   

HIM

WW

    
     
   

l

WWIWIWIWll‘ﬂW‘llHHllHl

(ID—s
0300
.b

ITH I

1007

This is to certify that the
thesis entitled

EXTENDING VALIDITY EVIDENCE FOR THE COACHING
EFFICACY SCALE
WITH VOLUNTEER YOUTH SPORT COACHES.

presented by
Nathan Roman

has been accepted towards fulﬁllment
of the requirements for the

MS. MS. degree in Kinesiology

 

 

(Dwim

Major Professor’s SigWature
O M 17L EQDO £9

Date

 

MSU is an Afﬁrmative Action/Equal Opportunity institution

 

 

‘ LIBRARY
Michigan State
University

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

 

Bari BUB

DB

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/Dale0ue.indd-p.1

 

EXTENDING VALIDITY EVIDENCE FOR THE COACHING EFFICACY SCALE
WITH VOLUNTEER YOUTH SPORT COACHES.

By

Nathan Roman

A Thesis

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Kinesiology

2006

ABSTRACT
EXTENDING VALIDITY EVIDENCE FOR THE COACHING EFFICACY
SCALE WITH VOLUNTEER YOUTH SPORT COACHES
By
Nathan Roman

The purpose of this study was to validate the structural validity of the Coaching Efﬁcacy
Scale (CES), developed by Feltz et a1. (1999) with youth sport coaches. The CBS has four
dimensions: motivation, strategy, technique, and character building. Coaches (N = 395)
from various sports such as hockey, volleyball, basketball, and football participated in the
study using a condensed scale as suggested by Myers et al. (in press). Using conﬁrmatory
factor analysis, the scale was found to be in line with previous studies (F eltz et al., 1999;
Lee et al., 2002; Myers et al., 2005; Sullivan & Kent, 2003). Multivariate multiple
regression results showed sources of efﬁcacy such as team improvement and athlete
support were signiﬁcant predictors of motivation, strategy, and technique efﬁcacy. Years

of playing experience was a signiﬁcant predictor of strategy and technique efﬁcacy.

Table of Contents

LIST OF TABLES .................................................................................... v
LIST OF FIGURES ..................................................................................... vi
CHAPTER 1:

INTRODUCTION ................................................................................... 1
Nature of the Problem ...................................................................... 1
Purpose of the Study ....................................................................... 5
Research Questions ........................................................................ 6
Deﬁnitions ...................................................................................... 6
Delimitations ................................................................................. 7
Limitations ..................................................................................... 7

Basic Assumptrons7

CHAPTER 2:

Literature Review ..................................................................................... 8
Youth Sports .................................................................................. 8
Youth Sport Coaches ....................................................................... 9
Coaching Effectiveness Model .............................................................................. 9
Coaches Inﬂuence of Prosocial Attitudes ............................................................ 12
Horn’s Coaching Effectiveness Model ................................................................ l4
Horn’s Model of Coaching Effectiveness ............................................................ 16
Coaching Efﬁcacy ......................................................................... 17
Psychometric Properties of the CES .................................................... 25

CHAPTER 3:

Methods .............................................................................................. 28
Participants ................................................................................. 28
Measures .................................................................................... 28
Procedure ..................................................................................... 29
Treatment of Data ............................................................................ 30

CHAPTER 4:

Results ............................................................................................... 32
Conﬁnnatory Factor Analysis ........................................................... 32
Sources of Coaching Efﬁcacy ........................................................... 32

CHAPTER 5:

Discussion ............................................................................................. 37
Sources of Coaching Efﬁcacy ........................................................... 37
Limitations ........................................................................................................... 41
Future Direction ................................................................................................... 42

APPENDICES ...................................................................................... 44

iii

REFERENCES ....................
.......................................................... 62

iv

LIST OF TABLES

Table 1 Means, Standard Deviations and Frequency of Response for Each Item
...................................................................................................... .52

Table 2 Descriptive Statistics for the Proposed Sources of Coaching Efﬁcacy, Coaching
Efﬁcacy Dimensions, and Team Variables ...................................................... .53

Table 3 Correlations of Sources with Total Coaching Efﬁcacy and Dimensions of

Coaching Efﬁcacy ................................................................................. .54
Table 4 Summary of the MMR Analysis and Standardized Beta Coefﬁcients for
Predictors CES ..................................................................................... .55
Table 5 Summary of the Canonical Correlation Analysis Between Sources of Coaching
Efﬁcacy and Coaching Efﬁcacy Subscales ...................................................... 56
Table 6 Descriptive Information of Coaching and Playing Years Experience. .......... .57

Table 7 Descriptive Information of Hours of Preparation In-Season and Out-of-
Season ................................................................................................ 58

Table 8 Descriptive Comparisons Of Efﬁcacy Sources Between Ice Hockey Coaches and
Other Sport Coaches .................................................................................. 59

Table 9 Descriptive Comparisons of Efﬁcacy Dimensions Between Ice Hockey Coaches
and Other Sport Coaches ........................................................................... 6O

LIST OF FIGURES

Figure 1 Horn’s Model of Coaching Effectiveness ............................................ 61

vi

CHAPTER 1
INTRODUCTION

Nature of the Problem

Participation in non-school youth sports is prevalent in America today as millions
of children participate in programs sponsored by recreation departments, Little League
baseball, Pop Warner football, and many other organized groups. These are programs that
are organized by adults for children and youth, typically between the ages of 7 and 18
years, which have designated coaches, organized practices, and scheduled competitions
(Gould, 1982). Approximately 2.5 million coaches volunteer each year to help teach and
lead approximately 13 million children who participate in youth sports (Seefeldt &
Ewing, 1996). These youth sport coaches are integral in properly teaching the
foundational aspects of sport, and their role is similar to the role teachers have with their
students. They can have a strong inﬂuence on the beneﬁts, negative experiences, and
motivation surrounding youth sports participation (Gould, Hodge, Peterson, and
Giannini, 1989; Guivemau & Duda, 2002). Research has shown that a coach’s
effectiveness in fostering the foundational and beneﬁcial aspects of sport is inﬂuenced by
self-efﬁcacy (or self-conﬁdence) beliefs for coaching (F eltz, Chase, Moritz, & Sullivan,
1999; Myers, Vargas-Tonsing, & Feltz, 2005). However, the coaching efﬁcacy of
coaches at the youth sport levels has not been studied despite its potential inﬂuence on
the sport experience of children and youth.

