schools might use

adult advocates, and the types o daily

responsibilities adult advocates might ex-

pect to undertake.

1. Coose aduts wo are committed to in-

vesting in te students persona and aca-

demic success, eep caseoads ow, and

purposeuy matc students wit adut ad-

vocates. Te adut advocate needs to be abe

to devote time and energy to mutipe ac-

ets o te students ie. Atoug teacers

can act as advocates, teacers are uniey

to ave te time needed or te advocacy

reationsip. Scoo counseors migt serve

as advocates, but te arge caseoads o

most scoo counseors typicay precude

an intensive advocacy reationsip. Te adut

advocate coud be a resource teacer, com-

munity member, or socia worer. Te adut

advocate soud be based primariy at te

scoo, and soud embody ey persona

caracteristics, incuding persistence, beie

in te abiity o a students to succeed, a wi-

ingness to wor cooperativey wit amiies

and scoo sta, and advocacy and commu-

nication sis.26

Building trust is critical or the develop-

ment o the relationship between the stu-

dent and the adult. With this in mind, ad-

vocates should not have caseloads larger

than 15 students, and matches should take

individual student needs into account so

that the adult can eectively advocate on

the students behal and adapt activities

according to the students interests and

goals.27

Purposeully matching studentsand adults increases the likelihood that

the relationship will thrive. Increasing

school engagement depends on the com-

mitment and practices o adults in the


al. (1998).

27. Sinclair et al. (1998); McPartland and Net-

tles (1991); Smink (1990).

school.28 It is important to provide ad-

vocates whom the students eel they can

identiy with, including advocates who

reect the cultural and ethnic diversity o

the students they are mentoring.29

2. Estabis a reguar time in te scoo day

or wee or students to meet wit te adut.

It taes time or meaningu reationsips

between aduts and students to ave an

impact.30 Consistent meetings between te

advocate and te student provide account-

abiity and te opportunity or te advocate

to suggest guidance or praise successes.

Students aso need time to communicate

rustrations or detais about persona en-

counters. Suc conversations merit reguar

meeting times in order to estabis a trust-

ing reationsip. Te amount o time needed

or meetings depends on te severity o te

students probems: some need daiy meet-

ings, wie weey meetings may be suf-

cient or oters.

3. Communicate wit adut advocates

about te various obstaces students may

encounterand provide adut advocates wit

guidance and training about ow to wor

wit students, parents, or scoo sta to ad-dress te probems. Adut advocates soud

be prepared to ep students overcome ob-

staces tat may range rom transportation

to scoo to poor reationsips wit teac-

ers. Te adut may spend time woring wit

te student on attendance by conducting

intensive attendance monitoring, possiby

incuding contacting te student directy or

contacting parents i te student is not in

cass. Te adut can ep te student deveop

career goas and postsecondary pans (seerecommendation 6). Aduts can wor wit

students on academic progress by moni-

toring te competion o omewor assign-

ments, or wit teacers to earn about te

28. Weinberger (1992); Wehlage et al. (1989).


30. Gunn and King (2003); Letgers et al.

(2002).


26/72


( 20 )

students academic difcuties. Advocates

may even ep te students amiy by reer-

ring te parent to potentia jobs or scoo

training programs, or by maing appoint-

ments or providing transportation to socia

service agencies.

Orientation and training or adults are crit-

ical to the success o a student-advocate re-

lationship.31 Training or advocates should

include inormation about resources avail-

able to assist the student and amily.

Proper training o adult advocates may

be particularly important during times

when the student is struggling with aca-

demic or behavior problems and may not

be interested in meeting with the advocate.

In addition, training and support can help

alleviate burnout caused by the time and

emotional demands, on even those sta

who are interested in this role.


1. Many scoos ac ongoing systems and

processes or meeting wit students. Scoo

scedues eave itte room or discussions

about ie events, strugges, rustrations, andpossibe soutions.

