Developmental guidance for school success skills

MISSING IMAGE

Material Information

Title:
Developmental guidance for school success skills a comparison of modeling and coaching
Physical Description:
viii, 216 leaves : ill. ; 28 cm.
Language:
English
Creator:
Cuthbert, Marjorie Irene, 1947-
Publication Date:

Subjects

Subjects / Keywords:
Educational counseling -- Psychological aspects   ( lcsh )
Students -- Attitudes   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1987.
Bibliography:
Includes bibliographical references (leaves 209-215).
Statement of Responsibility:
by Marjorie Irene Cuthbert.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 16881836
ocm16881836
System ID:
AA00011208:00001

Full Text












DEVELOPMENTAL GUIDANCE FOR
SCHOOL SUCCESS SKILLS:
A COMPARISON OF MODELING AND COACHING











By

MARJORIE IRENE CUTHBERT


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMEeTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1987














ACKNOWLEDGMENTS


I would like to extend sincere thanks to Dr. Robert Myrick for

serving as chairman of my ccmnittee. His professional suggestions and

personal encouragement were invaluable to me and are greatly

appreciated. I would also like to thank Dr. Larry Loesch for serving on

my committee. His careful attention to detail and high expectations

elicited my best efforts in all aspects of this study. I also would

like to thank Dr. Cecil Mercer. His support and helpful input

substantially contributed to the study. Thanks are also extended to

Drs. Athol Packer and Esther Morgan for serving on the camnittee.

I would also like to thank Dr. Wiley Dixon, Director of Guidance,

Alachua County, Florida, Public Schools, for his support of the project

and his sincere interest in the results. Special thanks also go to the

school counselors who presented the units and carefully carried out the

research procedures.


Additional thanks are extended (a) to my husband, Bruce, for his

constant love and support throughout the entire graduate degree; (b) to

my daughter, Kristina, for her unselfish help and sincere desire to

assist me; (c) to my son, Scott, for his continuous belief in his

mother; (d) to my mother- and father-in law for their assistance with

data entry and help with the children; and (e) to my loving mother and

late father who willingly made innumerable sacrifices for me and who

always thought I could do anything I put my mind to doing.

ii















TABLE OF CONTENTS


Page


ACKIM LEDGME2NTS. .. .

LIST OF TABLES ... ...............

ABSTRACT .......... .. .............


ii

v

vii


CHAPTERS


I INTRODUCTION ........ ..............


Theoretical Perspective .
Statement of the Problem..
Need for the Study .....
Purpose of the Study .
Research Questions. .
Definition of Terms .
Overview of the Remainder of


the Study.


II REVIEW OF RELATED LITERATURE. . .

School Success Skills . .
Developmental Classroom Guidance. . .
Target Students .. .................
Modeling. . . .
Coaching.........................
Outcome Measures...................
Summary ................... ...

III METHODOLOGY ....................


Population .............
Sampling Procedure .
Resultant Sample . .
Research Design . .
Treatment Descripticn .
Assessment Techniques .
Hypotheses ..............
Research Procedures . .
Research Participant Training .
Data Analyses ............
Methodological Limitations .


iii









IV RESULTS ........ ... ..

School Success Classroom Behaviors.
School Success Classroom Attitudes.
Student Summary Sheet .
Summary ..............

V SUMMARY, CONCLUSIONS, DISCUSSION,
LIMITATIONS, IMPLICATIONS, AND
RECbOMMENDATIONS . .
Suay... .IOS .........

Sumnmry.. . .
Conclusions . .
Discussion ..........
Limitations ............
Implications. . .
Recacmendations . .


APPENDICES


A SKILLS FOR SCHOOL SUCCESS MODELING .

B SKILLS FOR SCHOOL SUCCESS COACHING .

C SCHOOL SUCCESS INVENTORY TEACHER FORM .

D SCHOOL SUCCESS INVENTORY STUDENT FROM .

E SCHOOL ATTITUDE INVENTORY . .

F DATA ANALYSIS SUMMARY TABLES .

G STUDENT SUMMARY SHEET . .

H STUDENT FEEDBACK SUMMARY. . .

REFERENCES . . ..

BIOGRAPHICAL SKETCH. .................










107

129

149

151

153

159

206

207

209

216


66

67
81
89
89




93

93
95
98
103
104
105


. .


S .


















LIST OF TABLES


Tables Page


3-1 Number of Students per Condition by Schools 53

3-2 Pre-Post Control Group Design . 55

4-1 Pre-Post Mean Change Scores for Total School
Success Inventory Teacher Form (SSI-TF) 68

4-2 Pre-Post Mean Change Scores for Total School
School Success Inventory Teacher Form (SSI-TF),
by Condition and Individual Schools . 70

4-3 Target Students: Pre-Post Mean Change Scores
for Total School Success Inventory Teacher
Form (SSI-TF) ................... 72

4-4 Pre-Post Mean Change Scores for Total School
Success Inventory Teacher Form (SSI-TF), for
Conditions by Race ................. 73

4-5 Pre-Post Mean Change Scores for Total School
Success Inventory Teacher Form (SSI-TF), for
Conditions by Gender ................ 75

4-6 Pre-Post Mean Change Scores for Total School
Success Inventory Student Form (SSI-SF) 76

4-7 Target Students: Pre-Post Mean Change Scores
for Total School Success Inventory Student
Form (SSI-SF) ................... 78

4-8 Pre-Post Mean Change Scores for Total School
Success Inventory Student Form (SSI-SF), for
Conditions by Race .. ............... 79

4-9 Pre-Post Mean Change Scores for Total School
Success Inventory Student Form (SSI-SF), for
Conditions by Gender . . 80

4-10 Pre-Post Mean Change Scores for School Attitude
Inventory (SAI), Total and by Dimensions 82









4-11 Pre-Post Mean Change Scores for School Attitude
Inventory Scores of Individual Schools for Total
Score and Dimensions ................ 84

4-12 Target Students: Pre-Post Mean Change Scores
for School Attitude Inventory (SAI) ... 87

4-13 Pre-Post Mean Change Scores for Total School
Attitude Inventory (SAI), for Conditions by Race 88

4-14 Pre-Post Mean Change Scores for Total School
Attitude Inventory (SAI), for Conditions by Gender 90















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


DEVELOPMENTAL GUIDANCE FOR SCHOOL SUCCESS SKILLS:
A COMPARISON OF MODELING AND COACHING

By

Marjorie Irene Cuthbert

May, 1987

Chairman: Robert D. Myrick
Major Department: Counselor Education

Two developmental guidance units were evaluated for their effects

on teachers' perceptions of student school success behaviors (School

Success Inventory, SSI-TF), on students' self-reported school success

behaviors (School Success Inventory, SSI-SF), and on students'

attitudinal ratings of school situations (School Attitude Inventory,

SAI). Six school success skills were presented and included paying

attention in school, listening to teachers and peers, volunteering in

school, using self-control, interacting with teachers and peers, and

utilizing self-assessment techniques at school. The units were

delivered by five state certified school counselors to third-grade

students in 15 classrooms in public elementary schools.

A randomized pre-post control group research design was used to

compare students in three conditions. Students in the modeling

condition (n=102) learned vicariously as they watched peers demonstrate

the skills. Students in the coaching condition (n=101)

vii








learned the skills directly by listening to the counselors; students

then practiced in small groups, where they coached each other's

performances of the skills. The control students (n-96) did not receive

the units until the study was completed.

Although an overall analysis of covariance (ANCOVA) on total SSI-TF

scores revealed no significant differences with respect to conditions,

subsequent ANCOVAs by individual schools showed significant differences

(p <.05) among the groups, with students in the coaching group improved

in three of the five schools and those in the modeling group improved in

one school. The students' corresponding perceptions of their behavior

did not show consistent improvement.

Scores on the overall SAI indicated a significant difference in

attitudes toward school for students in the coaching condition, as

compared to the modeling and control conditions. Further analyses also

indicated a significant effect for attitudes on the excitement-calm

dimension of the SAI, with students in the coaching condition rating

themselves as feeling more calm about school situations following the

guidance unit. It was concluded that developmental guidance units

designed to teach school success skills, especially one that included

coaching, can influence student attitudes about school situations.


viii















CHAPTER I
IIIRODUCTION


Traditionally, schools have concentrated on teaching children basic

academic skills of reading, writing, and arithmetic (Cartledge &

Milburn, 1978; McGinnis & Goldstein, 1984). Where time and money were

available, schools expanded academic subjects and taught skills in areas

such as athletics and the arts. However, there has been a noticeable

lack of teaching the task and interpersonal skills linked directly to

achieving success in the school setting.

Snme students come to school seemingly knowing what skills to use

to become successful. They understand what teachers mean and respond

when teachers say, "Class, pay attention!" or "Class, listen carefully

now." Other children, who are not as aware of such school success

skills or responses, often become frustrated and achieve less.

Direct instruction in skills that enhance a child's success at

school is typically absent in most schools. When offered, instruction

in school success skills has usually been fragmented, unsystematic, or

incomplete (Cartledge & Milburn, 1978). Administrators, school

counselors, parents, and teachers at each grade level have often assumed

that someone was instructing students in the school success skills that

they would need, only to find that most children have not accumulated

them. The acquisition of essential school success skills could be

greatly facilitated if direct instruction were incorporated as formal,

didactic components of school curricula.

1










Therefore, this study was designed to evaluate the effectiveness of

helping children learn school success skills. Two different

instructional methods (i.e., modeling and coaching) were incorporated

into developmental guidance units presenting school success skills of

(a) paying attention, (b) listening to teachers and peers, (c)

volunteering in school, (d) using self-control, (e) interacting with

teachers and peers, and (f) utilizing self-assessment. The effects of

these units on third-grade students were examined. Target students,

those identified by their teachers as the lowest in school success

skills, received particular attention in the study.


Theoretical Perspective


Behaviors can be learned. Therefore, behaviors can be taught to

children. If school success skills are expressed behaviorally, then

they can be taught to children. Krumboltz and Krumboltz (1972) stated

that new behaviors can be developed through (a) the successive

approximations principle (e.g., reward steps to the final behavior), (b)

the cueing principle (i.e., cue the correct performance before the

action is expected), (c) the discrimination principle (e.g., identify

differentiating cues and appropriately reward them), and (d) the

modeling principle (i.e., expose children to an appropriate person

performing the desired behavior). This study used the modeling

principle for the modeling condition, where behaviors (i.e., school

success skills) were presented by same-age peers to classrooms of

students who learned these behaviors vicariously from the models. This

study also utilized the successive approximations principle in the

coaching condition where peers, after being instructed directly on









specific behaviors (i.e., school success skills) by counselors, rewarded

and shaped each others' attempts of expected behaviors.

After behaviors have been emitted, they can be strengthened through

the positive reinforcement principle where an immediate reward is given

after each correct performance of the behaviors, thus increasing their

chances of occurring again. After behaviors have been established, they

can be maintained by the substitution principle (i.e., reinforcing with

a previously ineffective reward just before the more effective reward is

given) or the intermittent reinforcement principle (i.e., gradually and

intermittently decreasing the frequency of rewarding the behavior)

(Krumboltz & Krumboltz, 1972). These basic principles of behavioral

learning are consistent with those used in the current study. Since the

school success skills are representative of those behaviors teachers

look for in the classroom, and are thought to enhance learning, it is

assumed that teachers will attempt to maintain these behaviors with

positive reinforcement and that internal reinforcement will also occur

with students as they experience success.

Cartledge and Milburn (1978), in reviewing related literature,

noted that the acquisition of certain skills may be crucial to the

academic experience and to overall individual school success. Skills

which have been linked to school success from previous studies and

identified by teachers need to be taught to children. It appears that

children who are taught specific school success skills are more

productive and achieve more (Cartledge & Milburn, 1978). It also

appears that students will rate themselves more positively on

attitudinal statements related to school success as a result of










participation in classroom guidance units (Gerler & Anderson, 1986;

Myrick, Merhill, & Swanson 1986).


Statement of the Problem


Modeling, by which new behavior is learned vicariously after

viewing a model's demonstration, has been used successfully in a variety

of situations as a mode of instruction to teach new skills or behaviors.

Further, it has been shown to be effective with specific target

populations (Bandura & Barab, 1973; Evers & Schwarz, 1973; Gottman,

1977; Keller & Carlson, 1974; O'Connor, 1969; Sarason, 1968). However,

only a small proportion of studies has been carried out with

"unselected" elementary school populations (Conger & Keane, 1981). The

importance of teaching skills to children, particularly those

interpersonal and task skills related to school achievement, has been

well established, and the attainment of those skills shown to be crucial

to overall school success (Cartledge & Milburn, 1978). However, it is

not known if the use of modeling to teach a set of school success

skills, when they are presented in developmental guidance units, is

effective with groups of students in unselected elementary school

classrooms.

Coaching, where new skills are presented directly and practiced

while being coached or shaped by others, also has been shown to be

effective with isolated and withdrawn children both in preschool and

elementary age ranges (Bornstein, Bellack, & Hersen, 1977; Conger &

Keane, 1981; Gottman, Gonso, & Schuler, 1976; Gresham & Nagle, 1980;

Ladd, 1981; LaGreca & Santogrossi, 1980; Oden & Asher, 1977). It has

also been shown to be effective with learning disabled students










(Schumaker & Hazel, 1984). However, the effectiveness of coaching has

not been evaluated with groups of students in classrooms.

Developmental guidance units, where skills are presented by a

counselor or teacher to classes of students, have been shown to be

effective in teaching interpersonal ideas and skills to students (Cobb &

Richards, 1983; Cuthbert, 1984, 1985; Duncan & Gumaer, 1980; Gerler &

Anderson, 1986; Myrick et al., 1986). However, it is not known if

developmental guidance units can be effective with young students where

modeling or coaching are used to teach school success skills.


Need for the Study


There is a need to teach students skills that can lead to success

at school. Staggering numbers of youth continue to drop out of school.