Coaching efﬁcacy is the belief coaches have on effecting the learning and

performance of their athletes (Feltz et al., 1999). Feltz et a1. developed a model of

coaching efﬁcacy comprised of four dimensions of coaching: game strategy, motivation,
teaching techniques, and character building. These dimensions were based on Bandura’s
(1997) self-efﬁcacy theory and Denham and Michael’s (1981) model of teacher efﬁcacy.
Game strategy efﬁcacy is the conﬁdence that coaches have to lead their teams to a
successful outcome, such as being able to understand the competitive strategies of the
sport and recognizing the strengths and weaknesses of the opposing team. Motivation
efﬁcacy involves the belief that coaches have to affect the psychological attributes of
their athletes, such as how well they can motivate the athletes and build team conﬁdence
and cohesion. Teaching Technique efﬁcacy is concerned with the ability to effectively
demonstrate skills, recognize talent, and diagnose skill errors. Finally, Character Building
efﬁcacy is deﬁned as the conﬁdence that coaches have in their ability to foster a sense of
fair play and responsibility toward sport and other participants. They also proposed a
unidimensional conceptualization, total coaching efﬁcacy, that combines all four factors.
Within the coaching efﬁcacy model, Feltz et a1. (1999) hypothesized that
coaching efﬁcacy has an inﬂuence on coaching behavior, player/team satisfaction,
player/team performance, and player/team efﬁcacy. Preliminary ﬁndings showed more
conﬁdent coaches display more ﬁequent uses of positive reinforcement and general
encouragement behavior, having higher winning percentages, and having more satisﬁed
athletes. Other studies have shown the Coaching Efﬁcacy Scale (CES) dimensions to be
correlated with leadership styles (Sullivan & Kent, 2003), and athletes’ self-efﬁcacy,
satisfaction with the coach, and team performance in line with tenets of the coaching

efﬁcacy model (Myers et al., 2005; Vargas-Tonsing, Wamers, & Feltz, 2003).

Coaching efﬁcacy, in turn, was proposed, by F eltz et al. (1999), to be inﬂuenced
by coaches’ extent of coaching experience/preparation, prior success (won-lost record),
perceived skill of athletes, and school/community support. In addition to the sources of
coaching efﬁcacy proposed by Feltz et a1. (1999), Chase and her colleagues (Chase,

F eltz, Hayashi, & Hepler, 2005) identiﬁed additional sources of coaching efﬁcacy
information through structured interviews with coaches. They found that coaches also
relied on evidence of player improvement and development, support from their players,
and their own previous playing experience to judge their coaching efﬁcacy. Sullivan,
Gee, and F eltz (in press) also found that coaches’ own playing experience uniquely
predicted Game Strategy efﬁcacy.

Feltz et a1. (1999) developed an instrument to measure the four dimensions of
coaching efﬁcacy. One component to developing the measure was a seminar involving 11
coaches who were graduate students in sport psychology. The seminar lasted 5 weeks and
the coaches had varying coaching education backgrounds. The four dimensions that
emerged from the seminar with coaches and former coaches of high school sports -- game
strategy, motivation, technique, and character building -- led to the generation of 41
items. Nine collegiate and scholastic coaches evaluated the relevance of the items and
determined that all were potentially important indicators of coaching efﬁcacy. However,
17 of the original items were later dropped as a result of factor analysis (F eltz et al.,
1999).

Conﬁrmatory factor analysis (CF A) of the 24-item CES with its four dimensions
demonstrated psychometric ﬁt indexes that were minimally acceptable but consistent in

samples of high school and small-college level coaches. Fit indices ranged from 87-89

for Comparative Fit Index (CF I), and 080-085 for Root Mean Square Residual Error of
Approximation (RMSEA) across four studies for the multidimensional model (F eltz et al.
1999; Lee, Malete, & Feltz, 2002; Myers, Wolfe, & Feltz, 2005; Sullivan & Kent, 2003).
Feltz et al. (1999) also tested the ﬁt of the unidimensional model. Slightly greater misﬁt
was observed for this model, ,1; = 844, non-normed Fit Index (NNF I) = .87, CFI = .88,
and RMSEA = .09, as compared to the multidimensional model but is still useful in
instances when coaching efﬁcacy is but one of a host of variables used to predict a
general outcome (e.g., performance), and/or when subscale scores are highly related and
likely to cause problems associated with multicollinearity within the speciﬁed data
analysis (Myers et al., 2005).

Myers et al. (2005) also examined the degree to which high school and small-
college level coaches employed the original 10-category rating scale structure of the CES
(where 0 signiﬁed “not at all conﬁdent” and 9 signiﬁed “extremely conﬁdent”)
systematically. They reported that coaches did not employ the original rating scale
structure systematically because they were being asked to distinguish between too many
levels of coaching efﬁcacy, which was congruent with previous ﬁndings for the optimal
structure of an ordered response efﬁcacy scale (Zhu et al., 1997) and long-standing
recommendations for Likert scales (Likert, 1932). A post-hoc analysis suggested that a
4-category structure was optimal for the Myers et al. sample. Accordingly, they
suggested that subsequent users of the CES should use a condensed rating scale structure.

For youth sport coaches concentrating on the fundamentals of the sport, the factor
structure of the measure may not be appropriate due to the dimensions that may vary

depending on the age and level of the athlete or level of the coach. Youth-sport coaches

have an impact on athletes during the developmental stages of sports participation. Their
inﬂuence includes teaching the rules and strategies of the sport, the fundamental skills
needed to perform, and even developing the character to handle social-emotional
situations in sport (i.e. sportsmanship). In addition, many youth sport organizations
downplay the emphasis of outcome goals such as winning. Thus, the factor structure of
the CES should be examined through CFA before hypotheses within the coaching
efﬁcacy model are tested with volunteer youth-sport coaches.