Suggested Approach: Administrators

should consider reallocating daily sched-

ules to provide a specifc period or adult

advocate-student meetings to occur, or re-

quiring meetings to occur during breaks

within the school day, such as lunch or ad-

visory periods. Allocating enough time or

the student and adult advocate to meet is

the key to building the relationship; super-fcial meetings rarely result in trusting (and

useul) relationships. Administrators can

also be supportive o the exibility advo-

cates may need in meeting with students.

2. Scoos may fnd tat sta or oter aduts

are not avaiabe or not interested in woring

wit students as adut advocates.

31. Grossman and Garry (1997); Sipe (1996).

Suggested Approach: Schools should not

orce sta to be advocates i they are not

interested in or committed to developing a

trusting relationship with the student. Ide-

ally the adult should be available beore,

during, or ater school hours to work ulltime as an advocate. Resource constraints

may limit administrator ability to imple-

ment this type o intensive relationship

(see roadblock 4). Schools can consider

partnering with local social service agen-

cies or aith-based organizations to pro-

vide adult advocates.

3. Resistance rom sta wo tin tat stu-

dents aready receive adut advocacy rom

scoo counseors. Scoo counseors (as

teir jobs are typicay defned) and teac-

ers may not ave adequate time to address

individua student needs to te extent rec-

ommended, and may assume tat tis type

o intensive mentoring and engagement is

someone eses probem.

Suggested Approach: Administrators frst

need to clearly defne and explain the role

o the advocates or the sta so that the

advocates can eectively work with teach-

ers and counselors on behal o the stu-dent. Administrators also must encourage

a cultural shit that shares responsibilities

or ostering student success among all

school sta.

4. Insufcient resources are avaiabe to ire

sta as advocates.

Suggested Approach: Not all students in

a school are in need o an advocate. One

way to keep costs manageable is to en-sure that other recommendations in this

guide are eectively implemented so that

the students who are most at risk o drop-

ping out are the ones who are assigned

advocates. For instance, better identif-

cation o those who are in need (recom-

mendation 1) can allow more eicient

targeting o limited sta time, and person-

alizing the learning environment within

the school (recommendation 5) can reduce


27/72


( 21 )

the number o students who need an adult

advocate.

5. Sta perceptions tat students at ris o

dropping out are receiving specia treatment

despite poor academic perormance, absen-teeism, or oter negative beavior.

Suggested Approach: Administrators need

to be the most enthusiastic supporters o

the adult advocates. This includes clearly

explaining the role o the advocate to sta

and encouraging advocates and teachers,

counselors, or other sta members to work

together or the beneft o the student. Ad-

ministrators should seek input and eed-

back rom both teachers and advocates

about student progress so that improve-

ments within the school can be imple-

mented. Schools can also adopt more ex-

ible policies or the advocates in helping

students. For example, changing student

schedules mid-year typically is not accom-

modated, but may be required or a strug-

gling student.

6. Students migt not want to meet wit anadut advocate.

Suggested Approach: The job o the adult

advocate is to establish and maintain a

trusting relationship with the student.

This will likely require persistence on

the part o the adult, and highlights the

importance o the adult advocates tak-

ing responsibility or the success o the

student. I the student is not interested

in meeting, it is the adults responsibil-

ity to fnd the student and establish a re-

lationship or to determine an alternative

approach or helping the student that the

student agrees to.


28/72

( 22 )

Recommendation 3.Provide academicsupport andenrichment to improveacademic perormance(targeted intervention).Providing academic supports eps

improve academic perormance and re-

engage students in scoo. Te pane

suggests tat tis recommendation

be impemented in conjunction wit

oter recommendations in tis practice

guide.

Level o evidence: Moderate


this recommendation as moderate because

o the varying eect o dierent interven-

tions on dropping out and the varying

level o intensity o academic supports

among the evaluated interventions. The

panel considered 12 rigorous studies o

8 dropout interventions that included an

academic support component. Four inter-ventions included academic support as a

substantial component o the intervention.

