* * * * * *.4 0 H c+P. 0) M 0 01fic *

rh 1-1 0 (l) Irre- 17' 0 Cr 11 t7 *0 (1) cl rt 1- 0 *(t) cruller)

E3 0 1-I *1,* b 01 N 1-. 121 En El* tr t-t td tr m ** H 11 1-, (-1- 1=1 b*

Q 1-1 Cr 0 ct **tVM tit (1) c+ ** 0 0 0 c+* ex, 11 0 0w b 111 fle 4e:* to 0 Cr 0 .1 ** rr G (1) P. (n fl) **a b EI 1-1 1- *-et ri P. ** m b b C1 at fy ** LEI rr a0 0 b 11 ** 0 M Ry 0-) ** b 03 ix) I=1' 0 I-4 ** M 1-1 M a M Cr ** N Ry 11 Cs P '<I kit* b c+ 11 ca C t-1*mw 0 c)* ta O b P. b GO ** + 0 rz ft, 0 1-- 1-1 ** 0 -:1 0 111 0 n ** t" b rr **b 0 P. II 0 1-1b P. ** 31) 17 0 ..1) 4-f" m m b** (+ 0 0 '0 ID t-1 n *ti H ** 0 0 f) 0 ,(1) a' En ** 11) 11 cD ta 1:1 ID 0 Ca) ** b e-h H 0 .1 0 m ** -&C) cl 0 at ED 1-1tr 11 C 15 ** 0 0 c+ 1:11 ** a (D Q Cr En b ** ia LI rt. a) Le ** - En :7' t-* a, a. ti Wm ttt P. *aF M 0 0-tti (1) b ** 0 trt Mt H t-h **MG (AHD). 00 **H 19 """ n M H ** o a) HI -1. 1311 6I ** la 121 omrt111 111 ** rf tti 111 Ei), >c ** re- 0 Pc. rh ** b t1.1 (t) U) Ul to 0 **m t:4 (i) 121 ** m I+ 0 M c1 ** 0 t1 I-J 111 b M.4 ** H H Ul C M m b** P. 0 11) El 0 1-1 ** Q O. 0 W1103 11) :1 P. ** P. 0 0 1-1 0 1 Ul ** 0 0 re- fa fle ub m 1=r ** a re- Cr } P. 1-p m *

F-'N 0 M a ** 0 (1) c+ Co 0 ** :1 1-1 H ** N rt ** ** * * * * * * * * *

,gym Cr' Eni re, ub tit tO tr (0Da. ti 0 0 1-1 rt a) a) at o

En c+ a) tr. 1-. ri ct 0 13tn. IA C. ry: 1-1

0. 11 J. a) to a) .a) fat0 a) 1-1 lu tn E3 111. M 1-1

he c1 111 te 0 11 a.rt El H c+ 11) b 11)

tr. Cr 64 111 0 H rr 0, 1- 0 1(t) 0 0. < 7 ID m pe fre.0 Cr I a) (n a M 0.0 0 0 0 I. (0 0 ri

(1) tO ua ht, 0 11 IJ -t(11 H P. 11 1.11 IC rr c+

M rt fle b ID lit G-1 0 fle tr1-1 m ca, 11) ca tr (I)P. rt M re- %(-4 1=1

I.-. En t=t ri tr: rt tiI kol (t) o ca tt)1,1 (i) cf- III 01 a' illCo C +4 rr 0 c1 t-h 'InC' 17 b P C:: CI) un En to r t 13

0 I-. 1-1 0 0 ty 0 "Li hl0 0 (A 12) 11) u II H lttb H 00011111N 1

1-1-4 11 0 b & o0 to En re H M ,t1 0 11rit ID c 111 h m ti

0 N to 're a) 0 rtc1 0 Cl, 11 Cr 0 0 C1 0 011 Cr I)0 M 0) IC (1 (1* O 0)

( 0 1-1 En 1-1

tQ H b' 411011011 DI I' 1-1) 0 0 HRj 11. ta ' 11)

1 1 1 t r iv ch i 0 0 fl)E3 a fl) 1-' 1=1 0 CI 0 1-11 tD ti11 11) 15 ti CD 0. 1-1 0 Cr 0

1-1 P. 010 f+ 0M H Ry la) 0 U1 1-1 C1J Ry

M 1/1 t-1 El tr 0 fle0 1 1 ' C u 0 0 0 0 eir 111 cl

0.) t-h 0 to C (r) 0 H N re- Mrt t-h 0 0 re- M 0 1-1 cc 1=1 re- he

Izi 0 11 Ei) er) ta o m1n Et) trfa 1-1 En H 11. 11 1-1 S.1 P1 1-

rt 1-1 rt Lit o c1 N lat 01 ca 1-L70 111 0 I=r 41 0 U) 0 13- 0)

El 1-1 H 1-. tt (i)to u 0 En ti a rt En tit 0 ti0 in d 17' w c1 r1) Oh 71l=r Cu a) c+ rt r1 0 11 1;1 11 ub0 0 ti 0 Cr Ul 11 11) 0 0003 M 0 0 0I 41 0 fl 0

ca. 0 tr b NE) - 0M 0 M '0 0 II (I) N. fa) En mr bPr" la 0 a) 0 11) 111 M IN 0 c f-at t-4'U rt a) tit b110a)M11/ :fro 0 0 0

H t-h DI m to a 1:1) I-1 J 0 0 1-10 Cl (11 1-1 1:11 ) -1 (D b 1-1 I-1H 0 rt M LI lat c bla Ity Ft 0 rob hz, Cu 111 0-0) 111 En 11fle 0 0 re 1-1 m t 1 ..1D lD H 01.1 kr to ID 0 lat 131 17- m 0 Cs

(1) Ca LL) ti fle r .L1CD U1 H 0 h 1--.