In addition, youth-sport coaches may rely on different sources of efﬁcacy
information on which to base their conﬁdence in coaching than high school or small
college coaches. If outcome goals are downplayed in youth sports, a coach’s perceptions
of the tearn’s ability may not be as important as perceptions of how the team has
improved from one season to the next. Past winning record and school/community
support may be of no importance, but parental and player support may be very important.
Thus, the most salient sources of efﬁcacy information for youth-sport coaches should
also be examined to help expand the model and aid in helping coaches enhance their

coaching efﬁcacy.

Purpose of the Study

The purpose of this study was to examine the structural validity of the CES (F eltz
et al., 1999) with youth sport coaches using a condensed rating scale structure. The CBS
is currently the only instrument for measuring coaching efﬁcacy, thus it is important that
the measurement is validated for the assessment of youth sport coaches’ coaching
efﬁcacy. In addition, the sources of coaching efﬁcacy proposed by Feltz et al. (1999) and

Chase et al. (2005), namely, years of coaching experience, extent of coaching education,

 

parent, organizational, and player support, and previous playing experience, will be

examined as predictors of coaching efﬁcacy.

Research Questions

1. Can the factor structure of the CES be replicated with a sample of volunteer youth
sport coaches using a condensed rating scale structure?

2. What are the strongest dimensions of coaching efﬁcacy among male and female
volunteer youth-sport coaches?

3. What are the strongest predictors of coaching efﬁcacy among the sources
measured?

Deﬁnitions

. CES- Coaching Efﬁcacy Scale developed by Feltz et a1. (1999) assesses the

dimensions of game strategy, motivation, teaching technique, and character

building.

. Character-building Efﬁcacy—Coach’s belief that he or she can foster a sense of

fair play and responsibility toward sport and other participants.

. Coaching Efﬁcacy- Extent to which coaches believe they have the capacity to

effect the learning and performance of their athletes (Feltz et al., 1999).

. Game Strategy Efﬁcacy-Coach’s belief he or she can lead the team to a successful

outcome, effectively pinpoint opponents’ weaknesses, and make in game

adjustments.

. Motivation Efﬁcacy-Coach’s belief he or she can build team cohesion, increase

conﬁdence, and motivate athletes.

6. Teaching Technique Efﬁcacy-The coach’s belief he or she can teach fundamental
techniques in sport and recognize talent, and diagnose skill deﬁciencies in the
athlete(s).

7. Youth Sport Coach- Usually a person who volunteers his or her time to coach

children whose ages range from 4 to 12 years.

Delimitations

The study was delimited to volunteer youth sport coaches in the United States.
Results of this study may not generalize to other types of coaches, such as high

school, college, or paid club-level coaches or to coaches outside of the U. S.

Limitations

The study was limited in terms of nonrandom selection of coaches, potentially
causing a subject self-selection bias. A non-representative sample was the most likely
threat to external validity in this study. This can occur when participants are selected
on some basis other than random assignment. In the present study, the investigator
sought coaches of various youth sports. The sports were hockey, basketball,
volleyball, cheerleading, and football, however, 73% of the coaches came from the

sport of hockey.

Basic Assumptions

1. Participants responded honestly to all questions on the survey.
2. Participants comprehended all questions on the survey.
3. The questionnaires constructed to measure sources of coaching efﬁcacy

accurately measured these sources.

CHAPTER 2
REVIEW OF LITERATURE

In this chapter, an overview of research on youth sport coaches is followed by a
summary of the model of coaching efﬁcacy, developed by Feltz et a1. (1999).
Additionally, the development of the CES (F eltz et al., 1999) and the associated
psychometric properties are presented. Finally a discussion on the need for further
development of the CBS for the youth sport coaching population is presented.
Youth Sports

Participation in non-school youth sports is widespread in America today. Between
13 and 20 million children participate in programs sponsored by recreation departments,
Little League baseball, Pop Warner football, and many other organized groups (Martens,
1986; Seefeldt & Ewing, 1996). Youth sports beneﬁt the development of today’s children
in many ways. The beneﬁts include moral development (Chambers, 1991), perceptions of
competence (Feltz & Ewing, 1987; Feltz & Petlichkoff, 1983), self-esteem (Weiss, 1987),
and life skills such as developing self-control, persistence, teamwork, and learning to
cooperate with teammates (Shields & Bredemeier, 1995; Weiss & Bredemeier, 1990).

The success of these programs depends on youth sport coaches to volunteer their
time in an attempt to provide maximal development for the youth athletes.
Approximately 2.5 million coaches volunteer each year, usually following the sport
involvement of their own children, but are equally important to the areas of youth

development.

Youth Sport Coaches

Volunteerism has been the backbone of youth sport coaching. These coaches are
present in the lives of these young athletes as they start their athletic careers,
consequently these coaches have a potentially large impact on the sport experience of
their athletes (Litherland, 1996). Coaches believe youth athletes should learn life skills,
have fun, develop conﬁdence, and learn how to be a part of the team (Lesyk & Komspan,
2000). Research has evaluated the impact that a coach can have on an athlete in such
areas as self-esteem, performance anxiety, and enjoyment (Scanlan & Lewthwaite, 1986;
Smith & Smoll, 1997). Recent studies have shown that coaches do have an inﬂuence,
both positive and negative, on their athletes’ lives (Gould, Hodge, Peterson, & Giannini,
1989; Guivemau & Duda, 2002; Vargas-Tonsing et al., 2003). More research is needed
to understand the types of behaviors, expectancies, and attitudes of coaches that impact

young athletes in their sport experiences.

Coach eﬂectiveness model.

Smoll, Smith, Curtis, and Hunt (1978) presented a model for conceptualizing and
investigating coach-player relationships. Acknowledging the psychosocial impact that
sport has on children and the effect a coach can have on a player, the authors set out to
assess the interrelation of coaches’ behaviors, players’ perceptions of those behaviors,
and player attitudes. Using the Coaching Behavior Assessment System (CBAS) and
structured interviews of the players, results suggested that these athletes tended to engage
in positive attitudes towards their coach, sport, and team when the coach exhibited

supportive and instructional behaviors.