Among the studies o these interventions,

two ound positive or potentially positive

eects on progressing in school.32 How-

ever, the relationship between the com-

ponents in these interventions and out-

comes associated with dropping out, a key

consideration in making a moderate or

high level o evidence rating, is uncertain

because at least two studies o two inter-ventions that included academic support

as a key component ound no discernible

eects on outcomes related to dropout

prevention. The panel also considered fve

rigorous studies o our interventions that

included academic supports as smaller

32. Dynarski and Wood (1997); Kemple, Her-

lihy, and Smith (2005).

component o the intervention.33 O these

studies, eects ranged rom not discern-

ible to positive.

Brie summary o evidence to

support the recommendation

Research shows that low academic per-

ormance, absenteeism, and grade reten-

tion are related to dropping out.34 Provid-

ing academic supports, such as tutoring

or enrichment programs, helps address

skill gaps and oset a cycle o rustration,

and can enrich the academic experience

or students who may be bored or dis-

engaged.35 Academic struggles may also

play a role in students eeling alienated

rom school, so incentives such as lead-

ership opportunities in academic areas

or rewards or improved perormance

may help increase academic and student

engagement.36

Interventions aimed at preventing drop-

ping out primarily incorporate academic

support components in one o two ways:

1) by oering more intensive in or out o

school programs, or 2) through homework

assistance or tutoring programs. Bothtypes o academic support are intended

to help students reach profciency levels

in key academic areas, but because these

strategies dier and are oten oered in

conjunction with other services, the evi-

dence rom interventions employing these

strategies varies.

33. Constantine et al. (2006); Dynarski et al.

(1998); Sinclair et al. (1998); Sinclair, Chris-tensen, and Thurlow (2005); Larson and Rum-

berger (1995).

34. Lee and Burkam (2003); Rumberger (1995);

Rumberger and Thomas (2000); Rumberger

and Palardy (2005); Rumberger and Larson

(1998).

35. Balanz, McPartland, and Shaw (2002).

36. Wehlage (1989); National Research Council

(2004); Quint et al. (2005); Larson and Rum-

berger (1995).


29/72

3. PROVIDE ACADEMIC SUPPORT AND ENRIChMENT TO IMPROVE ACADEMIC PERfORMANCE

( 23 )

Two interventions that have been rigor-

ously evaluated provide evidence that stu-

dents at risk o dropping out who receive

academic support through specifc courses

may be eective. One intervention, a whole

school reorm model or high schools, o-ers additional reading or math courses or

students who need additional support.37

The reading course works in conjunction

with other core reading courses oered at

the school, and incorporates educational

sotware as an important component o

the curriculum. This intervention also in-

cludes summer school and Saturday school

components or students to both accumu-

late credits and receive academic help. Stu-

dents rom the intervention schools made

larger improvements on academic credits

and promotion than students rom the

comparison group. However, because the

intervention contains several large-scale

whole school reorm eorts, it is important

to note that the eects o the intervention

on dropping out may not be attributable

to academic support components.

A second intervention oers remedial

reading programs or students as part o

the daily schedule. However, the interven-tion itsel diers rom other interventions

in that it is an alternative high school

model where students are also provided

opportunities or credit accumulation

and independent study.38 Thus, while the

academic support through a specifc re-

medial reading program may be eective

or preventing dropping out, the type o

school in which the evaluation occurred

was somewhat dierent than that o other

schools.

The eectiveness o academic support in

the orm o tutoring or homework assis-

tance can vary and may depend on the

other services oered by the programs.