111

.0- d El o

T1 En Ira fle AC En t-t tort. 0 0 0 0 Cr 0 0 M13) 1-1 tr 11) fl) It) ry 0 0-U 0 CI) C1/ 0 11 rta.tr0 (i) m ty 0 (1)a 0 n n (1, W ru (1)t-1 44 1-t )1 np. I=1' 0 0 I 1)1. al on ri1.1 to a ti 1-rt 0 10cu 0 rr rr(J1}-1OrocD0rf 'Li 0 (D ti 0 riM U1 rt. re C 1- fle tptr) 0 (1) ti ta 1

re- 0 0 11 rr' AD (0(t) 1-- 1/1 a) 0 CI 1:11

0 (r), /-11 (Do M m 'Li 1-. (") ( t0 re al D. It ti m tr°I-J crit) UP fp °w rt-PFI)

'1:111 0)2. 10t-11:1 (t.(1;(k(D/1"b AI 0, 1-1ww'O f-'(t) 3C (1 Cniao

CA u, I- 0.0 0 u En

on tt; 41) rt.(11

p1) 41 (I) 01

re 00 re'

C) 1 fl)0 G Cr 0 0 1-),4. v I1-1 4 ti Li ,c3

(r) ut:: 1-4 ttrEl CI rt m 0 11

M I 0I-' 1 (1 I- I- 11) 1-'(ft (A En (n a) t.

H Cr I^in gc rt a ko 0r1

fl. (t) m eltr$ 0 ti ln ct rt rfub 0 (T) 1-

b' 0 m 0, 0 0(0 0 rt 0 11Q. la tit to tr)

't/ to 0 HM 1 1 ) b ( ) t1

t-

0 (1) C) C1 tOre 0 re rt)ID C1 0 fugai t t rt Cat trf

a) 0 a)14 Co M (0 1-1 1-1

rt rf W N 11)mH 0 0 11)Cu

. 0 I=1 (nPt, (l).60 ri I 0

0 riH0 AC wrt

t Ittstr.-1I I111

1-1111

11

U)

$ 4tr 0fle c.uc t-(1,

fa AI (D 4.0 11I 1; I 4 C4 iv .

t 11) .0 1-1 1 rt el

rr rJ I% I 0 LH t rt:() + lo () 1-0 Hi 0 ( 7lf ID C) 1/1 t I0 Di t n 11 a) roe toI- 0) M P. 00 (r) t,0 7

n oc ou t 4.4O cf ri) o (..1to ic 0 to 0.

0 0. o t1 t" a)- G co 11 Do O :1 oo,

," 'Ll P.(I 11 to 0 0 0 rt(* rtP 1-1 . L.*

Cu 10 10 Pr, 11) a) *t-t I-1 LI 41 11 Luc 1L1

03 0 0 0 r$' 4) 0 0 0tl kg.) 0 Ll to kr Ullat II 0 11 I I a) rt

at .1) 11,

Imp vl 041 fir) (1141 111

tze h1 I- 1 - 13rh 01 a (1)1-11 111

in m Pl ct t0(1 Mr, let

Ct P I 1 fn(-1. m

111 (1) N t1'ti Ii 1 'xi 171 1-,- Enty m 0 0 0 t0 I-c GI DI

LP 13 En W It-/1 - tpc 11 01 1 ) 0. 1 1 0 0 u

RI ID tr tr.) 1I II tn W ti (IM 0 re tt) r .4En 4-1 1' .4 Mri 1 re tl)in 0. k.14 rt -.- - En E.+

ft

rnC)

1-4

1-1

Itl

Lti

Vl

oviCU

HI

hl

1%1 txt t ) tl cxit tie ft) tv..0 10

n, 4,) .t It 1. c).rr ti

(

1-1m

(1)

rf

.0

00

LA1 117 '1/CP LI I1

I CU 't1hl '

L1 tvw 11

tNI Lr:

Q

I 1

(I)I-1

H-1C.1

1-r

tr.

IL.) Cu 10 01 Ill ,r)C:1 0 At: 11 At I'ti I 0. t ()

tit AI %LP Clt

J ( ) 10 I' 1-L I 111 P. 111

10 PII II) CI (/)

lJ H 1: 1t11 ) 41 1

Cr tAlU as

DI(I at 0)r1 0 11 rt

t ID ,t(t) 0H w to

t.. 0 III1- tr: rt.

0 -t.o Et)

u)1.)

M ttrt. 1 ft- U t

,1 ort*I DI ID((1 0I--1 11 114

1-h C 011)

ri ())tl)

m L)

.c1 1.40 LI/

r4 Ib

111 0t I 't C)

t,)

t't

C.1

13

CJ

C1

rn

?ALAVA SETTLEMENT. ,EFFECT OF TtiE ACADEMIC-ATFIETIC

PROGRAM ON FUTURE SC :1001. PERFORMANCE

r-C:D

C\J

.1)

(-NJJS of Astruc or,

wtoregC:) wa

tfr,..casoa,

t E ( (EsasE't .E: ,st..or Pt le:44. :4 Ca maw I4- : :4

4,c t.5+ t1-.tf: :. tlttit"( :g t.' ..E :tEC--L" tr. Pt; t t& .

a

Pal Se't:le=ent

819-Nort.1/7:ineya.rd'Blvd,

HonoiLlu,

C.eorge F. Schnack, M. ., ?reszdenc

Board of 7rbstees

Robert R. Higathino, Executive Director

This Report was prepaied by theSocial Welfare Development and Research Center

to cooperatirn with tht staffof ?Ala= Settlement

Social Welfare Development and ReseardhsCenterSchool of Social WorkUniversity of Hawaii

lack T. Nagoshi, Director

Cl4fford al O'Donnell, Ph.p.Associate Researcher

Barbara O'Donnell, M. S. V.Experimenter

July, L.V4.

2

TABT '2:11- (0.1.77!::c

Page

LIST Of TABEES ii

PR :ACEt4,1

ASTRAC7

rNTRAXCTTON

Relevance sf School :: Dellnquencf* 1

Revie.: of Literature . . 2

Desc;iption :f, ?a aria Setclement ......Researc. Question and Hypothesis 5

METHODOLOGY

S6mple See/ecticn . .N.

7

Measures of Outcome 7.

Description of D.= Collection 8

Statistics 1:s .- .!

87

FINDINGS

.

Lovz - Achieving Students: Is ?aline Reaching its Target Group? 10

C:eoeraltzarion.f Palama Results 11

CONCLUSIONS AND IMPLICATIONS OF STUDY

Major Conclusions , . / 14

kelation to OtWer Studies.

16

Implications 17

REFERENCES 19

.

:SIBLYOGRAPHY 21

'AFFENTI6ES

. -,,, A - Fatima Settlement Beard of TruiteesB Behavior Modification Football StaffC': UhavBr Modification Basketball Staff

ler

1.