Later, Smith and Smoll (1991) set out to look at the behavioral research and
intervention in youth sport. They recognized that adult leaders occupy a critical role in
youth sports and the impact made on young athletes. To understand this athlete-coach
relationship, Smith and Smoll used a multi-method approach.

The multi-method approach used included observer coding of overt behaviors,
measures of player perception and recall of those same coaching behaviors, and
measurement of players’ evaluative reactions. To measure leadership behaviors in this
study, the CBAS was again used. The coaches were given several self-report measures at
the end of the season and data from 542 players were collected during individual
interviews also conducted at the end of the season.

Scores revealed a relationship between coaches’ scores on the behavioral
dimensions and player measures. Players responded favorably to coaches who exhibited
higher proportions of supportive and instructional behaviors. Consequently these same
players liked their teammates more than players who did not play for a coach who
developed a supportive environment. Final ﬁndings revealed a signiﬁcant interaction
between coach supportiveness and athletes’ level of self-esteem.

Smoll, Smith, Barnett, and Everett (1993) set out to study the impact that a
coach’s behavior can have on an athlete’s self-enhancement processes. In a previous
study, Smith, Smoll, and Curtis (1978) discovered that athletes who played for coaches
who were highly supportive had higher postseason scores on the Coopersmith’s (1967)
General Self-Esteem Scale. This study attempted to assess the change in boys’ self-
esteem before and after being coached by coaches who were trained in how to be highly

supportive and in how to give quality instruction. Based on the self-enhancement model’s

10

assumption that boys who have lower self-esteem would beneﬁt from positive feedback
and Smith and Smoll’s (1990) previous ﬁnding that low self-esteem boys are highly
responsive to variations in supportiveness and instruction, the authors developed their
hypothesis for the study. Their hypothesis was that those who played for trained coaches
would experience an increase in self-esteem more than boys who played for untrained
coaches.

Participants were 18 male coaches and 152 male Little League Baseball players.
Eight coaches comprised the experimental group and 10 made up the control group. The
experimental group went through Coach Effectiveness Training (CET; Smoll & Smith,
1993)

The CET’s goal was to increase four speciﬁc target behaviors: reinforcement,
mistake-contingent encouragement, corrective instruction, and technical instruction. An
intervention was designed to increase the supportiveness and instructional effectiveness
of the coaches. Evaluation of the coaches’ behaviors and boys’ progress was
accomplished by comparing the experimental and control groups’ perceptions of their
coaches’ behaviors, their attitudes toward the coaches and other aspects of participation,
and their levels of self-esteem.

Three major results came from the study. First, players perceived behavioral
differences between trained and untrained coaches. Players thought the trained coaches
were more positive than the untrained coaches. The second result revealed player
evaluative responses favored the trained coaches who exhibited behavioral differences.

They liked playing for the trained coaches more than the untrained coaches. Lastly, the

11

authors found a statistically signiﬁcant increase in the boys’ self-esteem for those who
played for the trained coaches and who started with low self-esteem

Horn (1985) recognized the need to look at the relationship between coaching
behaviors and athletes’ self-perceptions. The purpose of her study was to observe the
kind of feedback that female athletes received over the course of a season in combination
with the skill mastery they achieved over that period of time.

Five coaches and 72 athletes from ﬁve junior high softball teams comprised the
participants for this study. The CBAS was used to record and classify coaches’ behaviors.
To assess changes in the athletes’ self-perceptions were the Perceived Competence Scale
for Children (Harter, 1982), the Multidimensional Measure of Children’s Perceptions of
Control (Connell, 1980), and the Generalized Expectancy of Sport Success Scale
(Coulson & Cobb, 1979), which measured the strength of the athletes’ expectations for
future athletic success. Determining success of performance was conducted through
evaluation by teammates.

Results revealed a statistically signiﬁcant relationship between a child’s
performance, adult’s evaluation of the performance, and the development of the child’s
perceived competence. These results imply that the coach has the potential to inﬂuence
how a child perceives success and failure based on evaluation that in turn can affect the

child’s own efﬁcacy.

Coaching inﬂuence of prosocial attitudes.

In addition to inﬂuencing young athletes’ perceptions of success, positive

experience, and self-esteem, coaches can also inﬂuence prosocial attitudes and behaviors.

12

Studying the effects of team’s collective norms on athletes’ behaviors was done by
Shields, Bredemeier, Gardner, and Bostrom (1995). The purpose of the study was to
examine team norms about cheating and aggression in relation to leadership, cohesion,
and demographic variables. Shields et al. alleged a player’s likelihood to use unfair game
tactics was the belief that their teammates would engage in the same tactics and the
motivational orientation of their coach, which superceded their own motivational
orientation.

Baseball (n=182) and softball (n=1 16) players from six community colleges and
six high schools were recruited for the study. A demographic questionnaire was given
along with the Group Environment Questionnaire (GEQ), the Team Norm Questionnaire,
and three versions of the Leadership Scale for Sports (LSS). These were given out the last
third of the regular season and administered by the researchers.

Assessment of the leadership styles of the coaches (as perceived by the athletes)
appears to determine the team norms of aggression and sanctioning of cheating. Also of
interest were female athletes who had a female coach. These athletes had a lower
expectation of team norms for cheating and aggression than female athletes who had a
male coach. Adding to this, female athletes coached by women believed behaviors of
cheating and aggression were less accepting than male coaches. These ﬁndings suggest
that the coach can impact an athlete’s moral judgment.

Guivemau and Duda (2002) found athletes’ coaches inﬂuenced them more than
any other signiﬁcant individual listed when determining the athletes’ likelihood to
aggress in soccer. They reason that moral reasoning maturity has an inverse correlation

with aggressive tendencies because individuals do not live in a vacuum. Interactions with

13

the community affect their decisions and ultimately their behaviors. These interactions
with signiﬁcant others (i.e. parents, coaches, and teammates) inﬂuence athletes’
attitudes, decisions, and actions.