Gadwa and Griggs (1985) noted that the 1979 Carnegie Council on Policy

Studies in Higher Education pinpointed the high school dropout rate as a

major problem facing youth and suggested that educators should give

serious attention to reducing the number of high school drop outs,

reducing rates of -bsenteeism, and bettering the high school experience.

According to the U.S Census Bureau (Gadwa & Griggs, 1985), 43.1% of

youth were not enrolled in school or did not graduate. Elementary

school is a critical time to attend to instilling school success skills

which may contribute to keeping youth in school. Reports dealing with

high school dropouts suggest that the causes can be traced back to

elementary school levels, where problems can be assessed as early as the

third grade (Schreiber, 1967).

Another major problem area for educators continues to be disruptive

behavior and student discipline (Bleck & Bleck, 1982; Cobb & Richards,









1983; Frith & Clark, 1983; Sorsdahl & Sanche, 1985). A steady increase

in personal, social, and behavioral problems has been seen for students

in regular classrooms during the 1970s, and 13 of 14 annual Gallup polls

have identified discipline as the most important problem in education in

the United States (Sorsdahl & Sanche, 1985). Bleck and Bleck (1982)

also stated that disruptive behavior can adversely affect relationships

with peers and teachers. If presenting school success skills to

children proves to be effective, disruptive behavior may be decreased,

thereby lessening the emphasis on discipline and increasing students'

chances for positive interactions. There would be more time for

productive learning, and teachers would have more time to teach all the

students.

In addition to the enhancement of the classroom environment, the

learning effectiveness and efficiency of individual children can be

increased if developmental guidance units improve specific school

success skills for children. For example, if children can learn to

control their aggressiveness, their achievement may increase. If

children learn skills to overcame the anxiousness that often accompanies

raising their hand to share ideas and are able to increase the number of

times they do so, they may also affect their teachers' positive

perceptions of them. If children learn skills to interact with peers

and teachers, they will feel more positive about themselves and have a

better opportunity for academic achievement (Cartledge & Milburn, 1978).

If the modeling condition were shown to be the more effective way

to teach certain skills, teachers or counselors could be trained to use

this technique to introduce new skills and concepts across many subject

areas. Because most new skills are presented didactically in schools,










the addition of models to illustrate the content would add the dimension

of vicarious learning to direct learning, thereby improving teaching

activities.

If the coaching condition proved to be an efficacious way to teach

new skills, then school counselors and teachers could provide more time

where students are coached as they perform new behaviors. Skills such

as interacting with teachers and peers, paying attention to teachers,

and using self-control are too often talked about in an abstract manner

and rarely practiced by the students as part of a specific lesson on

school success skills. Some children do not understand what they are to

do; consequently, they became frustrated and could become disruptive or

engage in misconduct. If coaching of skills, which in this study means

being coached by members of small groups, were shown to be effective,

teachers could use this technique to enhance learning in their classes.

If the effects of these conditions were known, same practices

within the educational profession also could be revised. School

counselors who have chosen to work with children on school success

skills, either individually or in small groups, might redirect more of

their efforts to other areas of helping children since school success

skills could be effectively presented to entire classrooms. Or, they

would have more time for follow-up with children who need more attention

or practice. Also, if students were able to feel more successful about

school and to have their teachers perceive them as having school success

skills, the number of referrals to school counselors could decrease.

This would free school counselors to create and present other

developmental guidance units or activities.










If guidance units using modeling and coaching proved to be

successful in improving school success skills, school counselors might

use similar units at different grade levels. They also might act as

consultants to teachers, alerting teachers to the skills that facilitate

school success. Counselors could also train teachers in the use of the

modeling and coaching techniques so that the teachers could teach skills

directly to their students. In addition, more impetus to provide school

success skills within approved curricula for schools at local, state, or

national levels might be possible.


Purpose of the Study


The purpose of this study was to determine the effectiveness of two

elementary school developmental guidance units designed to help children

learn school success skills and improve their attitudes about school.

More specifically, these two classroom guidance units, one incorporating

modeling and the other incorporating coaching, were presented to

third-grade elementary public school students and evaluated through a

research design which included three groups (i.e., modeling, coaching,

and control). Specifically, answers were sought to the following

research questions.


Research Questions


1. Will classroom teachers report changes in their perceptions of

student school behaviors after the students have completed developmental

guidance units focusing on school success skills?










2. Will the school success skills of target students be perceived

differently by their teachers after the students have participated in

the developmental guidance units?


3. Do race and gender interact with the effects of the developmental

guidance units on teachers' perceptions of students' school behaviors?


4. Are students' perceptions of their school success behaviors affected

by participation in developmental guidance units focusing on school

success skills?


5. Will target students change their perceptions of relevant school

behaviors after participating in the developmental guidance units?


6. Do race and gender interact with the effect of the developmental

guidance units on students' perceptions of their school behaviors?


7. After students receive the developmental guidance units, will their

attitudes about school situations change?


8. Will target students who receive the developmental guidance units

change their attitudes about school situations?


9. Do race and gender differentially affect students' attitudes about

school situations following participation in the developmental guidance

units?


Definition of Terms


Coaching. A technique for instilling new behavior by direct

instruction and practice with shaping by observers. In this study,










school success skills were presented didactically by a school counselor

to students. In turn, students practiced the skill in small groups

where peers within the group assisted in coaching and praising the

performance of the skill.


Control. A dimension of the School Attitude Inventory used for

assessing attitudes about school. The scale ranges from being

controlled by others at one extreme of the rated continuum to being in

control at the other extreme.


Developmental classroom guidance. A form of guidance whereby a

counselor or teacher interacts with students in a classroom to enhance

students' personal and/or academic growth.


Elementary school counselors. Professional educators who are

certified by the state in which they work to provide counseling and

guidance services in elementary schools.


Excitement. A dimension of the School Attitude Inventory used for

assessing attitudes about school. The scale ranges from being excited

at one extreme of the rated continuum to being calm at the other

extreme.


Guidance unit. A series of classroom guidance sessions related to

a particular topic. Each session has a goal and specific objectives

related to accomplishing the general purpose of the unit.


Modeling principle. A technique for teaching children new behavior

by vicarious learning. In this study, school success skills were










presented verbally to students by a school counselor while they observed

another child (model) performing the school success skill.


Pleasure. A dimension of the School Attitude Inventory used for

assessing attitudes about school. The scale ranges from happy at one

extreme of the rated continuum to sad at the other extreme.


School situations. Circumstances in school to which students

respond, including teacher calling on students, asking questions in

class, organizing school work, talking in front of the class, accepting

suggestions, asserting one's self, participating in class discussions,

and accepting teacher's corrections and suggestions.


School success skills. Behaviors (i.e., skills) which help

students perform effectively in school and the classocm.


Target students. Students who were identified by their teachers as

being low-skilled (i.e., those students who were rated a 4 or under on a

scale of 1 through 7 with 7 indicating the most skills) in terms of task

and interpersonal behaviors that students use to make an effort to learn

successfully in the classroom.


Overview of the Remainder of the Study


The remainder of the study is organized into four additional

chapters. A review of the related literature is presented in Chapter

II. Chapter III contains the research methodology, where procedures of

the study are described. The results are presented in Chapter IV.

Chapter V includes a summary, conclusions, discussion, limitations,

implications, and reccnmendations.















CHAPTER II
REVIEW OF RELATED LITERATURE


In 1983, the National Carnission on Education issued a report, A

Nation at Risk: The Imperative for Reform, which declared that American

education was not in step with the technological times:

Our nation is at risk. Our once unchallenged preeminence in
commerce, industry, science, and technological innovation is
being overtaken by competitors of the world. .educational
foundations of our society are presently being eroded by a
rising tide of mediocrity that threatens our very future as
a nation and a people. (National Commission on Excellence
in Education, 1983, p. 12)

This report listed many indicators of the "risk," including (a) lower

average achievement scores than during the "Sputnik" era 26 years ago,

(b) falling college board scores, and (c) fewer students having superior

scores on the Scholastic Aptitude Tests.

McDaniel (1985) noted that while the initial focus of the several

"reform" or "excellence" carnissions was at the secondary level, many

recent reports dealing with excellence have been directed to elementary

and middle schools as well. He estimated that about 30 national reports

have been issued and that the 50 states have appointed about 300 task

forces in the name of excellence.

Cassil (1986) attempted to identify the factors emerging from the

reports on educational excellence. The basic premise is that if

education is doing so poorly, "it must be broken and therefore should be

fixed." Subsequently, a variety of remedies have been proposed to

improve the educational system. However, these efforts, for the most

12










part, appear to have treated only the symptoms of the problem (Cassil,

1986). For example, merit pay has been proposed as a way to recognize

and motivate teachers. However, Watts (1985) argued that merit pay was

not the answer to improved instructional programs. lie proposed that

teachers should be paid for what they accomplish with students, not

rewarded for certificates, degrees, and years of experience.

Short (1985) also addressed the need for excellence and called for

strengthening curriculum through content restructuring or improvement as

the answer to bettering America's schools. He stated that reform is

unlikely to came unless attention is given to daily content and to the

purposes upon which student learning is centered.

Several of the major reform reports (e.g., "Educating Americans for

the 21st Century," by the National Science Board Ccmnission on

Precollege Education, 1983; "Secondary Education in America," by the

Carnegie Foundation, 1983; "A Nation at Risk: The Imperative for

Educational Reform," by the National Cranission on Excellence in

Education, 1983) called for increased high school curriculum for

students. The implementation of this demand may benefit those who stay

within the system or go onto college, but also may have a negative

effect on those who struggled to meet the original requirements

successfully. Thus, while for sane students excellence means further

enhancement of academic prowess, for many marginal or disadvantaged

students success involves maintaining sufficient skills and motivation

to stay in school.

The Business Advisory Cannission of the Education Caomission of the

United States stated that many students are still "disconnected" fron

society and that (a) 700,000 students drop out of school each year,










(b) three million youths ages 16 to 19 are looking for work, (c) one

million teenagers become pregnant, (d) juvenile arrests jumped 60-fold

between 1960 and 1980, and (e) suicides among teenagers grew by 177%

among white teens and 162% for nonwhites between 1950 and 1978 (Business

Advisory Ccmnission, 1985). It has been noted that school counselors

can play an important part in helping these students and called for drop

out prevention programs. An added emphasis on prevention at early ages

in school also could help prevent the "disconnected" youth of later

years (Schappi, 1985).

Because academic achievement is critical in helping students at all

levels reach their maximum potential, educators at all levels have been

charged with striving toward excellence. However, educators have not

yet addressed the fact that students may not have the basic skills with

which to meet the new standards of excellence. Students may not know

the skills that help them learn optimally and thus may not be achieving

to their fullest capacities. Perhaps if students can learn school

success skills at early ages, the standards of excellence can be nore

readily met.


School Success Skills


Cartledge and Milburn (1978), after reviewing the literature on

behavioral skills in the classroom setting, found certain skills linked

to academic success and to school success. These skills can be

expressed behaviorally within the general categories of personal

interaction skills and task-related skills. Since these have been

linked to school achievement and success, it seems logical to expose

children to them. If these findings are not put to use, children may be










cheated and not experience their maximum academic and social potential

in school. The specific school success skills selected for this study

are a subset of interpersonal and task skills. They were chosen because

they have been shown to help students become successful at school and

also have been shown to have positive outcomes for skills training. The

school success skills for this study were paying attention in school,

listening to teachers and peers, volunteering in school, using

self-control, interacting with teachers and peers, and utilizing

self-assessment. Each is discussed below.


Paying Attention (Attending)

One important school success skill is that of attending. The

discussion of attending as a complex neurological and physical process

is beyond the scope of this review. Rather, it is defined in terms of

classroom attending which "involves the social element of orienting

toward the teacher or to stimuli defined by the teacher at times and

under conditions specified by the teacher and as such, may be regarded

as an academically-relevant social skill" (Cartledge & Milburn, 1978, p.

138).

Cobb (1972), in a multiple regression analysis study of the

relationship of discrete classroom behaviors to fourth-grade

achievement, defined attending as doing what was appropriate in an

academic situation, e.g., looking at teachers when they presented

material, writing answers to arithmetic problems, or looking at other

students who are reciting. Results showed attending to be the most

powerful of eight behavioral categories in each of two schools,









contributing multiple Rs of .40 and .47, respectively, to arithmetic

achievement.

Cobb and Hops (1973) studied the effects of academic survival skill

training on low achieving first-grade students. In earlier research,

Cobb (1972) had found a relationship between achievement test

performance and the behavioral classes of attending to teacher,

following teacher instructions, and volunteering to answer academic

questions. These behaviors, therefore, were included as the "survival"

skills to be studied. Attending was redefined from earlier studies to

include two categories: attending and work. Eighteen first-grade

children chosen for low rates of academic survival skills combined with

low scores on standardized reading tests comprised the control group

(n=6) and the experimental group (n=12).

Results showed that the experimental groups increased their level

of survival skills by 24% and the control group only 3%, and that such

gains were related to reading achievement gains. Because the results

for students who had been identified as having low reading scores and

low academic skills showed positive gains (Cobb & Hops, 1973), in the

present study those students low in school success skills (by teachers'

perceptions) were identified in order to assess treatment effectiveness.

Hops and Cobb (1974) investigated the survival skills of attending,

following teacher instructions, and volunteering by training two

first-grade teachers in contingent reinforcement of the survival skills

and a third teacher in a new method of programmed, individualized

reading instruction; a fourth teacher whose class served as a control

group received no new instructions. Both experimental approaches

resulted in improved reading achievement for first-grade students in the









regular classroom when compared to the control group. The proportion of

children's survival-skill behaviors increased only for the groups who

received reinforcement of survival skills.


Listening

The skill of listening also contributed to the positive results of

the studies described previously (Cobb & Hops, 1973; Hops & Cobb, 1974).

It also has been identified by several other authors as an important

skill.