LIST CYF TABLES

Table Page

I ".ears of Pre-Progrez School Perfor=ance for the1970-71 'Scnop.1 year 11

2 Means of Past-Program School Performance onSTEP mac SCAT Math Tests

3 Correlations Between Prograr. Xath Gains and Post-Progra=Scnool Perfoczance kteas'..res for tne Experimental Group' .

-s

44,

4

11

13

iii

PREFACE

Ln December, 1973, the Sociel Welfare Development and Research Center

issued a report entitled New Approaches and Innovations at Palama Settlement.

It summarized two important programs at Palama Settlement, the In-Community

Treatment Project and the Behavior Modification Sports Program. At that time,

Palace Settlement-was characterized as being "one -of a few programs in this

011

State that have adopted a research strategy to continually assess effectiveness

through objective data.' This report testifies to the accuracy of that state-

ment. The data presented are part of this continuing research strategy to

4111 assess the effect of the acadeTic-at-letic program on future school performance.

;he results have important implications, not only for the program at Palama

Settlement, but for many similar programs in this State as well as on the

Mainland.

For this accomplishment we once again recognize the facilitative role

played by the Board of Trustees and Ehe Executive Director, Robert H. Higashino.

The prime mover.of this research effort is Barbara O'Donnell who engaged in

this study in preparing a Research Paper for her M. S. W. at the School of

Social Work, University df:---Rawaii. This report has been adapted from her

Research Paper withlitinor revisions. Many people helped make her study possi-

ble. Of prime importance are Earlene Chambers and Kenneth Ling of Palarna

Settlement who continue to be both committed to this project (and receptive to

innovations; Joel .Fischer, Professor in the School OTSoci I Work, University

of Hawaii, who served as a most able research advisor; and lifford R. ODonnell

.of the Social Welfare Development and Research Center Who pr vided technical

assistance rn research design and analysis, and co5sialtation.

5

Jack T. Nagoshi, DirectoSocial Welfare Developmen

and Research Center

. ABST'RACT,,

iv

This stJd:. exa-.ines trt hypothesis that participation in a Palama Settle-

ment louse after-school, acaceric-athletic program leads to an increase in

academic performance in tne targeted subject (mat) in school. It was also

of interest to know ...nether or not the progra was- raaching its targergioup

of low-academic achievers. A sample of youngsters who completed at least one

academic-at'nletic season (basketball or football) Ln the Settlement programf

were matched for sex, grade Lr scnool and scn :oi attended with a non-partici-

pnCln contr:1 group. For both 1..roups, school grades in math, verbal and'

math scores on stardard..zed tests (ST E? ond SCAT) and attendance in schdol

were co-pared both for the year prior to and following the program. %Mese

scores were obtained from scbool. records.

_ .

There wefe no differences between the two grbups pnpre-program school -*

math grades and verbal or math achievement tests- Thus the pre-program scores'

show that volunteers were the same academioally as non-volunseers. The post-'

program scores show that academic gains made in the program result in improved'

post-program STEP and SCAT math scores. but do not show up in post-program

school math grades:,"(

These findings underscore both a success in the Palana program for im-

proving math ability and a failure for this to be reflected in school,math

grades. This implies the need for closer cooperation between youngsters'

schbol'teachers and the Palama academic coaches (programmers of academic

material)" to ensure ehe generalization of program results to school performance.ht

1

1

r

INTRODUCTION

The problems 000r ser7.00l performance and non-attendance, dropping out

of school, school vandalism and dfug ab,..se are ;ell known to educators, law

enforcers and those in t-e nelping professions. Polk and Schafer propose

"mat educational failure - by schools as well as by students - is directly

related to delinquency.,1

7ochios stated tnat, "failure to master the tools

.of learning - Atticularlv reading - leads cniicren to err.ploy coping

Aktn:ds .of brava:,o, 'acting out', and truac.cy to escape eMbarrassment,'anxiety

and fa.iure.*2

i'alama Settlement -ouse in "awaii ios an after-school program to improve

academic performance in a scnolastically low-achieving target population.

Their basic concept is rooted in the behavior modification approach of rein-

forcing,theprosocial behaviors of studying and team cooperation, The program

has.been shown to improve students' math'abilliy from .5 to 1.2 school grades

in 10 weeks. 3This study is being undertaken to determine! if thii improvement

'results kn higher school performance in math grades, achievement tests and

attendance.

Relevance of School to Delinquency

"Okrer one Million young people are referred to the, juvenile courts each

,ear, the nuthber representing close 3% of all children between the ..a.g4.6

Of 10 and 17 (the population at risi)."4 "At the same time, available evidetice

srtrangly suggeti that-delinquenticommttments result in part from adverse or

negative school experiences.

famiiies and'slum conditions/ re not denied. Eliminstion of juvenile delinquency.

, i .. . / . , .

,

s .

waits 'upon major social nge,; howevfer2 as studies have shown a connection

The influences of poverty, brOken and troubled

2

between juvenile delinquency and the school, theie =ust be change as an inevi--,

table part of prevention.

For low-performan-e students, stho-,1 is associated with punishment because

of failed courses, negate. s. of te.=;.ners,and parents, :and a :-.egative self-image.

t

Association by school fail..ires with others who.nave failed ind a comnitment to

a. delinquent community are possible avenues for tle drop-out. -"The essential

notion tie vie; is that. 3s a reoult of the lowering of an individual's

commitment-4...c success and conforr-ity, there is an increased probability cf his

'corning to a character_sticlly rebe"--,.coeer .oultLre or siznificant social

attacn7-ents wnich, in t_r-r,_ _':crease t-e c-.7nces of irvolve-7.ert in delinquency.' 6

Such a notion is supported and 1.,..ce:CK7-whc found a high correlation

between dropping out of sc-.00l and delinquenc!..

Review of Literature

The past 15 years have seen an enormous amount of liteiature-publish'ed,-

,,on the subject of behavior modification. Material on operant conditioning,

in particular token ecnoomies whic) rely on the principle of immediate rein- '

,

forcement for desired behrior,8.forms the basiS for a program such as Palama's

Behavior Modification Sports Program, wherein athletics are used as the rein-_

forcer to increase academic behaviors. Studies on various types of programs,. .

including programs for youth in a ComractiOnal facility (Col-ien and Filipczak,9

O'Donnell and DeLeon,10

andNcKee,and Clemen-s11) and for drop-,outs from school

(O'Donnell and Sianley12),hgve suggested there may'be-considerabce merit in

utilizing such' an approach.