Guivemau and Duda examined the relationship between the atmosphere of
athletic teams to the athletes’ self-described likelihood to aggress (SLA). One additional
purpose was to assess whether there was a “predominant ﬁgure most inﬂuential to
athletes’ SLA.”

Participants were 194 male and female soccer players ranging in age from 13 to
19 years of age who played for the same coach on average of 1 to 3 years. The study was
administered during a mid-western university’s summer camp, and the instruments
consisted of demographic information, J udgrnents about Moral Behavior in Youth Sport
Questionnaire (J AMBYSQ), and Team Norm Questionnaire (TNQ).

Results revealed support for the hypotheses. A signiﬁcant positive correlation was
found between athletes’ perceived team pro-aggressive norms and their SLA. In regards
to the inﬂuence of signiﬁcant others on athletes’ aggressive tendencies, the data revealed
that the coach was most inﬂuential in the athletes’ decision to engage in an inappropriate
act. These ﬁndings support Shields et al. (1995) who said team norms sanctioning,

cheating, and aggression may be inﬂuenced by leadership style.

Horn ’s coaching effectiveness model.
Horn (2002) developed a model to explain the relationship between coaching
effectiveness and the athlete (see Figure l). Hom’s coaching effectiveness model gives a

comprehensive look at a coach’s effect on athletes. The model comprises 10 boxes with

14

the coaches’ expectancies, values, beliefs, and goals as the central box (Box 4). The
antecedents for these (Boxes 1-3 respectively) are socio-cultural context, organizational
climate, and coaches’ personal characteristics that have an interrelation between the three
as well as on Box 4. Coaches’ behavior (Box 5) is inﬂuenced by Box 4 and has an impact
on athletes’ performance and behavior (Box 6) and athletes’ perceptions, interpretation,
and evaluation of their coaches’ behavior (Box 8). Athletes’ personal characteristics (Box
7) have an effect on Box 8 as well leading to an effect on athletes’ self-perceptions,
beliefs, and attitudes (Box 9). This leads to an affect on athletes’ level and type of
motivation (Box 10) and both Boxes 9 and 10 have an effect on athletes’ performance
and behavior (Box 6). The only effect Box 6 has in the model is on the coaches’
expectancies, values, beliefs, and goals (Box 4). Horn laid out a model of the
interrelationship between various variables in sport on coaches and athletes; she also
looked at coaches’ feedback and its effect on children’s perceptions of physical
competence. The model incorporated the Feltz et al. (1999) model of coaching efﬁcacy.
The Feltz et al. model is the only one that has a measure of coaching efﬁcacy

expectations.

15

 

 

Box 1
Sociocultural
Context

 

 

A

 

 

Box 7
Athltes’
personal
characteritics

 

 

 

 

 

 

 

Box 2
Organizational
Climate

 

 

 

Box 4
Coaches’
expectancies,
values, beliefs,

k and goals.

r—>

 

Box 5
Coaches’
behaviors

/

 

 

 

A

 

 

Box 3
Coaches’
personal
characteristics

 

 

 

 

 

Box 8

Athletes’
perceptions,
interpretation, and
evaluation of their
coaches’ behavior

 

 

 

 

 

 

Box 6
Athletes’
performance
and behavior

 

./

Box 9
Athletes’ self-
perceptions,
beliefs, and
attitudes

 

 

 

Figure l. Hom’s Model of Coaching Effectiveness

 

 

 

 

Box 10
Athletes’ level
and type of
motivation

 

 

 

Coaching Eﬁ‘icacy

Bandura (1986) states efﬁcacy to be, “peoples’ judgments of their capabilities to
organize and execute courses of action required to attain designated types of
performances. It is concerned not with the skills one has but with judgments of what one
can do with whatever skills one possesses” (pg. 391). Feltz et al. (1999) developed a
model of coaching efﬁcacy based on Bandura’s construct of self-efﬁcacy and models of
teaching efﬁcacy (e.g., Denham & Michael, 1981). Feltz et al. used the teacher efﬁcacy
model as a basis from which to develop their own because teachers and coaches have
similar end goals for their students and athletes respectively. Denham and Michael (1981)
deﬁned teaching efﬁcacy as, one’s sense that the ideal teaching can bring about positive
changes in a student than their assessment of their abilities to bring such changes.

However, Feltz et a1. considered the teacher efﬁcacy model to be inadequate for
the study of coaching efﬁcacy because coaches have unique objectives in the
development of athletes’ performance and they are also inﬂuenced by different
organizational variables (e.g., voluntary participation). For instance, the actual level of
importance placed on performance and performance measures may be more salient in
sport, because performance in sport, even at the youth level, is typically a public event.
Therefore, Feltz et al. (1999) deﬁned coaching efﬁcacy as “the extent to which coaches
believe they have the capacity to effect the learning and performance of their athletes”
(p.765).

In examining the teacher efﬁcacy literature, Feltz et al. (1999) found a number of

relationships that could extend to coaches. For instance, teachers’ behaviors are

17

inﬂuenced by their efﬁcacy (Coladarci, 1994; Denham & Michael, 1981; Fisher, et al.,
1978; Gibson & Dembo, 1984; Hoy & Woolfolk, 1993; Stinnet, 1970). Hoy and
Woolfolk (1993) suggest that teachers who possess high levels of efﬁcacy believe they
can motivate even the most difﬁcult students. A teacher’s ability to be ﬂexible with his or
her plans is related to high efﬁcacy (Fisher et al.,l978), suggesting an efﬁcacious teacher
adjusts his or her behavior according to the goals set out (i.e. give feedback when
appropriate, create a task-oriented atmosphere, etc.) and does not solely adhere to a strict
set of guidelines. Teacher dropout appears to be inﬂuenced by efﬁcacy as well according
to Stinnet (1970). Other studies give support to efﬁcacy’s impact on teachers’ behaviors.