McGinnis and Goldstein (1984) identified listening as an important

skill in their curriculum for teaching prosocial skills as part of

mainstream and special education programs. They pointed out the

importance of teaching students how to show that they are listening to

others. They listed steps indicative of listening which included (a)

looking at the person who is talking, (b) remembering to sit quietly,

(c) thinking about what is being said, (d) acknowledging ideas by saying

yes or nodding your head, and (e) asking a question to seek more

information on the subject.

Myrick and Bowman (1981) stressed that listening is an important

part of learning and that students who listen closely do well in school,

follow directions better, and make fewer mistakes. They identified four

behaviors of careful listeners, including (a) looking at persons when

they are talking, (b) paying attention to persons' words, (c) being

aware of persons' feelings, and (d) verbally responding to the speaker.

Foster (1983) pointed out that being a good listener necessitated

acting like a good listener. Acting like a good listener included (a)

maintaining eye contact; (b) sitting attentively and leaning forward at









times; (c) looking as if listening is enjoyable by raising eyebrows,

nodding, smiling, and laughing appropriately; (d) asking questions; and

(e) appearing alert but not tense.

In the current study, listening skills were built upon the previous

attending skill and included actively listening for content and feelings

of teachers' and peers' cotnunications. Students were instructed in how

to show that they were actively listening and in how to respond to show

that they were listening.


Volunteering

Volunteering, in the sense of participating in classroom

discussions, has been positively correlated with academic achievement

(Hops & Cobb, 1973). Teachers were trained in reinforcement techniques

designed to increase "survival skills" (volunteering, looking around,

attending, and working) in first-grade students. After 20 days of

intervention the students in the experimental classrooms increased in

frequency of these "survival skills" and were significantly different

from the control students. Students in the experimental groups also

showed gains in reading achievement which allowed the researchers to

conclude that increases in skills would lead to increases in

achievement.

Zimmerman and Pike (1972) were able to increase question-asking

behavior in disadvantaged Mexican-American second-grade boys by adult

modeling of skills and contingent praise. The authors noted that the

modeling plus praise group was significantly different from the other

groups, with more response transfer to natural settings. This same










experimental group also produced more questions than the control group

and was significantly different from it.

Wittmer and Myrick (1980) also spoke to the importance of teachers

using open-ended questions to promote a facilitative atmosphere in

classrooms. They noted that a question is a facilitative response when

it seeks more information, provokes discussion, or is used to query

individuals on particular matters. The present study included the

presentation of open-ended questions as part of volunteering skills to

help students learn the advantages of this type of questioning to

enhance learning. Asking questions or contributing answers was believed

to reflect careful listening and active processing of material just

presented.

Noble and Nolan (1976), in a study of high school classes, found

that students could shape the frequency of the teacher calling on them

by the amount of volunteering they did in class. After three months of

school, observers found that teachers directed more questions to those

who volunteered more, where initially no relationship had existed.


Self-Control

The fourth school success skill used in this study was

self-control. The use of self-control has been shown to be an important

part of school behavior. Dunn and Kowitz (1970) submitted a list of 40

adjectives, believed to relate to academic success, to 100 secondary

school teachers who were then asked to rate the characteristics on a

5-point scale in terms of significance to school achievement. The

results showed the composite of a hard-working, mature student.

Included in this composite was self-control.









Principals, teachers, school counselors, and researchers have long

been concerned with finding ways to help reduce aggressive behaviors by

teaching self-control mechanisms. Drabman, Spitalnik, and O'Leary

(1973), as one of their experimental conditions, taught eight 9- and

10-year-old boys who were in classes for students with academic and

emotional problems to control themselves by earning bonus points for

matching a teacher's evaluation of their academic efforts and social

behaviors. The authors reported that the students maintained very low

levels of disruptive behavior and high rates of academic output. They

also noted generalization of appropriate behavior at times in the day

when the token program was not in effect.

The skill of using self-control is enhanced by the teaching of

assertive behaviors where students are taught the differences between

passive, aggressive, and assertive behaviors (Bower, Amatea, & Anderson,

1976). Students can be more in control if they can recognize situations

in which they can be assertive. They can also utilize the skills

associated with being assertive to act constructively, rather than

turning to aggressive behaviors where self-control is lost.


Interacting

The fifth school success skill, interacting with teachers and

peers, also has been reported to have a positive relationship to success

in learning. Cartledge & Milburn (1978), in their review of selected

studies, found a growing recognition of the reciprocal nature of

interactions between teachers and students. Experimental studies showed

that varying students' behaviors could bring changes in teachers'

behaviors.









Klein (1971), in a study of college students, found that positive

student behavior influenced teachers to use positive behaviors. She

pointed out that other researchers have studied interactions between

students and teachers through various types of student feedback and

found that written student feedback influenced teachers to change their

behaviors and become more like the model the students suggested.


Self-Assessment

The sixth school skill included was self-assessment. Students who

monitor themselves can increase appropriate behaviors. Bandura and

Perloff (1967) found that 7- to 10-year-old children who reinforced

their own behaviors generated more responses and were significantly

different fran control groups.

Broden, Hall, and Mitts (1971) demonstrated that self-recording of

classroom behavior can be used to increase appropriate behavior. An

eighth-grade female, using a plus to indicate that she was studying and

a minus for not studying to self-monitor study time, increased study

time to 78%. When the recording slips were taken away, the subject's

study time decreased from 70% the first day to an average of 27% for the

next four days. When the self-recording slips were reinstated,

performance increased to 80%, which was at a level where the classroom

teacher was able to praise her for a high rate of performance. It was

the school counselor who was able to institute the procedures that led

to increased appropriate behavior, which was maintained by the classroom

teacher and "natural" classroom reinforcers.

Kaufman and O'Leary (1972) also found that self-evaluation was

effective in maintaining a low level of disruptive behaviors that had










been established by a token program. Sixteen students from a

psychiatric hospital, who were deficient in reading and showed high

rates of disruptive classroom behavior, were instructed to make their

own judgments about how they followed the rules and to tell the class

and the teacher how many tokens they deserved. For the six class days

that the procedure was in effect, disruptive behavior remained at the

previous low level.

Johnson and Martin (1973) showed that second-grade children in a

self-reinforcement group yielded significantly different response rates

on visual discrimination tasks than the control group. Drahnan et al.

(1973) used self-assessment procedures to reduce disruptive behavior and

to effect transfer of behavior to times when the system was not in

effect. The present study had school counselors teach the students ways

to monitor their assignments and to extend these ideas to any area that

may have been of help to the students.

Cox and Gunn (1980) noted that instruction in acceptable

interaction skills in education has been limited and quite informal,

with the primary function of education being that of providing students

with academic skills. However, the social functioning of students

within the school has direct impact on the acquisition of these academic

skills. They further pointed out that the lack of appropriate social

skills may be the result of the students not (a) knowing what behavior

is appropriate, (b) knowing the correct motor skills to execute the

known behavior, and (c) being able to deal with the emotional response

which inhibits the skill performance, even when the correct skill is

known and has been practiced. Thus, it would seem logical to teach

children the behaviors they need to perform well in school, provide them










with a safe environment in which to practice the skills, and teach them

coping mechanisms to overcome emotional responses that inhibit maximum

performance.

Studies have documented many positive outcomes from exposure to

school success skills. In view of these findings, the issue of teaching

school success skills in the classroom should not be left unattended.

The problem then becomes one of finding the most effective and efficient

way to reach a large number of children with school success skill

instruction. Large numbers of children can be reached through

developmental guidance units where school counselors present

age-appropriate information to entire classrooms of children. However,

it had not been determined if a specific set of skills related to school

success can be taught to third-grade students in regular classrooms

through developmental guidance units and particularly to target

students. Therefore, the problem investigated in this study was whether

the presentation of school success skills, through a developmental

guidance unit using peer modeling or through a developmental guidance

unit utilizing a coaching technique, would distinguish from each other

or a control group on selected outcome measures.


Developmental Classroom Guidance


Developmental classroom guidance is a form of group guidance.

Group guidance is distinct from group counseling, which places emphasis

on remediation and is problem oriented. It is also distinct from group

psychotherapy wherein the main thrust is reconstructive or reeducative

and aims toward personality change (Brammer & Shostrom, 1976).









Branrmer and Shostrom (1976) characterized group guidance as (a)

preventative, (b) developmental, (c) emphasizing personally relevant

information, (d) being cognitive, (e) having an environmental emphasis,

(f) being delivered through presentation and discussion, (g) leader

directed, (h) topic oriented, (i) using planned activities, (j)

emphasizing skills, and (k) using camnon goals.

Duncan and Gumaer (1980) stated succinctly that, "Developmental

classroom group guidance is a systematic, sequentially planned

humanistic education program that provides an environment for

integrating affective and cognitive learning experiences" (p. 91). This

emphasis in guidance evolved slowly, but is now being implemented in

many school systems across the nation.

Growth of Developmental Guidance

Developmental guidance as it is known today emerged after much

evolution had taken place in the field of school guidance. Aubrey

(1982), in his historical review of guidance, stated that in general,

school guidance began with Jesse B. Davis, who is credited with

integrating guidance into the school curriculum. Frank Parsons also is

noted as being very influential in guidance, for in 1908 he established

the Vocation Bureau in Boston designed to help children who left school

to work.

Aubrey (1982) further noted that two major trends evolved in

guidance-vocational guidance and educational guidance. He stated "the

emergence of vocational guidance in public schools was consequently a

direct result of rapidly changing conditions in American industry and

quite unrelated to the accepted process of schooling known at that time"

(p. 199). In fact, the movement had no strong philosophical or










psychological support. The incorporation of psychometric approaches led

to a new interest with the development of predictive aptitude and

selection tests. This movement continued to influence guidance into the

1940s.

The other movement, educational guidance, developed in the 1920s

and encompassed two distinct views. One view saw educational guidance

concerned with distribution (i.e., pupil assistance in school, home,

recreation, vocational and social pursuits) and adjustment (i.e., help

for students when they are unable to learn at school). The second

emphasis was a widening of earlier vocational and moral guidance issues.

This view was a forerunner of developmental guidance. However,

developmental guidance did not evolve as a smooth transition from the

educational guidance movement. It emerged on the strengths of its

advocates, not in response to economic or societal pressures. Aubrey

(1982) also noted that the real roots of developmental guidance can be

traced to Robert Mathewson, who as early as 1949 advocated that the

guidance process moves with the student in a developmental sequence up

to the age of maturity. Aubrey (1982) stated that Mathewson also felt

the guidance process helped individuals gain in self-understanding, gain

in perspective on surroundings, and questioned whether teachers alone

could monitor this process.

Other writers also made cases for guidance in the schools. Wilson

(1950) stated that guidance was not a privilege, but rather "a necessity

for every normal child" (p.47). Kowitz and Kowitz (1959) wrote a book

for teachers that emphasized the need for guidance at the elementary

level. It also was intended to help the teacher increase sensitivity to

the child and provided specific techniques for problem solving. Some










basic ideas of modern developmental guidance were emerging in the 1950s,

but the literature was still filled with justifications for elementary

guidance and arguments about who should deliver it. Arbuckle (1950)

purported that counseling should be performed by all teachers, who would

have training in the field of human development and adjustment.

Controversy among writers and educators existed over who would deliver

the humanistic-type education.

Major events influenced guidance in the late 1950s and early 1960s.

America's reaction to the launch of Sputnik influenced guidance in that

much attention and money were directed toward education. Munson (1970)

pointed out that the National Defense Education Act of 1958 established

institutes to train school guidance personnel, the Higher Education Act

of 1965 provided fellowships for teachers to train in guidance, and the

Elementary and Secondary Education Act (1965) allowed a greater number

of schools to develop programs, all of which had great impact on the

guidance profession. As a result of these new provisions, school

counselor-student ratios lowered and school counselors had greater

access to more students and could begin to devote time and attention to

developmental needs.

The writings of the 1960s time period were influential in school

guidance and counseling. C. Gilbert Wrenn (1962) wrote the Counselor in

a Changing World wherein he expressed strongly that guidance should be

for the full range of students and be designed to meet their

developmental needs. Dinkmeyer (1968) also wrote that guidance must be

developmental in addition to remedial and preventative.

Faust (1968) stressed that the approach to children in the school

setting should be of a developmental focus for all children rather than










just for children in "crisis." Faust further noted that the school

counselor works with students in a developmental stage when students'

lives are centered in the classroom and in academic learning activities.

This perception lends strength to this study because the counselor

worked with students in a developmental capacity in the classroom.

Blocher (1966) and Munson (1970) also wrote in support of elementary

school counselors as established positions since they act as agents of

change and help education keep pace with change itself.

Developmental guidance had at its roots, and incorporates into its

basic principles today, the idea that appropriate information should be

taught to all children within appropriate developmental stages so that

they can realize their fullest potentials academically and socially.

This belief supports a basic theoretical position of this dissertation,

that appropriate school success skills or behaviors can be learned when

presented to children in age-appropriate contexts. Developmental

guidance is a valid media through which specific skills can be delivered

to children.


Classroom Guidance Units

Classroom guidance units have been shown to be effective and

efficient vehicles to use with large groups of students. Duncan and

Gumaer (1980) addressed the strengths of classroom units by stating that

"they use human developmental stages and growth patterns of children to

enhance the entire learning process" (p. 95). The success of this

process has been demonstrated in the following studies.

Myrick et al. (1986) demonstrated that school counselor-led

classroom guidance units resulted in improvements in school attitudes in










fourth-grade students ranked both high and low initially. Students from

67 schools were randomly assigned to a treatment group which received

the guidance unit. The unit consisted of six, 30- to 45-minute

sessions. Each session used the format of an introduction, an activity

for large groups, an activity for peer-led small group interactions, and

a closure where the whole group processed discussion together. The

others were assigned to a control group. The lowest and highest

students were determined by teachers who rated each students' attitude

on a scale of 1 through 7, with 7 being the most positive. Students

responded to a 20-item inventory that included classroom behaviors

related to achievement in school and general attitudes. Teachers also

responded to these items in terms of their perceptions of the students'

behaviors and attitudes. Results indicated that there were positive

changes and significant differences in school attitude for both low and

high students as a result of intervention.