,A

One study which closely parallels the Settlement. program under investiga-

tion here was done by Wolf. Giles and Hail.13 This program had a sample of

16 low-achieving students from a low-income neighborhood in Kansas' City. All

8

3

of the= scored at least two years below the norm for their grade'level (5th

and 6th grades) on tne Stanford Achievement Test (SAT). The parents were con-

tacted by a social worker who explained the program to. them; all parents

contacted er.rolIed tneir children. Thestudepts attended,"school" in a nearby

church basement eacn weekday after regular school for two and a half hours, and

an Saturday morning dLring the academic year. Ine Program was fn progress for

15 ,T.onM-:s. Points were given for academic work completed either from the regu-

lar :r tne re-edial scool; for example, a student received 100 points

for ;.ni: 75 pc.n:s for a B, etc. Reinforcers, coLld be obtained with

the .ed o:in:i, incl-ded field trips, snacks, novelties and clothes.

r.tn r con:Inzencics were provide'd for in tre program, such as bonuses for

instructors of children whose schook grades.improved, bonuses for good student

,behavior and attendance, parties for report card improvement, and an opportunity

y

for their regular school ,teachers to add points for'excellence in regular

schodl orto subtract points for inappropriate behavior.

The primary goal was to help students make larger than usual gains in

their academic skills,. The remedial group was randomly selected from the same

pool df"low-achie-7ers as a no-"Treatment control group. Students were ranked

according to reading. scores as measured by the SAT. The lowest scoring student

was assigned to the remedial, group, the next lowest to the control group, and

soon. Sixteen students were assigned to each group. Fifteen were in each. .

group at the end of the program. SAT 'scores and report cardgrades were recorded

for.the two years preceding Ehe program and for the year following. the program.

The median gain after the 15 month program on report cards for the remedial

group was 1.1 grade points, from slightly below a D to a C average, while the

control group gairltd 0.2 grade points. The median gain of the remedial group

on the SAT was 1.5 years as compared to 0.8 years for the controls. The pre-

vious' median gain on the.SAT for both grOUps was 0.6 yearS.

4

The cost per student was about $250 a year:, The authors suggested that

costs could be, reduced by using reinforcers which already exist in the educe-.

tional setting. One of the suggested reinforcers was athletics, which the

Palama program uses as a consequence for increasing academic behavior.

Description of Palama Settlement

Palama Settlement House, one of Hawaii's.oldest-social work agencies, was

founded in 1896. Since that year, it has pioneered many social programs such

as out-patient clinics, family planning clinics, programs for the elderly,

young and many more. Historically, Palama Settlement efforts have gone into

aiding the disadvantaged. One of these programs is the Behavior Modificatiolia

Sports Program. 14 It was reasoned that a sports program could provide an out'lg

fop aggression and at the same time teach lessons about sportsmanship and com-

petitiveness. The sports program is also used for the additional purpose of

providing motivation to improve academic skills. Boys participate as athletes -

and girls participate as cheerleaders. This participation is contingent upon

studying and improving academic performande. Points are' earned'by studying

in an individually planned program and for sports participation. Students,

between 1016gears of age, From a Model Cities area who volunteer for the

program must sign a contract,specifying behavior to be performed, such as:

0 1. Do their best on all academic testing.

2. Accept the academic program designed by their 'academic coach' (the4learning engineer at the, Learning Center).

3. Complete their academic program with a least 90% accuracy each week

to be permitted to-practice with the athletic team.

4. Participate in daijy athletic practice.

5. Eat at the Settlement daily during the athletic season accordihg Co

dietary prescriptions.

10

\ 5

6. Follow all recommendations based-on medical and dental exam,inations.

Once the contract is signed and the student enters the Sports program,

\e_academic skills are assessed nd an individualized program is prepared; resting

is done at the Center and must reflect 90% accpracy. Points are earned for

hours of academic work and for rformance on the athletic field. These points

and gold stars are publicly Rosted and can be exchanged for sports equipment

and sportswear. Volunteer academic coaches also receive start redeemable for

clothes and dinners, etc. A sample daily schedule for the students might be'

as follows: practice 4:30-6:00 p.m.; shower 6:00430 p.m., dinner 6:30-7:00 p.m.;

. study hall 7:30-8:30 p.m. The use of Palama facilities for study hall is op-

tional, but points are awarded for using it. Sports competitions are held on

weekends.

A previous study on,the program under investigation here by O'Donnell,

Chambers and Ling15 found that reliable gains o£ .5 to 1.2 grade levels on

the Stanford and California Achievement Tests were achieved during the LO -week _

. ,

athletic season between pre- and post-testing in all areas in which academic

work was programmed. These results suggested that youth whose' mean pre-score.i

are below average on national norms can achieve sizable gains in math in 10

weeks.

Research Question and Hypothesis

The preSent study was designed to answer two questions not addresied in

the above mentioned study. First, were those who volUnteered forthe program

higher, equivalent or lower in school performance as compared to students from

thesame area.and schools who did not volunteer for the program? In other

words, is Palama reaching its low-achieving target group? If the students

who volunteered for the program were performing higher in school to begin with,

. 4 6

this could-be the reason for the program's success. It would also meap the

target group was missed. if.the academic levels were equivaleqt, results could

be generalized to similar students fn Kalihi-Palama schools, but the target group

would again be missed. If the program volunteers were of lowen academic back-

grounds, then the target group would be reached.

The second question seek to answer whether or not the program results

generalize to post-program school performance. The hypothesis Is that Palama

prograreefforts do result in higher academic performdnce in the targeted subject

in regular school.

c?

r

12,

..../

I

I

MET4ODOLOGY

.1-

L4.Samo_le Selection

-r

- C

Youth who voluntarily participated in the athletic-'academic program at

Settlement`during the Fall, 1971, comprised thk experi- mental group.

Non-participants, were randomly Lleoced, made u? the control group. The

random procedure to select non- participants the same at each Kalihi-Palama

school that was attended by a Palama voltitteer. This procedureconsisted of

matching random nUmbcrs room a random number table a List of students in

the same grade as t:c vplunteer. If :he random number.matched a girl wi,t,h a,

male volunteer, that number was discarded and the next one was tried until

each Palama volunteer was matched with a random selectidn of stvdents from the1

game school, in the same grade (grades 6-10), and of the same sex.