Denham and Michael (1981) sought to investigate teacher efﬁcacy and its effects
on teaching effectiveness. Through the literature, consulting with various specialists in
the ﬁelds of psychology, sociology, and education they developed a model to describe
Teacher Sense of Efﬁcacy. A teacher’s cognitive and affective components have a
reciprocal effect on their consequences and antecedent convictions. The cognitive aspect
is comprised of a sense of likelihood a teacher can bring about positive change in a
student and an assessment of their own ability to bring about such changes. The affective
aspect is a teacher’s sense of efﬁcacy and the pride or shame associated with the sense of
efﬁcacy. Measurable consequences were deﬁned as teacher behaviors and student
outcomes. Sense of efﬁcacy, in particular the cognitive component, was composed of two
phenomena: the extent to which a teacher believes an ideal or normative teacher can help
a student change positively given circumstances and the extent to which that teacher
believes he or she can bring about positive changes in a student under given

circmnstances.

18

Antecedents for teaching efﬁcacy are similar to antecedents for coach efﬁcacy:
teacher training, teaching experience, system variables, personal variables, and causal
attributions (Denham & Michael, 1981; Fuller, Wood, Rapport, & Dombusch, 1982; Hoy
& Woolfolk, 1993; Park, 1992; Ramey—Gassert, Shroyer, & Staver, 1996). Teacher
training is to have increased efﬁcacy by giving them knowledge they feel essential to
teaching and directly increasing their sense of self-efﬁcacy.

Throughout scholarly literature, a theme was represented that alluded to the fact
that the more experience a teacher had in the ﬁeld the more efﬁcacy that came ﬁ'om it.
Learning from failures and feedback from students helped in forming the experience
(Denham & Michael, 1981; Gibson & Dembo, 1984).

Teacher behaviors and student behaviors comprise measurable consequences.
Teacher behaviors include classroom behaviors, support of innovation, professional
activities, and commitment to teaching. Student behaviors are achievement outcomes,
affective outcomes, and behavioral outcomes. More efﬁcacious teachers were found to
spend more time instructing/teaching, maximizing the time students spend in productive
activities (Bandura, 1997; Gibson & Dembo, 1984).

The most prominent ﬁnding in these studies was the strong relationship between
teacher efﬁcacy and improved student performance and the students’ gain in learning. As
a teacher, the goal is to see improvements in the pupils and learning to occur. According
to Berman et al. (1977) teacher efﬁcacy has a strong relationship with the variables
concerned.

Coaching efﬁcacy stems from the research done in teaching efﬁcacy. Coaching is

similar to teaching as coaches give instruction, are concerned with the learning and

19

performance of their athlete, and give feedback. Much of the teacher efﬁcacy literature
was used to develop the Feltz et al. (1999) model of coaching efﬁcacy. The model
consists of four sources of coaching efﬁcacy information: 1) extent of coaching
experience/preparation, 2) prior success (won-loss record), 3) perceived skill of athletes,
and 4) school/community support. These affect the four dimensions of coaching efﬁcacy
1) game strategy, 2) motivation, 3) teaching technique, and 4) character building.

The dimensions of coaching efﬁcacy were developed as a result of a 5-week
seminar, a review of the National Standards for Athletic Coaches, a review of the
coaching education literature, and preliminary work done by Park (1992). Teaching
technique efﬁcacy is the conﬁdence a coach has in teaching the skills of the sport and
includes discipline, training, and conditioning. Game strategy is a coach’s conﬁdence in
identifying an opponent’s weakness and making in-game adjustments. Motivation
efﬁcacy involves the belief that coaches have to affect the psychological attributes of
their athletes, such as how well they can motivate the athletes and build team conﬁdence
and cohesion. Character building efﬁcacy is deﬁned as the conﬁdence that coaches have
in their ability to foster a sense of fair play and responsibility toward sport and other
participants. From these four dimensions the Coaching Efﬁcacy Scale (CES) was
developed.

The CBS emerged from this study as an instrument to be used to measure the
nature of coaching efﬁcacy. The instrument includes the stem, “How conﬁdent are you in
your ability to...” then the sentence is completed with an item regarding one of the
dimensions. The CBS originally was scored on a 10-point Likert scale ranging from 0

(not at all conﬁdent) to 9 (extremely conﬁdent). A total of 24 items for the four

20

dimensions comprise the survey. There are a total of eight items for teaching technique
efﬁcacy for example, “How conﬁdent are you in your ability to demonstrate the skills of
your sport?” Eight items for game strategy; “How conﬁdent are you in your ability to
recognize opposing team’s strengths during competition?” motivation efﬁcacy: “How
conﬁdent are you in your ability to help athletes maintain conﬁdence in themselves?” and
character building efﬁcacy; “How conﬁdant are you in your ability to instill an attitude of
fair play among your athletes?” Subsequent studies led to reﬁnements of coaching
efﬁcacy sources.

Four sources of coaching efﬁcacy were proposed in the initial study by Feltz, et
al. (1999) and other studies have added to these sources (F eltz, Hayashi, & Hepler,
(2005); Lirgg, DiBrezzo, & Smith, 1994; Lee, Malete, & Feltz, 2002; Myers, Vargas-
Tonsing, & Feltz, 2005; Sullivan, Gee, & F eltz, in press). Myers, Vargas-Tonsing, and
F eltz (2005) examined the pr0posed sources of efﬁcacy information on dimensions of
coaching efﬁcacy among college coaches. Their study was conducted as an extension of
the coaching efﬁcacy literature because none of the prior research provided a
comprehensive test for the proposed model in non-high school settings.

Participants were 179 head coaches from Division II and ID universities from
several mid-western states. The coaches were sent packets at two different points in the
season. At Time 1 a packet containing a questionnaire of demographic items, measures of
proposed sources, and the CES were sent to the coaches. At Time 2, at the three-quarter
mark of the season the coaches that responded after Time 1 (75%) were sent the packets

once again. The packets contained the same items with the addition of a questionnaire

21

looking at their perceptions of their efﬁcacy-enhancing behaviors with athletes, athlete
satisfaction with their head coach, and winning percentage for the current season.