The present study utilized the basic format of an introduction, an

activity for large groups, an activity for peer-led small groups, and a

whole group discussion for closure for one of the two treatment

conditions called coaching. It was not known if this vehicle would be

effective in teaching specific school success skills to third-grade

students. The peer-led small group component of the Myrick model was

chosen for third-grade students in the present study because it was

effective with fourth-grade students.

Dinkmeyer and Caldwell (1970) attested to the importance of small

group work, noting that it provides the safety and support which are

necessary for self reinforcement. Muro and Dinkmeyer (1977) also noted









encouragement is powerful when given by group members rather than from

the leader.

Cobb and Richards (1983) found the combination of group guidance,

small group counseling for target students, and teacher consultation to

be an effective intervention for a sample of fourth- and fifth-grade

students to increase student self-awareness and understanding of the

unique characteristics of others. Even though the results were in a

positive direction, it was unclear exactly how each component

contributed to the end result. In the present study the two components

were separated in an attempt to assess each of their separate merits.

Sorsdahl and Sanche (1985) found that classroom meetings, where the

entire class meets with the teacher for the purpose of problem solving

or discussion, were effective in changing classroom behavior in the

positive direction for fourth-grade students. The authors concluded

that, "Classroom meetings also seem to be effective means of providing

preventive counseling to entire classes of children by enhancing their

problem-solving skills, their decision-making skills, their acceptance

of responsibility, and their interpersonal skills" (p. 55). The current

study utilized this basic concept in the modeling condition, where the

model demonstrated a skill and the class watched and discussed it as a

whole unit. However, in this study, the discussion was led by the

school counselor instead of the teacher, as was the case above.

Classroom meetings have been found to be effective for some purposes;

however, it remained to be shown if they will work to teach school

success skills to third-grade students.

Cuthbert (1984) also found developmental guidance units to be

appropriate in teaching communication skills to fourth-grade students










(N=108). Two fourth-grade classes were assigned to the experimental

group and two to the control group (which received the treatment later).

The treatment group received training in six specific communication

skills delivered through a counselor-directed unit. The outcome measure

was a student/teacher, Likert-type rating scale designed to assess

passive and aggressive behaviors and school behaviors related to working

carefully on school assignments and handing them in on time. Results

showed that students in the experimental groups were significantly

different from the control group and showed improvement on school

behaviors on the teacher scale, but pretest differences on passive and

aggressive subscales confounded the findings. The present investigation

used the basic format of this study, but used different outcome measures

to assess treatment. The ccrmunication skills study did demonstrate

that teacher's ratings of school behaviors (e.g., working hard on school

assignments, handing in assignments, having materials ready to work, and

doing work carefully) were improved after the students had participated

in developmental guidance units.

Cuthbert (1985) utilized a developmental guidance classroom unit to

enhance students' feelings about self. It consisted of six, 30-minute

sessions and was delivered by the counselors to second-grade classroans.

One classroom received delayed treatment so that it could serve as a

control group. The format of the unit did not incorporate the small

group practice segment. Each session consisted of the school counselor

leading the students through a classroom activity, through group

processing of related questions, and through large-group closure or

sunnary. The outcome measure was a pictorial Self-Assessment Manikin

(Lang, 1980) where children marked, for each question, how they felt









about themselves in terms of pleasure, excitement, and control on a

5-point pictorial scale. Even though results did not strongly support

inferences regarding treatment effects, several key items (e.g., How I

feel about myself and How I feel about being different from others)

indicated that the unit had had some positive impact on self ratings

with greater pleasure, less excitement and more control being indicated

by those who had not used a maximum rating on the pre assessment. This

study used the SAM pictorial scales as an outcome measure for school

success skills.

In a large study (N=896), Gerler and Anderson (1986) investigated

the effects of classroom guidance on children's classroom behavior,

attitudes toward school, and achievement in language arts and

mathematics. After school counselors presented classroom guidance units

to fourth- and fifth-grade students, results showed that the treatment

group improved on two measures of classroom behavior and on a measure of

school attitude, whereas the control group declined. Language grades

improved for the treatment group, but there were no significant

differences in language and math grades between the groups.

There is support in the literature to show that developmental

guidance units are effective ways to teach children new ideas. It was

believed that the use of such units for the teaching of academically

relevant school success skills would be further supported if the

developmental guidance unit using modeling and/or the the unit using

coaching had an effect on outcome measures for children in third-grade

classrooms. More particularly, the use of the guidance units would be

further supported if target students (i.e., those ranked by teachers as

having lower skills) showed improvement on dependent measures.











Target Students


As covered in the review of literature relating to specific school

success skills, target students, or those selected from a larger

population for particular intervention or study, increased reading

achievement (Cobb & Hops, 1973), increased question-asking behavior

(Zimmerman & Pike, 1972), decreased disruptive behavior and increased

academic output (Drabman et al., 1973; Kaufman & O'Leary, 1972),

increased study time (Broden et al., 1971), and increased response rates

on visual discrimination tasks (Johnson & Martin, 1973). Other studies

also have shown that target students are of particular interest in terms

of finding effective methods to help students better themselves both in

their behaviors and their attitudes.

Lewin, Nelson, and Tollefson (1983) noted that the behaviors

performed by students in the classroom affect the way their teachers

respond to them. Students who are too aggressive and disruptive are

often rejected by their teachers, and those who are high achievers,

motivated, and skilled socially are often favored. To determine if

teachers' attitudes toward students could be altered as a result of the

students changing their disruptive behaviors, they had each student

teacher (N=35) identify a target student whose behavior was disruptive.

The experimental group (n=19) and control group (n=16) teachers observed

their target students for seven school days. Student teachers in the

experimental group also received training in behavioral techniques of

alternative reinforcement of other behavior in combination with

extinction. Results showed that teachers' attitudes toward disruptive









target students did not change even though they were successful in

decreasing disruptive behavior.

Myrick and Dixon (1985) found that teachers rated their students

more positively after intervention. They noted that students who do not

feel good about themselves at school often have learning problems,

perform poorly in class, and disrupt others. In an attempt to help

students develop positive attitudes, Myrick and Dixon had target

students participate in six, small-group counseling sessions. They

found that small-group counseling with fifth- and sixth-grade target

students (N=59), who were identified by their teachers as needing

improved attitudes, helped them became aware of changes in themselves.

Teachers reported differences in classroom behaviors. In the present

study an attempt was made to increase positive school behaviors and

attitudes by working with the target and other students in their

classroom settings through developmental guidance units.

Myrick et al. (1986), in the Florida Classroom Guidance Project,

also found that target students and top students benefited from

participation in developmental guidance units delivered by school

counselors. The target students (n=623) were ranked by their teachers

as the six lowest in each class regarding attitudes toward school.

After participating in six, 30- to 45-minute sessions dealing with

understanding feelings and behaviors, experimental condition target

students improved over the control students on their ratings of (a)

finishing assignments on time, (b) saying kind things to others, (c)

believing that others were interested in what they had to say, (d) being

good workers at school, and (e) reporting that school made them feel

happy. Teachers rated the target students improved on three of eight










classroom behaviors and on nine attitude statements. Top students also

benefited.

Similar results were also seen in the Indiana Project, a

replication of the Florida study (Myrick et al., 1986). Again, target

students who received the classroom guidance unit rated themselves

significantly different from the control students on 4 items and

teachers rated the students significantly different on 13 of 20 items.

Top students also improved on 4 items.

Because it has been shown that target students benefited from

developmental guidance units and from the presentation of specific

school skills in other studies, in the present study outcome measures

were also examined for target students as well as for total third-grade

classes to see if either or both would rate themselves as improved on

school success behaviors and on attitudes about school situations.

Specific school success skills were presented to all students in

developmental guidance units, one incorporating modeling and the other

coaching, as techniques for counselors to deliver the content of the

units.


Modeling


Modeling, one of the fundamental means by which an observer can

acquire new behavior vicariously from the demonstrating person's display

of that behavior, has been shown to be an effective teaching method for

adults and children. Bandura (1969) pointed out that a wide variety of

response patterns have been transmitted through modeling, including

diverse behaviors such as stylistic response patterns, modes of









aggressive behavior, dramatic play patterns, prosocial frustration

reactions, and teaching styles.

Toner, Moore, and Ashley (1978) used models to teach the prosocial

behavior of self-control. First- and second-grade boys who were told

that they would model self-control for other children were themselves

more self-controlled and touched prohibited toys less than those not

told they would be models. In addition, Sarason (1968), in a series of

studies using models (primarily psychologists) to present skills (e.g.,

applying for a job, resisting temptation to engage in antisocial acts)

to juvenile delinquent boys, found the results to be the strongest and

most positive for the modeling groups, followed by role-playing and

control groups. On rating scales, they were less defensive and showed a

greater willingness to admit to having problems.

Bandura, Ross, and Ross (1961) demonstrated that using live models,

or social imitation, rather than successive approximations, may

short-cut the acquisition of new behaviors, since subjects (36 boys, 36

girls; 37-69 months of age) watching adult male and female models

displayed more physical and verbal aggression than those who were

exposed to nonaggressive models. Bandura and Walters (1963) also noted

the efficiency of the social imitation modeling condition. They

criticized traditional approaches to treating unassertive children,

where the desired social behavior is waited for and then reinforced, and

suggested that modeling can be more effective in treating inhibited

behavior.

The use of modeling has been effective in many areas, including

reduction of aggressiveness and the development of prosocial behaviors.

Thelen, Fry, Fehrenbach, and Frautschi (1979) pointed out that the









cognitive processes that are important components of modeling also can

occur from watching filmed models. For example, Bandura and Barab

(1973) noted that snake phobic adults (16 to 54 years of age) made

comparable reductions in snake fear when they saw adult or child models

demonstrate approach responses toward snakes.

Filmed, or symbolic modeling, has also been used extensively to

teach social skills to isolated or withdrawn children in the preschool

age range. O'Connor (1969) designed a 23-minute modeling film, in which

six nursery school actors served as models, to demonstrate social

competencies, and, at the same time, to reduce fear of peer interaction.

The film was shown to an experimental group of six preschool isolates (4

females and 2 males) while a film about dolphins was shown to seven

preschool isolates (4 females and 3 males) for the control group. These

children were considered to be isolated and so were targeted by first

being nominated as such by their teacher, and then observed on total

number of social interactions. Nonisolates were also observed to

provide baseline data for frequency of interaction. The experimental

group increased in social interaction rate after the modeling

intervention, but no follow-up assessment was done.

O'Connor (1972) included follow-up data in a subsequent study where

the same modeling film and the same control film were shown to social

isolates who were also divided in terms of external reinforcement or no

external reinforcement conditions. Graduate students supplied the

external reinforcement, which consisted of giving each child a total of

five hours of praise and attention upon successive approximations of the

behaviors during a 2-week period. All of the conditions--modeling and

shaping (n=7), modeling alone (n=9), and shaping alone (n=8)-produced









an increase in social interactions compared to the control group. At

the third posttreatment assessment period, follow-up data showed the

conditions for modeling and modeling plus shaping to have maintained

their effect with no significant differences between them, thus

indicating no incremental effect for shaping.

Evers and Schwarz (1973), in studying isolated nursery school

children on the bases of teacher nomination and low interaction rate,

utilized the same O'Connor (1969) social skills film in their modeling

group (N=8, 4 females and 4 males). No control group was included.

They also had a modeling plus praise treatment group (N=7, 3 females and

4 males) where, for two days, the teacher provided praise verbally to

the targeted children as group members so as not to redirect their

attention from the peer group to the adult. The teacher also was

permitted one prompt per day to guide the isolated child into group

activity. The results indicated that the modeling plus praise condition

was not significantly different from the modeling only group, and at

follow-up all subjects maintained or improved their posttest scores.

Their rationale for using teacher reinforcement was that familiar adults

might be more effective reinforcement agents than graduate students. On

the other hand, O'Connor's (1972) rationale for using graduate students

was that teachers actually reinforce children for attention to adults,

thereby diminishing peer interaction; also, strangers have been more

effective reinforcers than parents, and presumably more so than

teachers.

Gottman (1977) also used the O'Connor film for his experimental

group (N=17, 8 females and 8 males) and an Alaskan travelogue film for

his control group (N=17, 8 females and 9 males). The latter was









different from O'Connor's Marineland control film, which was devoid of

people. Pretreatment to posttreatment changes were found for both

experimental and control groups, and thus failed to replicate O'Connor's

1972 findings. It was suggested that this might be due, in part, to the

use of the different film or to the use of different assessment times.

This study included and expanded its outcome assessment measures beyond

just the frequency of the social interaction measure and included

examining the proportion of positive peer and teacher interactions, time

alone, and a sociometric rating of acceptance and rejection. There were

no significant differences for main effects, sociametric main effect, or

treatment by sociometric interactions. The researcher suggested that

the failure to replicate O'Connor's results raised methodological

questions about observational techniques.

Even though the O'Connor studies (1969, 1972) and the Evers and

Schwarz (1973) study were conducted on small numbers of nursery-school

age isolated children, and with varied treatment conditions which

included modeling plus external reinforcement given by graduate students

and teachers, they all have relevance for the current study because they

demonstrated that modeling of social canpetencies by same-age peers had

an effect on outcome measures. Because of the above considerations, the

size of the sample in this study was increased, included all children in

regular third-grade classrooms (i.e., not just those identified as only

an extreme in natural populations), and had third-grade peer models

demonstrating the specific skills. Separate data were obtained on those

with the lowest school success skills ratings, but all children received

treatment.