Measures of Out9me

1

In order to determine if Palama was reaching its low-achieving target

group, pm-program entrance data were collected from the school records for-10

both,the.experimental and control group students. These data were: school,

attendance records, school math grades, and math and verbal, percentile scores

on the Sequential Tests of Educational Progress (STEP) and the School and

College Ability Test (SCAT).

School math grades and STEP and ,SCAT math scores are measures used.since

the Palama program in the Fall, 1971, focused only ion math skills. Every

other year, the program focuses on math skills; in alternate years, the focus

is on Verbal skills.

Since the focus in the Fall of 1971 was on math, verbal achievement test

scores were collected to serve as a comparison corit-rol for any differences.in

13

r

R' 8

math scores betweer the enperimancal and control groups. It was reasoned-Efi'at

ID

if the Palama pi.ogram had any effect on achievement scores, it should occur

only in the math section and not in the verbal secn.ons of the SCAT and STEP.

Improvement in both the math and verbal sections would negate assuming that

improvement in math was due to Palama's intervention.

Description of Data Collection

All data' were ootained from school records at the schools involved. The

pre-program data were collected for tne Fall, 1970, and the Spring, 1971, since

the program, being examined was in progress In the Fall of 1971. The pre-program

. -.0data were collected in order to compare tne academic level of volunteers with

non-volunteers.

These same data, school.attendance, school math grades, and STEP and SCAT

math and'verbal scores, were collected for the two semesters following the

program (Spring and Fall of 1972) to determine if the already known Palama im-

provement generalized to post-program school performance.'

Attendance records were avagable for 34 Palama students who were enrolled

in'school both before and after their participation in Palama. Math grades

were available for 28 of those 34 students as not Al 34 were enrolled in pre-

and post-Palama math clagses. The number of students who had taken both pre-

and post-program STEP and SCAT tests numbered 12. Since complete data were

40"therefore not available for all students, numbers of students fange from 12 on

achievement test comparisons to 34 students for attendance comparisons.

Statistics Usedoe

The t-test was used to measure differences between the control and experL-

mental group pre-Palama (Table 1) and to measure the differences between the two

14

9

groups on school performance after Palama (Table 3).

The Pearson product moment correlation coefficient was used to determine

if there was a relationship between tie performance of students in'the Palana

program and their post-Palama performance in regular school. This was intended

to determine whether the mach gains made in Palana corrdlated.with post-program

school performance.

4

15

IP

This study is desizned tr ans.er two qt.estionT% firs.: is wnet'ler :ne

Palama program' is reac-ini. lcw-ac-tevinz tarze,t. grouo and tne second Is

whether invclve7ent in the Palama program will impro,...e pcst-proczram sonool per-

formance. For cleric:. tnese twr q_estions are dealt separately; tre

former question is discussed firs: and t-ne :atter qestion follows.

Sr.:dents Is rc.ac-:n z its' -arse_

Tne Palama :7et-t/e-ent academ:c-3: 1TEt.0 orogrF-- is essentially Si -Ed at

low-achieying su-dents .-setause tney a andidates to dm? Ott of school.

As discussed earlier, tne drop-ol,t a pr. e tial delinquent, .this potential

drop-out group of low-at'hievers is alama tlement's target (group. In order

to determine if this group of low- ,niever was volunteering for the program,

pre-program entrance data on scno absen*

scnool at ades, and STEP

and SCAT test scores were collected for e two semeters' rior to the program.

The purpose of this study, in part then, is to determine f youth who volunteered

for the program differed'from non-volunt ers, andif so, on which measures.

The experimental group wou.ld have tb be significan ly lower than the

[control group on at least some of the me sures listed i .Table 1 for Palama

to be considered as successful in attracting the low-a ieving target group.

Pre-program differences between Palama volunteers and e control group of non-

/volunteers are presented in Table 1..

'Since no predictions were made, the ata in Table were analyzed by two

tailed t-tests. This analysis revealed that the eXperirbenta/ and control

groups differed significantly only in scho41 attendance. This shows that the

Pa lama project was rot reac 'hing its low-achieving target group.4

16

Table 1

Means of Pre-Program School ?erfoimance for the

44.

1970-71 School Year

Measures Experimental .. Control

X S .D . v S .D . Lev. of Sig.

Days Absent 13.4 12 -7.1 8%8 34 2.70 .01

Math Grade' 7c. 5 71 8 28 .31 NS

STE? Math 45 Z8 3.6 29 12 .78 NS

SCAT Quant 44 32 38 27 12 .55 NS

STEP Beadinn.

SCAT Verbal

(.9

.,-.

28

27

L.C,

38

224

12

12

.8C

.15

NS

NS. .

eGrades were converted fror let:ers to numbers as foil A . 90,

.

B .. 80, C = 7':," D = 60, E = ':-., wit-. a "- or ---'' counting for 5.

Gevralization rf Pala7a Results

The second question in this study is Inetner the math gains made as a

result of the program generalize to post-program school performance. This is.

assessed by two different methods, first by comparing the means of the post-

program STEP and SCAT (Table 2) for both the experimental and control group,

and second by correlating; for.the experimental group only, program math gains.

and post-program school performance. (Table 3).

Table 2 presents data to assess whether the size of the program math gains

was sufficient to raise the SCAT and STEP.math percentile scores ofthe expexi-

mental group in relation to the control group.

4Tatire, 2

4..4

Means of Post- Program School Performance on STEP and SCAT Math Tests

Measures Experimental Control

X S.D. X S.D. Lev. of Sig.

STEP Math 42 34 39 12 , .28 NS.

SCAT,Oluant. 42 34 32: ao .ps

These data were analyzed using one-tailed' t- tests. "None of the differences

were statistically significant between the expert entel and Control groups. It-

.

17 9

12

is apparent that the aveoage gains made by students in the experimental group

eere not lerge enough to make their percentile SCAT and STEP math scores signi-

ficantly greater as a whole than these students in the control group.

The second way to assess tne impact of the Palama ppgram (Table 3) is to

correlate the program matt: gains with post - program school attendance, achieve-

went tests (STEP and SCA: taken at regular school) and school math grades.

It was assumed t^at tre results pf tne program would generalize to future

school performance. For this as'sumption to be supported, program mah.,-gains

and post-program bErformance on tre --atn sections of the SCAT and STEP, school

math grades, attens'ance tt correlate at a significant level.