Three dierent interventions that have

37. Kemple et al. (2005).

38. Dynarski and Wood (1997).

been rigorously evaluated provide tutor-

ing or students, either during or ater

school. One experimental study that in-

cluded more than 1,100 students exam-

ined an intervention that includes tutoring

among a variety o other strategies, suchas fnancial assistance or postsecondary

planning, but demonstrated no discernible

eects.39 Another intervention included

tutoring our days a week or 12 hours

daily, as part o a combination o services

designed to help low-income and potential

frst-generation college students complete

high school, and a rigorous evaluation o

the program ound that students partici-

pating in the program completed school

at a signifcantly higher rate.40 Mentors

provide homework assistance, in varying

amounts and intensity, or groups o 12

studentsas part o another intervention

that showed positive eects on staying in

school.41 Finally, two interventions that

had positive or potentially positive eects

on dropping out have the adult advocate

(see recommendation 2) provide academic

support or assistance or students when

needed (recommendation 2).42

Evaluations o other interventions thatinclude tutoring and ater school home-

work assistance have also been conducted,

but have not measured staying in school,

progressing in school, or completing

school.43


1. Provide individua or sma group support

in test-taing sis, study sis, or targeted

39. Shirm et al. (2006).

40. Constantine et al. (2006).

41. Dynarski et al. (1998).


al. (1998); Sinclair, Christenson, and Thurlow

(2005).

43. See Cardenas et al. (1992); Harrell et al.

(1998); Mehan et al. (1996).


30/72


( 24 )

subject areas suc as reading, writing, or

mat. Individua or sma group support

provides a comortabe pace or strugging

students to earn and eps students persist

in caenging courses. Academic support

can appen troug one-on-one interactionsor sma group interactions, and can incude

test-taing and study sis or enricment

courses. Enricment courses, in particuar,

oer opportunities or students to receive

additiona support. Tese types o courses,

about 1012 wees in engt, target a par-

ticuar subject area suc as reading, writing,

or mat and incude teacing strategies de-

signed to engage students suc as woe

cass discovery essons or dierentiated in-

dividua and sma group instruction.44 Aca-

demic support can be conducted by aduts

or by peers, and can occur during advisory

periods, unc, or during study sis periods

buit into te scedue.

2. Provide extra study time and opportuni-

ties or credit recovery and accumuation

troug ater scoo, Saturday scoo, or

summer enricment programs. Summer

programs, ater scoo, or Saturday scoo

sessions may oer students personaized

opportunities to improve academic sis.45Ater ours and summer scoo programs

address severa primary academic needs or

students at ris o dropping out, incuding

support or students during te transition

rom midde scoo to ig scoo, credit ac-

cumuation, and academic enricment aimed

at increasing engagement. Students wo

ai to pass required courses require sup-

port to catc up on bot academic sis and

credit accumuation. Ater scoo or summer

scoo programs can uf bot needs. Intis type o program, students soud wor

cosey wit teacers eiter individuay or

in sma groups to compete coursewor or

credits required to graduate.46 Scoos can

44. Kemple et al. (2005); Kemple and Herlihy

(2004).

45. Roderick and Engel (2001).

46. Kemple et al. (2005).

aso provide students at ris o dropping out

wit summer scoo enricment programs

tat increase engagement. Tese programs

can tae pace on scoo campuses, at dis-

trict ofces, or at partner sites suc as oca

postsecondary institutions or communitycoeges. During te program, wic can

ast 46 ours per day or our to six wees,

students soud be exposed to a variety o

experiences tat target ey academic areas

suc as mat, science, or reading.47


1. Tere is no time in te scoo day to add

anoter academic program.

Suggested Approach: Schools may want to

consider oering enrichment activities or

academic support at times dierent rom

core classes so that students do not miss

key curriculum while they are receiving sup-

port. Schools can oer enrichment courses

as electives, or oer additional support dur-

ing advisory periods at the beginning or end

o the day. Another alternative to providing

academic support during the school day

is to select students to participate in aterschool or summer school programs aimed

at providing support to develop academic

skills or earn credits toward graduation.

2. Students do not want to participate in tu-

toring outside o scedued casses.