In an attempt to more precisely identify specific social

interaction behaviors, and thereby further clarify the process by which

interaction increases, Keller and Carlson (1974) developed four 5-minute

videotapes depicting the specific social behaviors thought to be

components of social interaction (i.e., imitation, smiling, laughing,

token giving, and affection giving). These tapes were shown to isolated

preschoolers in the experimental group while the control group saw four

nature video tapes. An accompanying sound track described these

actions. All subjects were observed on the behaviors of interacting,

giving reinforcement, and receiving reinforcement, with the last two

being broken down more specifically to include imitation, smiling or

laughing, giving tokens, giving affection, and verbalization. Results

revealed increases in all three categories for the experimental group

(i.e., the control group remained the same) at posttreatment, but the

differences were not significant at follow-up. The results also showed

that treatment procedures raised the probability of occurrence of

responses that the subjects had already learned (i.e., verbalization,

imitation, and smiling) but did fail to implant new responses (affection

and token giving) into the subjects' repertoires. This study

highlighted the importance of defining skills specifically and of

assessing outcomes on quality of interaction to gain specific

information about the relationship of skills and behaviors. Student and

teacher perceptions were included in this current study, as well as an

attitudinal measure to determine how the students (the vicarious

learners) feel about themselves in relation to school.

As was shown from the studies reviewed above, modeling has been

used extensively to teach skills to isolated or withdrawn preschool










children. However, it is not clear if these findings would be the same

for elementary school age children or if they would work for a large

number of subjects treated at one time. The present study included

specific school skills which are relevant to success in the classroom.

They were presented by same-age models to classrooms of students.

Student and teacher behavioral ratings were used along with a student

attitudinal instrument to obtain a comprehensive set of outcome

measures.


Coaching


Coaching, a technique where new skills are presented and practiced

or shaped while being coached, has been shown to be effective in

training or facilitating behaviors in children. Although the exact

procedures used vary across studies, the efficacy of coaching has been

established in several studies.

Gottman et al. (1976) used a form of coaching as part of a

treatment to teach skills to two third-grade females. The experimental

treatment condition consisted of (a) showing them a videotape on how to

initiate interaction, (b) having them role-play how to make friends with

a male college coach who took the part of a new child in the class to be

befriended by the subject, and (c) participating in a game designed to

have each take the "special perspective" of the listener. The control

group (n=2) played games with a female experimenter. Even though

results indicated that the girls in the treatment group were

significantly different in sociometric position at follow-up and

redistributed their interactions to peers, the study had limitations.

First, the number of subjects is small, and second, the treatment group
n










was coached by a different male for each subject and the control group

by a single female. The true effect of coaching cannot be seen here

because it was only one part of three in the treatment condition.

Bornstein et al. (1977) showed that skills training consisting of

instructions, feedback from a therapist, behavior rehearsal by the

subject, and modeling of the skill resulted in considerable improvement

in component behaviors and overall assertiveness for all of four

subjects (8 to 11 years old). A type of coaching occurred when the

therapist gave feedback to the subject on the chosen skill until the

therapist felt the criterion for that behavior had been reached. It is

hard to determine the true effect of the coaching component here since

modeling (i.e., a model presented the prompt while the therapist

presented a chosen scene) was a part of the treatment. Again, the

number of subjects was small, which makes generalizability of results

questionable.

Oden and Asher (1977) used a clearly defined coaching component as

one of three conditions in their study. The number per condition (n=ll)

was expanded in relation to the studies cited earlier, and included

third- and fourth-grade socially isolated children (low-acceptance

students, total N=33). The coaching condition consisted of three parts

where (a) the student was verbally instructed in social skills by the

experimenter, (b) the student was given an opportunity to practice the

skills with a peer, and (c) the student and peer had a postplay review

session with the coach. The skills coached had to do with making

game-playing fun and included (a) participating in a game, (b)

cooperating, (c) talking and listening, and (d) validating. The coach

proposed a concept (e.g., cooperation), probed understanding of it,










rephrased the child's examples, asked for the description of opposite

behavior, and asked the subjects to try out ideas in a game. In a

postplay review, the coach discussed the experience with the subject.

The other treatment conditions were peer-pairing, where no coaching of

skills was given, and a control group where the students played.

Results showed that the coaching group was significantly different from

the other two conditions and increased on a play sociametric rating.

This group also gained in friendship nominations.

The Oden and Asher study (1977) added strength to the skill

training literature by clearly defining the coaching condition and

specifying the exact skills being taught. However, it failed to show if

these techniques, particularly the coaching technique, are effective

with more than two children at a time since the targeted, socially

isolated children and a paired partner were pulled from their classrooms

and returned after intervention. It also is unknown if the skills

learned in a created environment transferred to other real life

situations.

Ladd (1981) also examined third-grade children with low sociometric

ratings or peer acceptance and low behavior ratings (N=36). Three

treatment conditions were utilized and included (a) skill training or

coaching, which used the same three components as Oden and Asher above;

(b) attention control; and (c) nontreatment control. All three

treatments were implemented across three elementary schools. Each

condition had 4 students (1 male and 1 female from each of two

classroans) within each school, for a total of 12 per condition. The

skills coached in condition one included asking questions, leading, and

offering support to peers. Children were trained in dyads in an










experimental room for a total of eight, 45- to 50-minute sessions on

alternating days. The attention control dyads were also separated from

the classroom, where they received only experimenter attention and peer

interaction. The control group dyad was not separated and did not

receive training. Results showed that children who had been coached in

the concept of the skill spent a greater amount of time engaged in

question asking and leading and were significantly different from the

other groups. They were also significantly different in classroom peer

acceptance. The design of this experiment, the analysis (where the

effect of class was examined statistically in the pretest measures), the

identification of specific skills, and the careful detailing of the

coaching condition brought clarity to the use of this procedure.

However, again, the students were taken in dyads from their natural

classroom environment to an experimental room for training. It was

still not known if these techniques wrk in larger groups or for

moderately socially deficient students since low social acceptance

students in dyads were studied.

Gresham and Nagle (1980) studied isolated third- and fourth-grade

students who had been so labeled by sociametric methods. The results

indicated that coaching, which consisted of presenting rules for

behavior, rehearsal with the coach, and peer partner and feedback;

modeling; and an abbreviated combination of the two, were functionally

equal in training isolated third- and fourth-grade students in social

skills (i.e., participation, cooperation, carnnunication, and

validation). Again, children were coached in dyads or triads apart from

the classroom.










La Greca and Santogrossi (1980) worked with an isolated sample of

elementary school age children (grades 3 to 5) selected for low peer

acceptance ratings, but presented treatment in a group setting held

after school as opposed to individual, dyadic, or triadic groupings as

in previous studies. Subjects were divided into three treatment

conditions. The first was a skills training group (SK) where eight

skills (i.e., smiling, greeting, joining, inviting, conversing, sharing

and cooperating, complimenting, and grooming) were presented by first

having the students view the skills presented by peer-models, with

discussion following. Next, students were coached by receiving

suggestions from the trainers in the use of the skills provided and

given a chance to rehearse them. Role-playing situations were based on

situations that were relevant to the students and practiced with each

other. Third, in the SK condition, students were given homework

assignments that encouraged more practice of the skills. The attention

placebo group received the same training except that they saw movie

excerpts, played pretend games, and had unrelated homework. The waiting

list control group participated only in pre- and postmeasures. Results

indicated that the SK group showed increased skill in role playing,

greater verbal knowledge of skills, and more initiation behavior with

peers at school. The attempt to teach specific skills to groups was a

unique feature of this study; however, the true effectiveness of

coaching was confounded since it was coupled with modeling. It still

was not knriwn if coaching of skills with large groups of children

divided into smaller groups of five or six students where they coach

each other on skills will be efficacious.









The literature showed coaching to be effective with samples of

children who showed a deficit in skills both at the nursery school age

and at elementary school age. In the current study coaching was

examined as a separate condition to determine if it were effective in

presenting school success social skills to children in third-grade

classrooms. Students rated the lowest in school skills were targeted

for separate analysis so that impact of this treatment could be more

fully understood.

A promising link can be made between the modeling and coaching

literature, that was originally developed on socially isolated children

in nursery school samples and expanded to elementary school age

children, and the educational literature which identified skills as

being necessary for academic achievement and school success. Both

fields have successfully demonstrated that modeling and coaching have

positive effects on outcome measures of school success skills. The

developmental guidance literature also attested to the importance of

enhancing children through the presentation of various school success

skills. The classroom may be the area where all types of children can

make gains in acquiring school success skills and in achievement. It

also may be the arena where the strengths of these separate disciplines

can culminate and have impact on the development of large numbers of

children. The combined literatures document that school success skills

can be defined as behaviors, modeling and coaching are effective

techniques for teaching new behaviors, classroom guidance units are

successful vehicles to deliver new information to be learned by

students, and school success is the result of the utilization of

specific skills. Therefore, in this study these diverse findings were










integrated to investigate whether one or the other treatment components

(or both) proved successful in enhancing students' school success

skills.


Outcome Measures


In order to assess the effects of the treatments, several dependent

variables were used. Thelen et al. (1979) have stressed the importance

of multiple outcome measures in skills training. In the past, the use

of outcome measures has varied from using only one to using several.

However, results have been confusing to interpret because of

inconsistencies in methods and definitions.

Wanlass and Prinz (1982), in an extensive review of literature

related to childhood social isolation, found that the majority of

modeling treatments used only frequency of interaction as their

behavioral indicator of effects. They pointed out that controversy

exists over whether interaction rate has been established as a predictor

of later maladjustment and stated that there is a trend away from

relying on overall interaction rate as the primary outcome variable,

toward evaluating the quality of interaction. Gresham and Nagle (1980),

in their overview and analysis of effective ways to present social

skills to isolated children, also found modeling being assessed by

social interaction rate, while coaching treatment studies used quality

of interaction as primary outcome measures. They indicated that both

modeling and coaching researchers championed their own methods as best.

Foster and Ritchey (1979) stated that direct observations have

positive features; i.e., generate objective data, are under experimenter

control, can be checked among raters, allow for specification of










treatment effects, and "are the behavioral 'bottom lines' for

establishing functional relationships among antecedent conditions,

social behavior, and interpersonal consequences" (p. 631). However,

even with all of these positive assets, data generated from behavioral

observations have been confusing and inconsistent, with emphases on rate

of interaction and/or quality of interaction generated by various coding

systems. Because of the design of this study, where five counselors

worked in three different classrooms each, behavioral measures for which

raters tabulate scores were not used. This technique could not be

practically implemented in this study.

Goldfried and Kent (1972) pointed out that while behavioral

scientists generally are suspicious of self-report measures, they may be

reliable and valid under specified conditions. The Behavior and

Attitude Inventory (Myrick et al., 1986), a self-report measure where

children rate classroom behaviors in terms of self-perceived frequency

and rate attitudes about school in degrees of agreement on a Likert-type

scale system, has been used successfully for evaluation of developmental

guidance units. This study used the School Success Inventory-Student

Form which is modeled after Myrick's instrument. Students rated

themselves in terms of frequency with regard to school success

behaviors. They also rated themselves along continue of attitudes

toward school on the School Attitude Inventory in terms of pleasure,

excitement, and control with regard to school situations.

Teachers recorded their perceptions of students' behaviors on the

School Success Inventory--Teacher Form which reflected the same school

success skills upon which the students rated themselves. Edelbrock and

Achenback (1984) stated that teachers' perceptions often play a key role










in determining what is done to help disturbed children. Pekarik, Prinz,

Liebert, Weintraub, and Neale (1976) noted that teacher ratings are

usually regarded as relatively reliable and valid measures of children's

adjustment, and the combination of self report and teacher report are

thought to be a more valid index of disturbance than the teacher ratings

alone. Khan and Hoge (1983) cautioned researchers to be aware that

teachers may be sensitive to different aspects of behavior in boys and

girls and may reflect this in their ratings.


Sunmary


A review of the related literature has (a) shown that developmental

guidance is a viable vehicle to present information to children, (b)

shown that modeling and coaching have been effective techniques for

presenting new skills, and (c) demonstrated that self-report and teacher

ratings of behaviors and attitudes can be used to assess treatment, thus

their inclusion as dependent measures.

Confusion has existed and still exists in the literature about how

best to teach skills to children. Educators and researchers have tried

various methods to help students gain skills to make them more

successful in the school environment and continue to direct energies

toward this goal. This awareness of the importance of teaching school

success skills must continue so that interventions can help children

achieve the excellence sought in contemporary education. The problem of

how to help children learn school success skills most effectively still

remains.

Researchers in education studying ways to enhance academic success

have begun to identify skills for success at school, researchers in the










counseling professions have demonstrated that developmental guidance

units are effective methods for helping students learn and change their

attitudes, and researchers in the social sciences studying ways to

improve social isolation have shown that modeling and coaching are

effective ways to present skills. However, a specific set of school

success skills, presented through developmental guidance units

incorporating modeling and coaching, had never been delivered to

classrooms of children and evaluated for effectiveness. Such research

became the focus of this study.















CHAPTER III
METHODOLOGY


This study was designed to investigate the effectiveness of

developmental guidance units for improving school success behaviors and

attitudes toward school in third-grade children. A unit featuring

modeling and another involving coaching were compared with a control

group in terms of teachers' and students' perceptions of school success

behaviors and students' attitudinal ratings of school situations.


Population


The population for this study was composed of children in regular

third-grade classrooms in Alachua County, Florida, Public Schools. It

included special education students who were "mainstreamed" as well as

students who had been accelerated or retained.

Alachua County is located in north central Florida. The University

of Florida is located in the county. In the 1980 census, Alachua County

had a total population of 151,348 with about 71% designated urban and

29% designated rural. The 1984 median family income was $17,072

(Florida Department of Education, 1983-84).

In the academic year 1983-84, the Alachua County Public Schools had

approximately 22,345 students--64.29% white, 32.32% black, 1.45%

Hispanic, 1.28% Asian and .03% American Indian/Alaskan native. In the

population estimate of 1983, 51.2% of the total population for the age









range of 0-17 was male and 48.8% for this same age range was female

(Florida Department of Education, 1983-84).