Findings from the study support previous research as a signiﬁcant relationship
between the sources and dimensions of coaching efﬁcacy was found. Speciﬁcally,
important sources of coaching efﬁcacy were found to be perceived team ability, social
support from the athletes’ parents and the community, career winning percentage, and
experience as a collegiate head coach.

Feltz, Hayashi, and Hepler (2005) looked to expand on previous research by
interviewing 12 coaches from the Feltz et al., (1999) study. The purpose of the study was
to identify additional sources of coaching efﬁcacy that these coaches use.

Participants consisted of 12 coaches from phase two of the Feltz et al. (1999)
study. A set of questions was asked of the coaches with appropriate probing to get in-
depth analysis. Results found player improvement and development, support from their
players, and past playing experiences contribute to sources of coaching efﬁcacy as well
what F eltz et al. proposed.

Coaching behavior, in turn, is affected by coaching efﬁcacy (Feltz, Chase, Moritz,
& Sullivan, 1999; Fung, 2003; Myers, Vargas-Tonsing, & Feltz, 2005; Sullivan & Kent,
2003). Fung (2003) found that coaches with low game strategy efﬁcacy would be hesitant
to apply their knowledge and Myers, Vargas-Tonsing, and Feltz, (2005) found coaches
exhibited higher perceptions in these areas had a higher winning percentage, and scored
higher in their Coaching Efﬁcacy Scale scores. Coaches with a higher sense of coaching
efﬁcacy gave more praise and encouragement than coaches with a lower sense of

coaching efﬁcacy (Feltz et al, 1999; Sullivan & Kent, 2003). This behavior includes

22

positive reinforcement to a desirable performance, positive contingent reinforcement plus
technical instruction, spontaneous encouragement that does not follow a mistake, and
mistake contingent encouragement.

Sullivan and Kent (2003) sought to look at the link between coaching efﬁcacy and
its behavioral outcomes. This study differed from Feltz et al. (1999) as Sullivan and Kent
focused on perceptions of leadership style of coaches measured through the Leadership
Scale for Sport (LSS) allowing for a greater in—depth examination of the relationships
between efﬁcacy and leadership. Along with this the authors also looked to examine the
generalize-ability of coaching efﬁcacy.

The sample comprised of intercollegiate coaches ﬁom America and Canada.
Surveys were mailed to 570 coaches (300 American coaches and 270 Canadian coaches).
The instruments included in the packets were the Coaching Efﬁcacy Scale (F eltz et al.,
1999) and the LSS. The response rate was 165 male and 58 female coaches (39.3%).

Results showed the collegiate coaches scores on the CES dimensions correlated
with leadership styles. Motivation and teaching efﬁcacy predicted training, instruction,
and positive feedback use by the coaches; the higher their coaching efﬁcacy the more
likely coaches would use appropriate training and instruction. Winning percentage was
higher for coaches with high efﬁcacy. Coaches with lower coaching efﬁcacy ratings
provided more instructional and organizational behaviors and had a lower winning
percentage.

Myers, Vargas-Tonsing, and Feltz (2005) found total coaching efﬁcacy to
inﬂuence coaches’ efﬁcacy-enhancing behaviors (increasing conﬁdence as in verbal

persuasion) with athletes. Winning percentage and team satisfaction were also found to

23

increase with high total coaching efﬁcacy. In regards to women’s teams total coaching
efﬁcacy only predicted coaching behavior. These behaviors may have an effect on the
athletes.

Coaching efﬁcacy is shown to have an effect on athletes (F eltz, Chase, Moritz, &
Sullivan, 1999; Myers, Vargas-Tonsing, & Feltz, 2005; Short & Short, 2004; Vargas-
Tonsing, Wamers, & Feltz, 2003). Player satisfaction was found to be correlated with
high coaching efﬁcacy possibly due to coaches with higher efﬁcacy exhibiting coaching
styles more preferred by athletes (Feltz et al., 1999).

Vargas-Tonsing, Wamers, and Feltz (2003) said efﬁcacy was found to be a
signiﬁcant predictor of team efﬁcacy and team satisfaction. Using F eltz et al. is (1999)
model the authors set out to study the efﬁcacy of the players both individually and as a
team as it correlates with the coaches’ own efﬁcacy; their hypothesis was coaching
efﬁcacy would predict team and player efﬁcacy.

Participants were 133 female varsity athletes and head coaches from 12 high
school volleyball teams from a mid-west state. The athletes were given a self-efﬁcacy
and team-efﬁcacy questionnaire while the coaches were given the Coaching Efﬁcacy
Scale (F eltz et al., 1999).

Results revealed a signiﬁcant relationship between coaching efﬁcacy to be a
predictor of team efﬁcacy but not of self-efﬁcacy. The authors concluded this may be
because in team sports it is easier to assess the team’s accomplishments as a whole than
to asses individual accomplishments. The results imply the need for coaches to be aware

of their efﬁcacy’s inﬂuence on their team as a whole.

24

The literature has supported coaching efﬁcacy to have an impact on young
athletes’ lives. An instrument such as the Coaching Efﬁcacy Survey may help in
assessing any deﬁciencies in coaches and lead to improvements in the coaching education
literature. To validate the current CBS 3 series of statistics will be implemented similar to

what was already done in previous validations.

Psychometric Properties of the CES

Conﬁrmatory factor analysis (CFA) of the 24—item CES with its four dimensions
demonstrated psychometric ﬁt indexes that were minimally acceptable but consistent in
samples of high school and small-college level coaches. Fit indices ranged from 87-89
for Comparative Fit Index (CF I), and 080-085 for Root Mean Square Residual Error of
Approximation (RMSEA) across four studies for the multidimensional model (Feltz et
al., 1999; Lee, Malete, & Feltz, 2002; Myers, Wolfe, & F eltz, 2005; Sullivan & Kent,
2003). F eltz et al. (1999) also tested the ﬁt of the unidimensional model. Slightly greater
misﬁt was observed for this model, ,1/ = 844, Nonnormed Fit Index (NNF I) = .87, CFI =
.88, and RMSEA = .09, as compared to the multidimensional model but is still useﬁrl in
instances when coaching efﬁcacy is but one of a host of variables used to predict a
general outcome (e.g., performance), and/or when subscale scores are highly related and
likely to cause problems associated with multicollinearity within the speciﬁed data
analysis (Myers et al., 2005).