Since the. Palama program focusedl only on at skills, it was_ not' expected that

SCATand STEP verbal scores would ;nave a.significant correlation (a control to.

testthat any relationsnip to test performance is specific to the subject

studied in the Pajama program). Data appropriate for this analysis are school

absences, school math rrades, and STEP and SCAT test scores for all of theft

I

experimental group youth for. the mwt semesters following the Palama program.4

i

To, be certain that xorrelati9os made on the basis of these post-program

measures would not,be affected bi differences in pre-program scores all measures

were statistically adjusted for any'pre-score numerical differences. In this

way, any raMiining differences in the poit-scores cannot be attributed to ini-

tial, i.e., pre-score, differences. Thus the correlation here is actually

measuring the relationship between improvement in the Palama program and improve-;

went On the school SCAT and U.EP tests.

Pearson prod;Jct moment 'Ct)frelations betwaen,a4justed math gains made in

the Palama program and adjusted po3t-program school performance measures are

shown Table

4

13

Table 3

Correlations between Program Math Gams and Post-ProgramSchoolPerformance Measures for the Experimental Group

Measure

Ccmputationsa

Correlations

Lev. of

Sig. N

Lev. of

Sig. Concepts

Days-absent, Spring .29..' .NS .24 . NS 34

Days absent, Fall .02 NS . ,-.05 NS : 34

Math grade, Spring -.11 NS .03 'NS 28Math grade, Fall -.01 NS .07 NS 28

STEP Math .5 .35 .73 JDI. 12

SCAT Quantitative .54 .05 .39 .10 12STEP Reading -.3 NS .10: NS 12

SCAT Verbal . ,'' 'NS -.?8 NS 12

aCompttations and Concepts -were the. math- areas programmed andmeasured during tne Palama program.

As shOwn to Table 3, Palama program math gains generalize to future aca-

demic performance as measured by STEP and SCAT math tests taken at school.

One can,-therefore, predict from imprfvement in PalamA that there will tend to

be an iffproilement on tests in school '(STEP and SCAT). As expected, post-Palama

SCAT and STEP verbal and reading scores did not correlate with math gains made

. -

in Palama. Therefore, it is apparent that generalization to post-Palama

achievement test performance is specific to the subject studied in the Palama

program. Program math gains, however, did notgeneralize to either school

math grades or school attendance.

19

fONCIASIONS AND'IPLICATIONS OF STUDY

Major 'Conclusions

One of the questions in tATiS study is whether or not the Palasa program

results generalize to school perforTancr\in the targeted subject, matn. Results

in Table 3 reveal that there is a correlation between math gains made in the

program and performance on khe'ath sections of. the STEP and SCAT post-program

achievement tests tat:en in scnool. The s:gnifiCan.t.correlacions show that the

higher the 7ath score in Pa:a7a, the hiz-er :he STEP and SreA -:ath performance

in Table 3. he correlatIors .7cct:rred math and not tn the reading

tests of the STEP and SC:-T, an indication trat i7provement was likely a function

of tne PalaTa math program and not attrio..table'to other factors. Further,

since all scores were adjusted for their pre-program levels, this result is not

likely a factor that could be attributed to prior differences (e.g., "smart"

students do better at Palana and at school).

The lack of difference between the mewls of the experimental and controlA

groups (Table 2) on the postlprogram STEP and SCAT math sections indicates that

the average gain made,ip the program is not sufficient to raise a participant's

math performance relative to other students'. In other words, the control

group and experimental group means are,not different (Table 2). Since the

coerelatioirl results show that the STEP and SCAT are sensitive to the math

gains made in the program, It is reasonable to assume that longer programs which

resulted in larger gains might increase a participant's math p4rfotmance rela

tive to non-participating students.

These results indicate the-importance, for programs designed to improve

school performance, of assessing the relationship between performance in the

special program and performance in school. The assumption that program results

20

0

15

showing an improvement rn dcademic performance will automaticelly be reflected

in school performance is riot tenable.

A second question this study sought to answer 4.as whether or not the

Palema Settlement Behavior Modification Sports .Program was reaching its low-

academic achievemerit target group: The data from Table 1 reveal that essentially

there is no difference between the experimental and _control group when they

enter the program, except that the experimentafgroup is absent 5.3 more days

a year. Slnce the'control and experimental groups are very much alike, academic

gains achieved in the program cannot be attributed to brighter students parti-

cipating in the Palama program than are enrolled in the regular school.

Therefore, the Palama program is reaching students similar to a random sampling

of studedts from the same schools, but not the targeted low-achieving group.

The students in the Palama program are neither higher nor lower in academic

achievement than the control group.

If it is desirable for the program to reach low-achieving youth, it may

be possible to screen out the volunteers with higher grades and achievemeitt

test scores, By eliminating the relatively higher academic achievers, it may

be possible to enroll more low-achievers. One advantage of having Palama youth

similar to studerits enrolled in regular school is the ease with which these

results would then generalize to similar public school students should the

Department of Education wish toincorporate a program such as this in the public

sdhools.

In conclusion, the two major findings are that the program results (i.e.,

improvement in math) are discernible on the post7program STEP and SCAT math

sections, but do'not show up in post-program school math grades. Secondly, the

Palama program is reaching neither a high nor low achieving student, but one

which isthe Berrie 4s a random sampling of students (control group) Jn the same

grades and from the schools attended by the Palama program youth.

21.

ti

lE

Relation to (7ther studies

It ts per-aps s-rprtsIng za:n:,.=ade in t-e Fa:araa progra7 crrrela:

with i=prove=ent in .-ztr abilLty nave .ro cffec.t Jr at perfor=ance, i.e.,

school math graded. :1::wevez, sintlar -ave;iee:*. reported b.: Tharp and

Wetzel , and Fa and :.-DrnnA11.L because of.grades being more

a tunction cf the teas -.ers. expectatt=s :-an of the pLcil's actual ability

(Rosenthal and :aco:seh).1$3 In .7;.osentha: a.d :acc'csens study, teachers'

expectancies wer4. -; =- -er' 7^ :f

the la classes, an a,er,4e ,re tr t-e experir-ental

condition. Teachers ...ere to.d t-cs, s -.o unusual intelLec-

tual gains during _re schooryear. ' 71-1(: stt:dertts .tens actually selected by a

random procedure. These children'showed a significantly greater gain in IQ

scores than did the control children. Tie lower the grade level, the greater

the effect.