Sampling Procedure


Permission to conduct the study in Alachua County Public Schools

was obtained. First, a detailed description of the project was

submitted to the University of Florida Institutional Review Board which

gave permission to conduct the study. The study was then presented to

the University of Florida Extended Services Office where approval was

obtained frcm the College of Education. Approval was obtained from the

Department of Planning and Evaluation, Alachua County Public Schools.

The principals of the 20 public elementary schools in Alachua

County, Florida, were informed of the nature of the study and invited to

volunteer their schools for participation. Each of the principals

received a copy of the Application for Research in Alachua County Public

Schools form which explained the purpose of the research, gave a brief

summary, stated the population needs, and delineated the amount of time

involved to complete the study. Twelve principals and school counselors

decided that their students would participate. Six of the schools met

the criteria of having three regular third grades. Principals and/or

school counselors from eight schools chose not to participate because of

other commitments and schedule problems. Five of the six eligible

school counselors completed the study and one failed to complete the

unit within the specified time frame.

Within each school, all third-grade classrooms in the participating

schools were randomly assigned to three experimental conditions by using

a table of random numbers. All students from whom data were obtained









and included were present for at least five of the six guidance

sessions.

Within each of the three experimental conditions, target students

were identified by their teachers. On a class roster teachers rated

their students on a 1-7 scale, with 7 indicating high levels of task and

interpersonal skills related to success in school. Those with the

lowest ratings (i.e., 4 and under) were considered the target students.



Resultant Sample


The sample was composed of 299 students from 15 classrooms within

five elementary schools in Alachua County, Florida, Public Schools (see

Table 3-1). Each classroom of students was randomly assigned to the

three experimental conditions of modeling (student n=102), coaching

(student n=101), and control (student n=96) within each school. The

target students (n=74) were distributed across all three conditions (see

Table 3-1). All schools who (a) gave permission; (b) had at least three

third-grade classrooms within the school; (c) had an experienced,

certified school counselor; (d) and administered all experimental

conditions in accord with the protocol were included. The final sample

included 58 black, 236 white, and 5 other ethnic students, and 147 boys

and 147 girls.










Table 3-1

Number of Students per Condition by School


4

Conditions

School Modeling Coaching Control

Total Target Total Target Total Target


School A 20 4 18 7 23 5

School B 20 6 21 2 20 6

School C 18 12 23 5 14 3

School D 20 5 17 3 18 1

School E 24 5 22 4 21 6



TOTAL 102 32 101 21 96 21











Research Design


The research design used in this study was a pre-post control group

design with three conditions (Issac & Michael, 1981) (see Table 3-2).

Classrooms of students in each elementary school were randomly assigned

to one of three conditions by using a table of randan numbers.


Treatment Description


Two experimental guidance units were used to present school success

skills to third-grade children. Students who participated in them were

compared to a control group. The set of school success skills presented

in the developmental guidance units included (a) paying attention in

school, (b) listening to teachers and peers, (c) volunteering in school,

(d) using self-control, (e) interacting with teachers and peers, and (f)

utilizing self-assessment. These skills were depicted by models,

practiced by students, and were evaluated by selected outcane measures.

School success skills were incorporated in the developmental guidance

units and tested for efficacy in the following conditions-(a) modeling,

(b) coaching, and (c) control.

Developmental Guidance Units

Both experimental treatments (i.e., modeling and coaching) were

incorporated into two developmental guidance units one entitled Skills

for School Success-Modeling (SSS-M) and the other entitled Skills for

School Success-Coaching (SSS-C). Each unit was delivered to

third-grade children by their respective elementary school counselors.

The units presented the same set of school success skills (i.e., paying

attention in school, listening to teachers and peers, volunteering at










Table 3-2

Pre-Post Control Group Design


Conditions


Outcome measurement times

Pre Post


T1 (Modeling) 01 02 03 X 1 02 03

T2 (Coaching) 1 02 03 X 01 02 03

C1 (Control) 01 02 03 01 02 03


School Success Inventory -

School Success Inventory -

School Attitude Inventory


Teacher Form

Student Form


O1=

02=

03










school, using self-control at school, interacting with teachers and

peers, and utilizing self-assessment techniques), but each involved a

separate teaching technique of either modeling (see Appendix A) or

coaching (see Appendix B).

Both guidance units consisted of six, 30-minute sessions. Each of

the six sessions had a stated goal and accompanying objectives

consistent with the overall purpose of teaching school success skills

and with the goals of the School Board of Alachua County's (1983)

Developmental Guidance and Counseling Plan. The goals, objectives, and

content of the units were the same, with the units differentiated only

by the experimental treatments of modeling and coaching.


Skills for School Success--Modeling

Children in the modeling condition participated in a developmental

guidance unit entitled Skills for School Success--Modeling (SSS-M) (see

Appendix A). For each session the school counselor choose students, one

male and one female, to model and demonstrate the skill being taught.

School counselors identified the modeling students from those who spoke

well, appeared to be leaders, and were respected by classmates. The

other observing students learned the skills vicariously as they watched

the students modeling the behaviors.

The school counselor facilitated the modeling condition by guiding

the models through the skill, as outlined in the guidance unit. A new

skill was modeled for each of the six sessions, using different models

each session to promote class interest and participation. Each session

concluded with a discussion and summary.









Skills for School Success-Coaching

Children in the coaching condition participated in a developmental

guidance unit, Skills for School Success--Coaching (SSS-C) (see Appendix

B). For each session the school counselor presented didactically the

new skill After the school counselor presented the school success

skill for each session, the students were divided into groups of five or

six students where they practiced the skill for about 15 minutes. They

coached each other on performance as they practiced. Each session

concluded with a discussion and summary.


Control Group

Students in the control group responded to the same measures of

school success at the same 8-week interval as the two modeling and

coaching groups. Their participation in one of the two conditions was

delayed for eight weeks until after the study was completed.


Assessment Techniques


In order to assess the effects of the treatments, the following

measures were used (a) School Success Inventory-Teacher Form (SSI-SF),

(b) School Success Inventory-Student Form (SSI-TF), and (c) School

Attitude Inventory (SAI).


School Success Inventory--Teacher Form

The School Success Inventory--Teacher Form (see Appendix C) was a

paper and pencil measure on which teachers responded to 12 items related

to school behavior for each student. These items reflected the same

behaviors listed on the student form (i.e., attending, listening,

volunteering, using self-control, interacting, and assessing self);









corresponding items were keyed to the behaviors on the students' forms

with verb tense and pronoun use made appropriate for the rater.

The teachers marked each item in terms of perceived frequency of

occurrence on a 5-point scale of "Very Often," "Often," "Sanetimes,"

"Seldom," and "Very Seldom." Respondents indicated their choice by

marking with an X or a check mark. Each of these categories was

assigned a numerical score, with one indicating the "desired" behavior

and corresponding to "Very Often." The scores ranged from 12 (very

often) to 60 (very seldmn).

Test-retest reliability for the SSI-TF was .94, as obtained from a

pilot study of two third-grade classes. The SSI-TF was administered to

teachers in two third-grade classes who first rated their students

(N=54) on the scale described above and then responded to the same

instrument two weeks later.


School Success Inventory--Student Form

The School Success Inventory--Student Form (see Appendix D) was a

paper and pencil measure. It contained 12 items, 2 items to represent

each of the six classroom behaviors of attending, listening,

volunteering, using self-control, interacting, and assessing self. The

items comprise behaviors related to school success.

The students marked each item in terms of perceived frequency of

occurrence on a 5-point scale of "Very Often," "Often," "Sanetimes,"

"Seldom," and '"Very Seldmn." They indicated their choices by marking

with an X or a check mark. Each of these categories was assigned a

numerical score, with one indicating the "desired" behavior and

corresponding to "Very Often." The total score over the 12 items ranged









from 12 (behaviors occurring very often) to 60 (very seldom). A total

frequency score was recorded for each student.

Test-retest reliability for the SSI-SF was .79 as obtained from a

pilot study of two third-grade classes. The SSI-SF was administered to

students in two third-grade classrooms (N=49) to gather preliminary data

on its psychometric characteristics. Students responded to the 12

behavioral statements on the scale described above and then rated

themselves on the same instrument two weeks later.


School Attitude Inventory

The School Attitude Inventory was designed by the researcher (see

Appendix E). It consisted of 10 items dealing with success at school.

Each item was rated by the individual student using a pictorial scale

called the Self-Assessment Manikin (SAM) (Lang, 1980), which was based

on factor analytic studies of affective ratings (Osgood, 1962).

The evaluation scale was presented visually in three cartoon panels

of five pictures each. The first panel showed the cartoon figure with

variations of smiles (happy) to frowns (unhappy), thus measuring the

dimension of pleasure. The second cartoon panel measured excitement and

ranged from excited to calm. The third dimension measured control,

which ranged from a small-size figure (being controlled) to a large-size

figure (controlling). Children indicated their choices in relation to

each item on all three dimensions separately by marking a large X over

the figure they choose to represent their feelings. Each scale thus has

a range of five points. For analysis, each response was converted to

numbers, with 5 representing the most pleasure, 1 the most calm, and 5

the most in control. Total scores ranged from 30 to 150. Test-retest










reliability was .76 as obtained from a pilot study of students (N=49) in

two third-grade classes who responded to the items initially and again

after two weeks.

A general questionnaire was given at the completion of treatment.

It asked for personal reactions to the units, suggestions for

improvement and general ccmnents. It was evaluated subjectively and not

entered into formal, statistical analyses.


Hypotheses


The following hypotheses were evaluated at the .05 level of significance

in this study:


1. There are no significant differences among the three groups on

teachers' perceptions of student school success behaviors, as measured

by pre-post change on the School Success Inventory--Teacher Form.


2. There are no significant differences among the three groups on

teachers' perceptions of target students' school success behaviors, as

measured by pre-post change on the School Success Inventory-Teacher

Form.


3. There are no significant differences among the three groups with

respect to race or gender on teachers' perceptions of student school

success behaviors, as measured by pre-post change on the School Success

Inventory-Teacher Form.


4. There are no significant differences among the three groups on

students' perceptions of school success behaviors, as measured by

pre-post change on the School Success Inventory--Student Form.











5. There are no significant differences among the three groups on

target students' perceptions of school success behaviors, as measured by

pre-post change on the School Success Inventory-Student Form.


6. There are no significant differences among the three groups with

respect to race or gender on students' perceptions of school success

behaviors, as measured by pre-post change on the School Success

Inventory--Student Form.


7. There are no significant differences among the three groups on

students' attitudinal ratings of school situations, as measured by

pre-post change on the School Attitude Inventory.


8. There are no significant differences among the three groups on

target students' attitudinal ratings of school situations, as measured

by pre-post change on the School Attitude Inventory.


9. There are no significant differences among the three groups with

respect to race or gender on students' attitudinal ratings of school

situations, as measured by pre-post change on the School Attitude

Inventory.



Research Procedures


One week before treatment began, all pretreatment measures were

given. The school counselors at each school involved in the study

administered assessments in their respective schools. School counselors

were trained in the procedures for administering the School Success










Inventories--Teacher and Student Forms and the School Attitude

Inventory.

To lessen the teachers' burden, names of the students were printed

on their forms by the school counselors. All forms were recovered by

the school counselors in their respective schools. Within this

pretreatment week, school counselors also planned the room configuration

for the coaching treatment group, which required students to be grouped

into circles containing five or six students.

The six treatment sessions took place once per week and started the

week after the premeasures were taken. After six weeks in which

students participated in development guidance units, the school

counselors administered the School Success Inventory-Student Form and

had the teachers respond to the School Success Inventory--Teacher Form

as postmeasures. They also had the students fill out the School

Attitude Inventory. A general questionnaire was completed by those who

received the guidance units. School counselors filled out information

sheets on each student to include gender and race so that this

information could be included for analyses. The data were placed in

clearly marked envelopes and sent by inter-school truck mail to the

investigator.

Subjects were eliminated frgm the sample if complete data were not

collected for them or if attendance had been less than five of six

sessions. All scores were entered directly into the computer for data

analyses.










Research Participant Training


The participating Alachua County Public School counselors were

trained by the researcher. Only counselors who had at least one year

experience in teaching developmental guidance units, were certified by

the State of Florida, and attended a training session given by the

researcher participated. The school counselors attended a training

session where the study was described in detail. The school counselors

were given copies of the dependent measures and instruction on how to

present them. Specific information regarding the experimental

conditions were given in addition to giving each counselor copies of the

two development guidance units. The researcher demonstrated the

modeling and coaching procedures and explained how they would be used in

classrooms. The researcher also showed how the small-group technique

would be used as part of the coaching condition.

School counselors administered pre- and postmeasures to the

third-grade students in the study and returned the measures to the

researcher. The pretreatment measures were administered within the week

prior to the first sessions of the guidance units. The posttreatment

measures were administered one week after the completion of session six

in each unit.

School counselors delivered the guidance units according to the

instructions provided by the researcher. After completion of the study,

the school counselors responded to a summary sheet to ensure that all

procedures were followed in accordance with the study.









Data Analyses


For all tests used in this study, the significance level was set at

.05. Analyses were performed on the pre-to-post change scores for all

measures. To account for the effect of preintervention levels, analyses

of covariance were used in examining the data, using the prescore as the

covariate. As indicated above, the design involved three levels of the

treatment factor (i.e., modeling, coaching, and control), and at least

five classes nested within each treatment.

Both pre- and posttreatment data were initially analyzed using a

hierarchical design analysis of variance model (Myers, 1979). This

analysis was appropriate since each subject belonged to an intact

classroom, which may have partly influence his/her score due to

interactions with this particular set of individuals or occurrences. In

this analysis the effects due to group membership (class) were separated

from the within-group residual error term. Due to the small number of

classes in each group, treatment effects were also assessed with a

standard analysis of covariance (ANCOVA); the larger number of degrees

of freedom available yielded more powerful tests of the hypotheses.