Suggested Approach: Given the diculties

at-risk students may already have with at-

tendance, it is more realistic to schedule a

support class rather than expect studentsto attend tutoring sessions during typical

social times at school such as lunch.

3. It may be ard to fnd tutors.

Suggested Approach: Administrators can

be creative in implementing solutions

or individual or small-group tutoring by

47. Snipes et al. (2006).


31/72


( 25 )

establishing partnerships with local com-

munity organizations and postsecondary

institutions, recruiting parent or business

volunteers, or using sta in the school

who express interest. In some schools at-

risk students may be eligible or supple-mental services prescribed under NCLB:

Supplemental Educational Services are

available to low-income students who

attend schools that have missed ade-

quate yearly progress or three years.

In most school districts those services

have not been ully subscribed, with par-

ticularly low participation rates in high

schools.48

48. Stullich et al. (2006); U.S. Government Ac-

countability Oce (2006).


32/72

( 26 )

Recommendation 4.Implement programsto improve studentsclassroom behaviorand social skills(targeted intervention).Scoos can ep students identiy,

understand, and se-reguate teir

emotions and interactions wit peers

and aduts. Doing so can ep to

mitigate probematic and disruptive

beavior bot in and out o te

cassroom by teacing students ow to

interact and communicate positivey.An additiona beneft o tis type o

si deveopment is to ep students

consider ong-term consequences.

Severa o te dropout prevention

programs tat ave sown promise

in increasing students persistence in

scoo specifcay see to deveop

tese inds o sis. And activities

tat require students to tae on new

responsibiities and interact witengaged cassmates can promote

scoo membersip and deveop a

sense o se-efcacy.

Level o evidence: Low


this recommendation as lowbecause o the

varying eectiveness o dierent interven-

tions and the varying level o intensity oeorts to improve students classroom be-

havior and social skillsamong the evalu-

ated interventions. The panel considered

six rigorous studies o fve dropout inter-

ventions that included eorts to equip stu-

dents with such skills.49 One intervention

included a problem-solving curriculum


al. (1998); Sinclair, Christensen, and Thurlow

as a substantial eature o the program,

and the evaluation o the intervention

demonstrated potentially positive eects

on staying in school and progressing in

school.50 However, the external validity o

the study is low because it included ewerthan 100 students rom one school. Two

other interventions also had potentially

positive eects on staying in school, but

the eort to improve students classroom

behavior and social skills was a somewhat

less critical component o these interven-

tions.51 Finally, two rigorously evaluated

interventions showed no discernible e-

ects on progressing in school.52 The vary-

ing eects o the interventions and the

varying intensity o the eorts to improve

classroom behavior or social skills among

the interventions components preclude a

strong rating.

Brie summary o evidence tosupport the recommendation

School engagement includes a component

o behavior and a component o identifca-

tion with school.53 Disruptive behavior is

correlated with dropping out.54 Dropout

prevention interventions have sought todevelop students problem-solving and

lie skills basedon the need to enhance

student abilities to behave positively dur-

ing school, thereby increasing a sense

o school aliation and engagement.55

(2005); Shirm et al. (2006); Dynarski et al.

(1998); Snipes et al. (2006).


51. Sinclair et al. (1998); Sinclair, Christensen,and Thurlow (2005); Dynarski et al. (1998).

52. Shirm et al. (2006); Snipes et al. (2006).

53. See Voelkl (1997); Finn (1989); National Re-

search Council (2004).

54. Rumberger (1995); Rumberger and Palardy

(2005); Rumberger and Larson (1998); Swan-

son and Schneider (1999); Goldschmidt and

Wang (1999).

55. Rumberger and Larson (1998).


33/72

4. IMPlEMENT PROGRAMS TO IMPROVE STUDENTS ClASSROOM BEhAVIOR AND SOCIAl SkIllS

( 27 )

Developing theseskills goes beyond en-

suring students are not disruptive in

class, however. It teaches students about

Documents

Dropout Prevention_what Corks Clearing House