The factor of teacher expectancy may be a posible explanation for the

lack of carry -over of Palama gains to the public schools. It maybe that for

students who have not done well in the past, encouragement for work well dope

may be lacking. Students may pick up cues that the teacher doesn't think they

will do. really good work. In large classes especially, since these teachers

may not be aware they give these cues, attention pay be given to those students

who have proven themselves to be high achievers,with a corresponding lack of

attention to the 1pw achievers' If students have demonstrated that they can do

well when positivelyereinforced by something that is of value to thep (athle-

tics), the ability to do the work has proven itself to be there.

The Wry, Giles and Hall13 studysmentioned earlier-Anclud the regular

school teachers in their program by having students bring in hbmework 'rom .

. .,"

regular achbol to do and gain, points: They .also allow these same,teachers to

22

S.

17

add points for appropriate academic behaviors. In this way, the regular school

teacher is a part of the program and can see the gains made by the student in

the special program. The Palama program might do well to consider involving

the regular school teachers, if this is possible.

Studorts could earn points at Palama for completing school homework assign.-.

ments and for performance on school tests. Teachers would then become aware.Iff

the program and see it as an aid both to themselves and to the students. This

might bridge the gap between tne Palama performance and the school performance.

It might also affect trc presume "attitude factor' by giving teachers a

chance to see low - achiever's achieving at Pala:7a. A s:nilar method could be

used to affect'school attendance. Points could be given at Palama for school

attendance as indicated by school records. This might be done especially for

students with a high absenteeism rate.

Implications

Previously, Polk and Schafer ware quoted as saying, "that educational

'failure - by schOols as well as by students - is directly related to delin-.

quency.',49 Palama Settlement deals diree.tly with youth from a low-income

urban area who are possible future candidates fbr delinquency. The purpose

of the.programhis naoTfold; first, it-is simply to iltprove math performance

fdir its own sake, and second, ,an' improvement' ice math skills, as a result of the

program is assumed to carry over into regular school math performance: There,

success in copingyithmath would hopefullylmake school less aversive by tn-r

creasing successes and thus reducing chances' for droppt out.

They provide reinforcement (athletics) for prosocial behavior (studying,

team cooperatOn, renouncing of criminal activities) and block rewards for

antisocial behavior (crtsle: ignoring studiet, noncooperation). The program has

23

18

shown itself to be worthy by demonstrating, through the use of a token economy

system with athletics as tt?e reinforcer, that low-academic achieving students'

can be raised 0.5 to 1.2 grade levels in 10 weeks as measured on the California

,

and Stanford Achievement Tests. The present study has shown that although

students rake these gains, the improvement does not show up in post-program

school grades.

The implications for further innovation and study are cleaz. Perhaps

students could be partially reinforced between seasons in order to maintain some

of the academic behavlors,gained. Points could be earned between seasonA.for

grades or attendance be applied 7,p 7-e regular athletic-academic seasons

Also, these seasorz might be extended, or a third sports season might be added.

The need for carry-over is important if students are to succeed and remain

in school, hopefully less prone to delinquent problems. If, as studies have

shown, the drop -our is.more prone to delinquency, every effort must be made to

assure hiS remaining and finding success at school.

In conclusion, two points are stressed: 1) academic success in a special

program does not automatically result in academic improvement in school, and

2) if the results of such programs do not generalize to school performance,

then they clearly cannot be expected to prevent delinquency (since their ration-

ale is that such prevention occurs by improving school perfOrmancer

Two general changes need to be made as a result,of this study. First,

efforts to obtain generalization to school performance should be an integral'

part of every special academic program and secondly, data on future delinquency

are needed to evaluate the^assumed effects of such programs. These results

indicate that such untested assumptions may be unwarranted. These implications

not only apply ta the program at Palama Settlement but also to the many other

academic programS in Hawaii and the Mainland.

24

19

REFERENCES

1. Kenneth Polk and Waiter E. 5chafer Schools and Delinquency (New Jersey:

Prentice-Hall, Inc., 1972), p. v.

2. Paul S. Graubare., ed., Children Against School; (Chicago Follett

Educational Corporation, 1969), p. 329.

3% Clifford R. O'Donnell, Earlene Chambers and Kenneth Ling, "Athletics as

Reinforcement in a CommUnity Program for Academic Achievement', Journal

of Community Psychology, 1 (Apr_', 1973). p. 212-214.

4. Polk and Schafer, ca. cit., p 5.

5. Polk and Schafer, oo.

6. Pclk and Scnafer, a7_'. cit.

7. Lear T. Empey and' Steven G. Lube6. Ronald L. LaPorte, Explaining

Delinquency (!s:assachuse:ts: Heath Lexington Books, .1971), p. 77.

8. W. W. 'ienrich, A Primer of Behavior Modification (California: Brooks-

Cole Publishing Company, 1970), p. 37.

9. Harold L. Cohen and James A. Filipezak, "Programming Educational

Behavior for Institutionalized Adolesc'ents", in Behavioral Intervention

in Human Problems,,ed. by Henry C. Rickard (New York: Pergamon Press,

Inc., 1971), (. 179.

10. Clifford R. O'Donnell and Jean 1. DeLeon, "A Contingency Program for

Academic Achievement in a Correctional Setting: Gains and Time in

School",' Journal of Community Psychology, l'(JUly , 1973), p. 285-287.

11. Sohn M. McKee and Carl B..Clements, "A Behavioral Approach to Learning.

the Draper Model", in Belidvioral Intervention in Human Problems, ed. by

Henry C. Rickard (New York: Pergamon Press, Inc., 1971), p. 201.

12. Clifford R. O'Donnell and Kathleen Stanley, "Paying Students for Academic

PerformanCe: A Demonstration Project", Journal Of Community Psychology,

1 (April, 1973), p. 215116.

13, Montrose M. Wolf, David K. Giles and R. Vance Hall "Experiments with

Token Reinforcement in a Remedial Classroom'; in Children Against

Schools, ed. by Paul S. Graubard (Chicago: Follett Educational

Corporation, 1969), p. 307.

14. Social Welfa e Development and Research Center, Mew Approaches ana'

Innovations Palama Settlement, Author, 1973.