Pooling of the error terms in such analyses was considered appropriate

in this case, as there appeared to be no a priori grounds for regarding

the classes as differing systematically on any of the dependent

variables.

Tests were carried out on the pre-to-post change scores of the

various dependent variables following treatment. These included change

in the total scores (i.e., those obtained by sunming individual item

values) for the School Success Inventories-Teacher and Student Forms,









and the total and subtotal scores for the three dimensions of pleasure,

excitement, and control on the School Attitude Inventory. A limitation

of examining subtotal scores is that there are fewer items available for

each separate analysis, and thus lower reliability; however, these

analyses were performed in an attempt to determine which dimensions

might be most sensitive to students' attitude changes. ANOOVAs were

performed on the data of all students, and on the subsample of

low-skilled target students. Hypotheses regarding race or gender were

tested by analyses including both variables as factors, to allow

evaluation of potential interactions; these analyses were restricted to

black and white students only, due to the low numbers of other racial

groups in the sample. Significant effects were further evaluated with

Bonferroni t-tests, with alpha levels adjusted to control for the number

of contrasts performed (Myers, 1979).


Description of Methodological Limitations


Attempts were made to strengthen the internal validity of the study

in several ways. A control group was included to neutralize the effects

of history, equate for the effects of maturation, control for the effect

of regression, and equalize the effects of pretest measurement on

groups. However, in regard to the last point, no group received only

the posttest measures to control for test sensitization.















CHAPTER IV
RESULTS


To assess the effectiveness of developmental guidance units for

presenting school success skills to third-grade students in their

classrooms, analyses were performed on pre- and postdata obtained from

three dependent measures. These were (a) teachers' perceptions of

student school success behaviors as measured by the School Success

Inventory-Teacher Form, (b) students' perceptions of school success

behaviors as measured by the School Success Inventory--Student Form, and

(c) students' attitudinal ratings of school situations as measured by

the School Attitude Inventory. Two experimental conditions, modeling

(n=102) and coaching (n=101), were compared to a third control condition

(n=96). Data were collected from students in five public elementary

schools (total N=299).

In order to assess comprehensively all aspects of the three

experimental conditions, nine hypotheses were investigated. The

confidence level was set at .05 to determine statistical significance.

Each hypothesis is discussed in this chapter in relation to dependent

measures. To sharpen the assessment of experimental effects by removing

the influence of any preintervention variation, prescores were used as

covariates in computing analyses of covariance (ANCOVA) on the change

scores for each measure.

This study involved classrooms of students nested within the three

experimental conditions. Accordingly, a hierarchical ANCOVA was










initially employed for overall analyses because students' scores were

influenced not only by the conditions or treatments they received, but

also by membership in the classes to which they belonged. Because there

were three conditions in only five schools, resulting in a small number

of degrees of freedom with which to evaluate the hypotheses, standard

ANOOVAs were also performed on the dependent measures to provide more

pawer and lessen the chance of Type II error. Detailed listings of

means and statistical tests for all measures may be found in Appendix F.


School Success Classroom Behaviors


To investigate the effects of treatment on classroom behaviors,

two dependent variables were subjected to analyses. Teachers'

perceptions of student behavior were recorded on the School Success

Inventory-Teacher Form (SSI-TF). Students rated themselves in terms of

school success behavior on the School Success Inventory--Student Form

(SSI-SF). Overall scores for these forms were obtained by summing the

ratings on the 12 items in each scale. Results for each variable, with

their relevance to the hypotheses, are discussed below.


Teachers' Perceptions of Student School Success Behaviors


Hol: There are no significant differences among the
three groups on teachers' perceptions of student
school success behaviors, as measured by pre-post
change on the School Success Inventory--Teacher Form.

Overall analysis by conditions. Mean change scores showed that

teachers rated students in all conditions as using school success skills

more frequently at the time of the posttest (see Table 4-1). A negative

mean change score indicated improvement, because the highest frequency










Table 4-1

Pre-Post Mean Change Scores for Total School Success Inventory Teacher
Form (SSI-TF)



Modeling Coaching Control
n=102 n=101 n=96

Mean s.d. Mean s.d. Mean s.d.


Total -1.20 (7.85) -2.63 (6.91) -.70 (7.90)



Note: A negative change indicates improvement.
For further details see Table F-l (p. 160)
and Table F-13 (p. 181).
As the range for the total prescore or postscore on this
instrument was 12 to 60 (12 items x a score of 1-5 on each
dimension), the range of possible change scores was
+/- 48; in the total sample, the observed range of change
scores was -27 to 21.









of occurrence for each item was recorded as a one. Students in the

coaching condition increased more than those in the modeling condition,

and both increased more than students in the control condition.

However, a hierarchical ANCOVA revealed no significant differences among

the groups; therefore, Hol was not rejected. A standard ANCOVA also

failed to show significant differences among the groups. However,

further analyses revealed that there were differences between the

conditions which could be detected by tests of individual school data.

Individual school analyses by conditions. Individual ANCOVAs were

performed on each school's SSI-TF data separately to investigate further

effects possibly masked because of the pooling across schools. The mean

change scores in two schools (B & C) showed that students in the

modeling condition were rated as utilizing the school success skills

more often at posttest. The difference between modeling and the other

two conditions was significant only for students in school C (F(2,51) =

28.65, p <.001; see Table 4-2).

For the coaching condition, teachers rated students as using school

success skills more at posttest in four schools. AN]OVAs revealed that

significant effects for rated changes in coaching, as ccnpared to

modeling and controls, were seen for schools B and E (F(2,57) = 10.38, p

<.001; F(2,63) = 7.55, p <.001); however, this effect occurred for the

latter largely because of poorer scores for the other two groups. In

school D coaching was significantly different from modeling, with

coaching showing the most improvement (F(2,51) = 3.54, p <.04).

Teachers in school A in the control group rated their students

significantly different from the other two treatment groups. Thus,

significant differences were seen in four schools for the treatment









Table 4-2


Pre-Post Mean Change Scores for
Form (SSI-TF), by Condition and


Total School Success Inventory Teacher
Individual Schools


Modeling Coaching Control
n=102 n=101 n=96

Race Mean s.d. Mean s.d. Mean s.d.


School A 2.05a (6.54) -1.17a (7.84) -5.48b (10.44) *
School B -3.60a (5.30) -7.90b (6.42) -1.95a (3.05) *
School C -1.50a (4.26) .39b (4.72) -2.86b (4.44) *
School D 2.20a (6.27) -4.18b (6.45) -.28ab (4.85) *
School E 3.00a (6.34) -.77b (6.20) 6.81a (6.67) *



Note: A negative change indicates improvement.
= Overall F, p <.05
Means with different letters are significantly different
with a Bonferroni t test, p <.017 (after adjusting
for the covariate).
For further details see Table F-2 (p. 161)
and Table F-14 (p. 182).










conditions, with students in three schools more improved in the coaching

condition and students in one school more improved in the modeling

condition.

While the overall ANCOVA did not shcw significant differences among

the groups, some effects of the experimental conditions could be

observed when the results were examined in terms of overall scores for

individual schools by condition.


Ho2: There are no significant differences among the three
groups on teachers' perceptions of target student's
school success behaviors, as measured by pre-post change
on the School Success Inventory-Teacher Form.

An ANCOVA was computed to test for significant differences among

the groups for target students. The criterion for designation as a

target student was a score of 4 or lower on the teacher's global school

success skill rating. The mean change scores showed that teachers rated

their students as using school success skills more frequently at

posttest for modeling, coaching, and control (see Table 4-3). However,

differences among these changes for the three groups failed to reach

significance; therefore, Ho2 was not rejected.


Ho3: There are no significant differences among the three
groups with respect to race or gender on teachers'
perceptions of student school success behaviors, as
measured by pre-post change on the School Success
Inventory--Teacher Form.

For white students, improvement was seen in all three conditions,

with coaching showing the most improvement (see Table 4-4). Black

students were rated improved in both the modeling and coaching

conditions, but not in the control condition. These differences were

reflected in the ANCOVA by an overall effect for race (F(1,231) = 11.07,










Table 4-3

Target Students: Pre-Post Mean Change Scores for Total School
Success Inventory Teacher Form (SSI-TF)



Modeling Coaching Control
n=32 n=21 n=21

Mean s.d. Mean s.d. Mean s.d.


SSI-TF -4.50 (7.83) -3.71 (8.07) -.52 (8.48)


Note: A negative change indicates improvement.
For further details see Table F-3 (p. 162)
and Table F-15 (p. 184).









Table 4-4

Pre-Post Mean Change Scores for Total School Success Inventory Teacher
Form (SSI-TF), for Conditions by Race



Modeling Coaching Control
n=101 n=99 n=94

Race Mean s.d. Mean s.d. Mean s.d.


Blacks (n=58) -.11 (8.76) -.22 (4.88) .96 (10.02)
Whites (n=236) -1.41 (7.72) -3.26 (7.24) -1.26 (7.24)


Note: A negative change indicates improvement.
For further details see Table F-4 (p. 163)
and Table F-16 (p. 185).










p <.001), with white students (n=236) being rated as improving in school

skills to a greater degree than black students (n=58). However, there

were no significant differences among the experimental groups on race as

a function of condition; therefore, Ho3 was not rejected.

There were no overall significant differences for gender on the

SSI-TF (see Table 4-5). Teachers rated males as displaying the school

success skills more frequently at posttest in all three conditions, with

the largest increment attributed to coaching. Females were also rated

similarly; students in all conditions were rated as using skills more

frequently, with coaching once again increasing the most.


Students' Perceptions of School Success Behaviors


Ho4: There are no significant differences among the three
groups on students' perceptions of school success behaviors,
as measured by pre-post change on the School Success
Inventory-Student Form.

Overall analysis by conditions. To investigate significant

differences among the three experimental conditions (i.e., modeling,

coaching, and control) on students' perceptions of school success

behaviors, a hierarchical ANCOVA, where the prescores were used as the

covariate, was performed on the change scores of overall SSI-SF data.

Results of the hierarchical ANOXVA showed no significant differences

among the groups; thus, Ho4 was not rejected. A standard ANCOVA also

failed to show any significant differences among the groups.

The mean change scores did not show improvement in modeling or

coaching (see Table 4-6). Only the students in the control condition

rated themselves improved. Students in the modeling and coaching










Table 4-5

Pre-Post Mean Change Scores for Total School Success Inventory Teacher
Form (SSI-TF), for Conditions by Gender



Modeling Coaching Control
n=101 n=99 n=94

Gender Mean s.d. Mean s.d. Mean s.d.


Males (n=147) -1.29 (7.81) -3.12 (6.89) -1.38 (9.15)
Females (n=147) -1.06 (8.04) -2.29 (7.06) -.16 (6.76)


Note: A negative change indicates improvement.
For further details see Table F-4 (p. 163)
and Table F-16 (p. 185).










Table 4-6

Pre-Post Mean Change Scores for Total School Success Inventory Student
Form (SSI-SF)



Modeling Coaching Control
n=102 n=101 n=96

Mean s.d. Mean s.d. Mean s.d.


Total .13 (6.06) 1.06 (5.93) -.10 (7.47)


Note: A negative change indicates improvement.
For further details see Table F-5 (p. 165)
and Table F-17 (p. 186).
As the range for the total prescore or postscore on this
instrument was 12 to 60 (12 items x a score of 1-5 on each
dimension), the range of possible change scores was
+/- 48; in the total sample, the observed range of change
scores was -23 to 27.









conditions rated themselves as using school success skills slightly less

frequently.

Individual school analysis by conditions. In order to examine the

experimental effects more closely, ANCOVAs were run on data from the

five individual schools. The mean change scores from some individual

schools showed students rating themselves as more frequently using

school success skills at posttest. However, ANCOVAs showed no

significant differences for treatment effects on the overall instrument.


Ho5: There are no significant differences among the three
groups on target students' perceptions of school success
behaviors, as measured by pre-post change on the
School Success Inventory-Student Form.

To investigate significant differences among the three experimental

groups on students' perceptions of school success behaviors for target

students, an ANCOVA was computed on the SSI-SF change scores, using the

pre SSI-SF scores as the covariate.

The mean change scores showed that target students rated themselves

as using school success skills more frequently at posttest in the

modeling and control conditions and less frequently in the coaching

group (see Table 4-7). The ANCOVA revealed that the groups were not

significantly different from one another; therefore, Ho5 was not

rejected.


Ho6: There are no significant differences among the three
groups with respect to race or gender on students'
perceptions of school success behaviors, as measured by
pre-post change on the School Success Inventory-Student Form.

ANCOVAs were performed to investigate differences among the three

groups with respect to race and gender on students' perceptions of

school success behaviors as measured by the SSI-SF. There were no










Table 4-7

Target Students: Pre-Post Mean Change Scores for Total School
Success Inventory Student Form (SSI-SF)



Modeling Coaching Control
n=32 n=21 n=21

Mean s.d. Mean s.d. Mean s.d.


SSI-SF -.91 (7.90) .62 (7.24) -4.38 (8.48)


Note: A negative change indicates improvement.
For further details see Table F-7 (p. 167)
and Table F-19 (p. 189).










Table 4-8

Pre-Post Mean Change Scores for Total School Success Inventory Student
Form (SSI-SF), for Conditions by Race



Modeling Coaching Control
n=101 n=99 n=94

Race Mean s.d. Mean s.d. Mean s.d.


Blacks (n=58) -.83 (7.03) 1.22 (7.22) -3.36 (9.41)
Whites (n=236) .34 (5.89) 1.01 (5.73) .92 (6.63)


Note: A negative change indicates improvement.
For further details see Table F-8 (p. 168)
and Table F-20 (p. 190).










Table 4-9

Pre-Post Mean Change Scores for Total School Success Inventory Student
Form (SSI-SF), for Conditions by Gender



Modeling Coaching Control
n=101 n=99 n=94

Gender Mean s.d. Mean s.d. Mean s.d.


Males (n=147) -.65 (6.67) .52 (5.40) -.47 (7.94)
Females (n=147) .96 (5.35) 1.59 (6.55) .26 (7.22)


Note: A negative change indicates improvement.
For further details see Table F-8 (p. 168)
and Table F-20 (p. 190).









significant differences for either race (see Table 4-8) or gender (see

Table 4-9) on the SSI-SF for conditions; therefore, Ho6 was not

rejected. Black students rated themselves improved in the modeling and

control groups and slightly lower in the coaching conditions. White

students rated themselves less successful in all three conditions.

Irrespective of condition, there was an overall interaction of race and

gender on the SSI-SF (F(1,281) = 8.56, p <.005). Black females and

white males rated themselves improved, with black males and white

females indicating a decrease in use of school success skills.


School Success Classroom Attitudes


In order to investigate the effects of developmental guidance units

on students' attitudinal ratings of school situations, analyses were

performed on the School Attitude Inventory (SAI).


Ho7: There are no significant differences among the three
groups on students' attitudinal ratings of school
situations, as measured by pre-post change on the School
Attitude Inventory.

Overall analysis by conditions. A hierarchical ANCOVA was

performed on total SAI change scores and for the separate dimensions of

pleasure, excitement, and control. Results of the hierarchical ANCOVAs

for total SAI scores, pleasure, excitement, and control showed no

significant differences with respect to conditions. However, results of

the standard ANCOVAs revealed a significant difference among the groups

for SAI total scores, (F(2,295) = 4.36, p <.01). Students rated

themselves more favorably in all conditions, with coaching showing the

largest increment (Table 4-10). Post hoc analyses showed students in

the coaching condition rating themselves significantly different from









Table 4-10

Pre-Post Mean Change Scores for School Attitude Inventory (SAI),
Total and by Dimensions



Modeling Coaching Control
n=102 n=101 n=96

Mean s.d. Mean s.d. Mean s.d.


Total 1.89a (12.83) 6.68b (16.29) 1.30a (15.14) *

Pleasure -.01 (5.36) 1.08 (5.72) .91 (5.34)

Excitement -.65ab (8.27) -2.67b (9.10) .23a (9.80) *

Control 1.26 (6.35) 2.93 (8.87) .62 (7.84)


Note: A negative change indicates improvement for the Excitement scale.
= Overall F, p <.05
Means with different letters are significantly different
with a Bonferroni t test, p <.017 (after adjusting
for the covariate).
For further details see Table F-9 (p. 170)
and Table F-21 (p. 191).
As the range for the total prescore or postscore on this
instrument was 30 to 150 (10 items x 3 dimensions per item
x a score of 1-5 on each dimension), the range of possible
change scores was +/- 120; in the total sample, the
observed range of change scores was -35 to 49. Observed
ranges for the dimensions (possible scores +/- 40) were:
Pleasure, -14 to 16; Excitement, -30 to 36;
Control, -27 to 32.









those in the control and modeling conditions (F(1,295) = 6.16, p <.01; F

(1,295) = 6.85, p <.01). In addition, there were significant

differences among the groups on the excitement dimension (F(2,295) =

3.93, p <.03). Post hoc analyses also showed students in the coaching

condition to be significantly different from those in the control group

(F(1,295) = 7.05, p <.01). Students in the coaching condition rated

themselves the most calm, followed by modeling. Students in the control

conditions rated themselves less calm.

Individual school analysis by conditions. ANCOVAs were carried out

on total scores, and on the three separate dimensions, for individual

schools with respect to conditions. The mean change scores for total

SAI increased for three schools for modeling, in all five schools for

coaching, and in three schools for the control group (Table 4-11).

ANCOVAs performed on individual school data indicated significant

differences for two schools for total SAI scores in the coaching

condition (F(2,51) = 3.71, p <.04; F(2,51) = 4.67, p <.02), where

students rated themselves improved.

On the pleasure dimension, students in two schools showed an

increase in the mean change scores in the modeling condition, suggesting

that they felt happier about school situations at posttest. Students in

all five schools in the coaching condition also rated themselves

happier, as did students in four schools in the control group. ANCOVAs

revealed significant differences among the three experimental conditions

in two of the schools, with the coaching condition contributing the most

change in the favored direction of students rating themselves happier in

school D (F(2,63) = 4.14, p <.03); the control group was most improved

in school E (F(2,63) = 3.70, p <.03).










Table 4-11


Pre-Post Mean Change Scores for
of Individual Schools for Total


School Attitude Inventory Scores
Score and Dimensions


Modeling Coaching Control
n=102 n=101 n-96

School Mean s.d. Mean s.d. Mean s.d.


Total Score
School A
School B
School C
School D
School E

Pleasure
School A
School B
School C
School D
School E

Excitement
School A
School B
School C
School D
School E


Control
School
School ]
School 4
School i
School


5.60
3.45
-. 06a
-1. 00ab
1.38


1.80
-0.40
1.00
-1.35a
-.83a


-2.90
-1.25ab
2.33a
-1.45
.17


.90
2.60
1.28
-1.10
2.38


(14.59)
(11.77)
(15.86)
(11.96)
(10.18)


(6.38)
(3.91)
(6.86)
(5.04)
(4.27)


(9.09)
(5.84)
(11.50)
(7.76)
(6.60)


(5.89)
(7.84)
(5.59)
(6.34)
(5.79)


2.22
7.14
11.74b
11. 00a
1.27


.39
1.29
.65
2.82b
.54ab


-1.72
-2.14a
-6.52b
-3.35
.59


.11
3.71
4.56
4.82
1.32


(19.97)
(17.25)
(15.58)
(17.64)
(8.66)


(6.16)
(5.82)
(6.37)
(5.80)
(4.65)


(10.64)
(7.85)
(9.51)
(9.06)
(7.59)


(7.61)
(9.86)
(7.02)
(11.69)
(7.95)


1.83
-1.20
6.21ab
-.28b
1.19


1.00
-.35
1.71
.50ab
1.81b


-2.04
1.95b
-1.21ab
.83
1.52


-1.22
1.10
3.29
.06
.90


(15.30)
(17.52)
(10.23)
(16.78)
(14.34)


(4.25)
(6.31)
(3.77)
(5.79)
(6.06)


(14.15)
(9.06) *
(9.08) *
(7.65)
(6.48)


(6.67)
(9.47)
(4.51)
(9.03)
(8.10)


Note: A negative change indicates improvement for Excitement.
= Overall F, p <.05
Means with different letters are significantly different
with a Bonferroni t test, p <.017 (after adjusting
for the covariate).
For further details see Table F-10 (p. 171)
and Table F-22 (p. 195).









On the excitement dimension, students in three schools rated

themselves less aroused or excited for the modeling condition. Students

in the coaching condition in four schools rated themselves more calm.

Students in two schools in the control condition also rated themselves

in the favored direction. An ANCOVA showed students in one school

showing significant differences among the groups (F(2,51) = 3.22, p

<.05), with coaching showing the largest change in terms of students

rating themselves more calm at posttest. A second school showed a

marginal effect for coaching on this dimension (F(2,57) = 3.04, p

<.056).

On the control dimension at posttest, students in four schools

rated themselves as feeling more in control in school situations for the

modeling condition. For the coaching condition, students in all five

schools rated themselves in the favored direction. Students in four

schools in the control condition also rated themselves improved. No

significant differences among groups were found on the control

dimension. Only a marginal effect emerged from one school (F(2,51) =

3.01, p <.058), with the modeling condition contributing the most change

in the means.


Hog: There are no significant differences among the three
groups on target students' attitudinal ratings of
school situations, as measured by pre-post change on the
School Attitude Inventory.

ANCOVAs were performed on total SAI scores and the scores for the

three separate dimensions (pleasure, excitement, and control) to

investigate the differences among the three experimental groups on

target students' attitudinal ratings of school situations. The target

students showed improvement on mean change scores for total SAI scores









in the coaching and control conditions; modeling students rated

themselves less improved (see Table 4-12). There were no significant

differences among the groups for total SAI scores; therefore, }Ho8 was

not rejected.

Improvements were seen on each of the three separate SAI

dimensions. Students rated themselves happier on the pleasure dimension

for all three experimental conditions. Students rated themselves calmer

on the excitement dimension in the coaching and control groups and

somewhat more excited in the modeling condition. Improvement was seen

on the control dimension across all three conditions. However, ANCOVAs

revealed no significant differences among the groups on any of the three

dimensions.


Hog: There will be no significant differences among the
three groups with respect to race or gender on students'
attitudinal ratings of school situations, as measured by
pre-post change on the School Attitude Inventory.

To investigate differences between the experimental groups with

regard to race and gender, ANCOVAs were performed on the SAI total score

and on each of the three dimensions separately. In these analyses,

there was a significant condition by race interaction on the pleasure

dimension of the SAI (F(2,281) = 3.19, p <.05), with black students

rating themselves happier in modeling and whites happier in the coaching

condition (Table 4-13). There was also a significant condition by race

interaction effect on the excitement dimension (F(2,281) = 3.31, p

<.04); white students rated themselves more calm in the modeling and

coaching conditions, with black students calmer only in the coaching

condition. There were no significant differences for conditions by race

on the total SAI scores or on the control dimension. There also were no










Table 4-12

Target Students: Pre-Post Mean Change Scores for School
Attitude Inventory (SAI)


Modeling Coaching Control
n=32 n=21 n=21

Mean s.d. Mean s.d. Mean s.d.


Total -.56 (14.53) 6.24 (16.48) 4.86 (17.08)

Pleasure .38 (6.84) 1.76 (5.16) 2.67 (5.60)

Excitement 1.28 (10.01) -3.86 (8.66) -2.14 (10.70)

Control .34 (6.95) .62 (11.16) .05 (9.01)


Note: A negative change indicates improvement on Excitement.
For further details see Table F-ll (p. 172)
and Table F-23 (p. 203).









Table 4-13

Pre-Post Mean Change Scores for School Attitude Inventory
(SAl), for Conditions by Race


Modeling Coaching Control
n=101 n=99 n-94

Mean s.d. Mean s.d. Mean s.d.


Total Scores
Blacks -1.28 (12.17) 5.11 (21.41) 4.09 (20.43)
Whites 2.35 (12.85) 6.70 (15.16) .51 (13.39)

Pleasure
Blacks .61 (6.01) -1.11 (6.52) 2.41 (6.67)
Whites -.16 (5.27) 1.48 (5.50) .35 (4.83)

Excitement
Blacks 2.22 (10.46) -6.00 (10.54) .54 (11.92)
Whites -1.17 (7.66) -1.72 (8.60) -.10 (9.18)

Control
Blacks .33 (6.79) .22 (11.51) 2.23 (8.54)
Whites 1.34 (6.23) 3.52 (8.25) .07 (7.70)



Note: A negative change indicates improvement on Excitement.
For further details see Table F-12 (p. 173)
and Table F-24 (p. 204).










significant differences for conditions by gender on total SAI scores or

on the separate dimensions (Table 4-14). However, because of the

findings for conditions by race on the pleasure and excitement

dimensions, Ho9 was rejected.


Student Summary Sheet


At the conclusion of all treatment and posttest measurements,

students were asked to give informal feedback on a Student Surmary sheet

(see Appendix G). The purpose of this instrument was to gain added

insight into how students felt about the units so that improvements

could be made for further implementation of the units. Mbst of the

students viewed the units favorably. Some of the croments are

summarized in Appendix H.


Summary


A summary of the results of the ANCOVAs for this study is presented

below. It is organized by the dependent variables related to school

success behaviors and the relevant hypotheses. The third dependent

variable, related to attitudinal ratings of school situations, is shown

next, with related hypotheses.


Teachers' Perceptions of Student School Success Behaviors as Measured by

the School Success Inventory-Teacher Form (SSI-TF)


1. There were no significant differences among the three groups on

teachers' perceptions of student school success behaviors.










Table 4-14

Pre-Post Mean Change Scores for Total School Attitude Inventory
(SAI), for Conditions by Gender



Modeling Coaching Control
n=101 n-99 n=94

Mean s.d. Mean s.d. Mean s.d.


Total Scores
Males 1.29 (13.77) 8.12 (16.44) 3.22 (14.98)
Females 2.14 (11.69) 4.69 (16.24) -.37 (15.50)

Pleasure
Males .04 (6.22) 1.72 (5.63) .89 (5.86)
Females -.08 (4.41) .29 (5.84) .78 (4.91)

Excitement
Males -.52 (9.01) -3.48 (9.64) -1.73 (10.11)
Females -.61 (7.50) -1.49 (8.44) 1.69 (9.37)

Control
Males .73 (6.32) 2.92 (9.74) .60 (7.92)
Females 1.61 (6.32) 2.92 (8.18) .55 (7.99)



Note: A negative change indicates improvement on Excitement.
For further details see Table F-12 (p. 173)
and Table F-24 (p. 204).










2. There were no significant differences among the three groups on

teachers' perceptions of target student school success behaviors.


3. There were no significant differences among the three groups

with respect to race or gender on teachers' perceptions of student

school success behaviors.


Students' Perceptions of School Success Behaviors as Measured by the

School Success Inventory--Student Form (SSI-SF)


4. There were no significant differences among the three groups

on students' perceptions of school success behaviors.


5. There were no significant differences among the three groups on

target students' perceptions of school success behaviors.


6. There were no significant differences among the three groups

with respect to race or gender on students' perceptions of school

success behaviors.


Students' Attitudinal ratings of School Situations as Measured by the

School Attitude Inventory (SAI)


7. There were significant differences among the three groups on

students' attitudinal ratings of school situations on overall scores and

the excitement dimension of the SAI.


8. There were no significant differences among the three groups on

target students' attitudinal ratings of school situations.







92


9. There were significant differences among the three groups with

respect to race, but not gender, on students' attitudinal ratings of

school situations.