15. O'Donnell, Chambers and g.. op. zit.; p. 212-214.

16. Roland G. Tharp and fialph J. Wetzel, Behnvior Modification in the Natural

Environment (New York: Academic Press, 1969), p. 16p-169.4

/`

20

! ,

17. W.-S..0. Fo-and Clifford R. O'Donnell, "the Buddy System: Relationship

and Contingency Conklitions in a Co=unity.Intervention grogram Lor

Youth withAon-professionals as Behavior Change Agents"_, Journal of

Consulting and Clinical Psychology, 1974; p: 163-169.

18. Robert Rosenthal and Lenore Jacobsen; "Teachers' Expectihcies: Determinants

of Pupils' IQ Gains", Psychological Reports, 19(1966), p. 115-118.

19. Polk and Schafer, op. cit., p.

A.

r

2 6

4

a

21

' BIBLIOGRAPHY.3

-'ar

Empey, LaMar and LubeCk, Steven G. with LaPorte, Ronald L.Explaining Delinquency: 1,aisachusetts: Heath Lexington Books,'1971.

Grauiard, Paul S., ed.Children Against Schools. Chicago: Follett Educational...C=2.z_, 1969.

O'Donnell, Clifford R., Chambers, Earlene gnd Ling, Kenneth."Athletics as Reinforcement in a Community Program for AcademicAchieVement", Journal of Community Psychology, 1 (April, 1973), v. 212-214.

-O'Donnell, Clifford R. and DeLeon, Jean L.

Contingency PrCgrar for Academic Achievement in a Correctional ,Setting:

Gains and Time in School', Journal of Community Psychology, 1 (July,,1973), p. 285-287.

O'Donnell, Clifford R. and Stanley, Kathleen. \

"Paying Students. for Academic Performance; A Demonstration Project";

Journal of Communit> Psychology, 1 (April, 1973), p. 215-216.

Polk, Kenneth and Schafer, Walter?Schools and Delinquency. New Jareey: Prentice- Fail, Inc., 1972.

Rickard, Henry C..ed.Beh.i,Ctioral Intervention in Human Problems. New York: Pergamon Press,

Inc., 1971.

Rosenthal, Robert', and, Jacobsed, Lenore.

"Teachers' Expectancies: Determinants of, 'Pupils' IQ Gains"; psycho-

logical,Reports, 1966 19 {066), 115-113.

Social Welfare Development and Research Center.New Approaches and Innovations at Palama Settlement. Author, 1973.

Tharp, Roland G. and Wetzel, Ralph J.Behavior Modification in the Natural Environment. New York: Academic

Press, 1969.

Toby, Jackson and Toby, Marcia.Low School Status as a Predisposing Factor in Subcultural Delinquency.U.S. Office of Education, Cooperative Research Project No. 526, 1961.

Wenrich, W., W.

A Primer of Behavior Modification. California: Brooks/Cole Publishing

Co., 1970.

27

't

APPENDIX A

PALMA SE'ETLEKEWI''BOARD OF TRUSTEES

1974 .

OFFICERS

President - Dr. George Schnack1st V. P. - Gary Taki

2nd V. P. - John MurphySecretary - Mrs. Y. Shimamoto_Treasurer - Jtaroz4e Tarutani

HONORARY BOARD MEMBERS

George R. CarterCarter Galt

A. L. Y. Ward

BOARD MEMBERS

.

Martha Almon Jack NagoshiInes Cayaban .Raymond OshiroWillie Chang Vladimir OssipoffGilbert E. Cox *--Patt-ratterson

Frank C. Farinas Karen PerkinsFuifatu Fauolo Lena ReverioNora Furuno Major Roland SagumPeter Iha Dr. George SchnackVirgil J'Hoo Yoshiko ShimamotoWalter Kupau Gary TakiPeggy Melim Jerome Tarutani

Matilda Molina Roger S. Toguchi

John Mur0y Joette WheelonJohn C. Wright

BOARD ADVISORY COUNCIL

John ClarkeHenry DamonCecil Heftel

Robert Rath, Sr.A. L. Y. Ward

2 ik

APPENDIX B

FOOTBALL 1972

-STAF1'ING

Executive DirectorProgram DesignerProject CoordinatorLearning EngineerBehavior Modification Consultant

Robert HigashinoEarlene Chambers

. Kenneth Ling

Joyce TanjiRobert Omura

Social Welfare Developmentand Research CenterUniversity of Hawaii

James SwensonSharon Apo

Valerie ButcherKeithiChang

Hairy Churl -Hoon

Bob Cole

Doug CrosierDuval DutroClayton HigaCarol Koshi

Lance MattsonGlenn OgawaNelson TaniVernon Wong

Jeanette HirakawaTrinidad MaliaBertha Maunupau

Rosalind Nakagawa'Violet Viiohia

Catherine UchiharaKarel Ling

Vernon Wong

Neighborhood.Development Supervisor

AcademiC Coaches

Training Table Staff

Cheerleading CoachMidget A Bead CoachMidget A Assistant Coach Dennis Okimoto

Midget B Head Coach . Doug Crosier

Midget B Assistant Coach . Bob-Cole

Bantam Head Coach . . Kenneth Ling

Bantam Assistant Coach . Lance MattsonDuval Dutro

-

2 9

4

Executive Director.

APPENDIX C

BASKETBALL 192

STATING

Program Designer.Project CoordinatorAcademic CoordinatorBehavior Modification Consultant

Medical ConsultantDental ConsultantPhotographerNeighborhood DevelopmentAcademic Coaches

Supervisor

. Robert HigashiioEarlene ChaMbers

Kenneth LingKarel Ling

Robert OmuraSocial Welfare Development

an.- Research CenterUniversity of Hawaii

B. Allen Richardson, M. D.Ping Kse Yee, D.D.S.

. . ... .`Warren HooJames Swenson

,Sharon Apo

Glenn BlindePatty Buchanan_Janice Ching

Douglas CrosierSharon Cummings

Duval Dutro'Jeanne _Fleming

Trina GooErik Harada

Gerald HasselyDebbie Ho

Joy JasinskiDorothy KimKarel Ling

Lewis LillieClifford Lee

Ron Loo---hie /1 LumHoy

Darlene Matsumoto,,Bonnie Miura

Heather MiuraScott Nakagawa

Glenn OgewaNora Oyama

Merfalyn PhillipsSandra RasmussenRonald Rodrigues

Helene RyanMarie Schulz

John SharpGailSouza

Robert Stender. Joyce Tanji

Vernon Wong

.........

30