Myers et al. (2005) also examined the degree to which high school and small-
college level coaches employed the original 10-category rating scale structure of the CES

(where 0 signiﬁed “not at all conﬁdent” and 9 signiﬁed “extremely conﬁdent”)

25

systematically. They reported that coaches did not employ the original rating scale
structure systematically because they were being asked to distinguish between too many
levels of coaching efﬁcacy, which was congruent with previous ﬁndings for the optimal
structure of an ordered response efﬁcacy scale (Zhu et al., 1997) and long-standing
recommendations for Likert scales (Likert, 1932). A post-hoc analysis suggested that a
4-category structure was optimal for the Myers et al. sample. Accordingly, they
suggested that subsequent users of the CES should use a condensed rating scale structure.

The dimensions of coaching efﬁcacy may vary, however, depending on the age
and level of the athletes or level of the coach. The factor structure of the measure may not
be appropriate for youth sport coaches who are concentrating more on the fundamentals
of the sport. Youth-sport coaches have an impact on athletes during the developmental
stages of sports participation. Their inﬂuence includes teaching the rules and strategies of
the sport, the fundamental skills needed to perform, and even developing the character to
handle social-emotional situations in sport (i.e., sportsmanship). In addition, many youth
sport organizations downplay the emphasis of outcome goals such as winning. Thus, the
factor structure of the CES should be examined through CFA before hypotheses within
the coaching efﬁcacy model are tested with volunteer youth-sport coaches.

In addition, youth-sport coaches may rely on different sources of efﬁcacy
information on which to base their conﬁdence in coaching than high school or small
college coaches. If outcome goals are downplayed in youth sports, a coach’s perceptions
of the team’s ability may not be as important as perceptions of how the team has
improved from one season to the next. Past winning record and school/community

support may be of no importance, but parental and player support may be very important.

26

Thus, the most salient sources of efﬁcacy information for youth-sport coaches should
also be examined to help expand the model and aid in helping coaches enhance their

coaching efﬁcacy.

27

CHAPTER 3
METHOD

Participants

The research subjects comprised 394 volunteer youth sport coaches ﬁom the mid-
west region of the United States. Youth sport coaches were deﬁned to be a person who
coaches children between the ages of 7 and 12 years. The sample included both head
coach and assistant coaches. The sample was 94% male, 6% Caucasian, and 52% had a
minimum of a bachelor’s degree. The age of coaches ranged from 17-72 years (M = 38.8,
SD = 8.8) and years of coaching experience from 0-30 years (M = 3.59, SD = 4.11).
Coaches represented a number of sports including ice hockey (73%), basketball (18%),

soccer (4%), volleyball (3%), football (1%), softball (.3%), and other (.3%).

Measures

Demographic measures. Demographics were collected including the coaches’
gender, age, ethnicity, educational background, sport background, primary sport coached,
years of coaching primary sport, years playing in primary sport, hours spent coaching
primary sport, and volunteer or paid status. Coaches were also asked about the age and
gender of the team they are coaching.

CES. As explained in Chapter 1, the CES measures total coaching efﬁcacy, as
well as subsets within the instrument including: teaching technique efﬁcacy (TE), game
strategy efﬁcacy (GSE), motivation efﬁcacy (ME), and character building efﬁcacy
(CBE). As recommended by Myers et al. (2005), a condensed rating scale structure was

employed in this study where “1” was no conﬁdence, “2” low conﬁdence, “3” moderate

28

conﬁdence, “4” high conﬁdence, and “5” complete conﬁdence. Unlike the Myers et al.
study, which advocated for a 4-category structure with high school and collegiate
coaches, in this study a 5-category structure was deemed appropriate because it included
a category for “no conﬁdence”, which was reasonable in the given sample, because many
of the participating coaches (8%) had no prior experience coaching. Additonally (27%)
had no experience playing the sport coached.

Procedure

Permission was obtained from the Institutional Review Board. Coaches were
recruited from coaching clinics, tournaments and league meetings. The ﬁrst group of data
was collected at a youth ice hockey coaching clinic in the northeast. Surveys were handed
out to the coaches before their workshops and 496 coaches responded. Coaches who
responded coached athletes whose ages ranged from 6 to 17 years. Participant responses
were discarded if the criteria were not met leaving 387 responses. To get the number of
coaches down to a number that would be statistically acceptable to add with other data
and not thoroughly dominate the analysis, every fourth participant was deleted ﬁ'om the
data. The ﬁnal number of ice hockey responses were 289.

Responses were also collected from youth sport coaches in other sports in a mid-
westem state. Numerous phone calls and emails were sent to various leagues throughout
the State inquiring their willingness to allow contact with their coaches for the study.
Those who responded were contacted for speciﬁc information as to where and when the
events occurred. At various basketball tournaments 69 coaches responded to the survey.
At a volleyball tournament 13 youth sport coaches responded to the survey. Soccer

coaches from a local recreation league and club league were approached and all (16)

29

responded to the survey. A local youth football league agreed to help and submitted a list
of coaches available to contact for the study. Numerous phone calls and emails were
made and sent resulting in ﬁve agreeing to participate in the study. The surveys were
administered in their respective homes and completed promptly. A volleyball and
cheerleading coach were recruited at a coach’s education meeting held in a mid-west
rural town. In the end, 105 youth sport coaches in sports other than ice hockey
participated in the study. A total of 394 responses were gathered from the participants.

An explanation of the study was presented and informed consent was obtained
from all coaches. An explanation of the study was provided to the coaches. They were
told the following survey was to assess coaching conﬁdence. All youth sport coaches
approached agreed to participate in the study. The investigator was not present as the
participants ﬁlled out the surveys.

To minimize the threat of data loss ﬁ'om unanswered questions, the investigator
examined completed questionnaires to assure all questions were answered and if any
questions were skipped the investigator brought this to the attention of the respondent and
the overlooked question was then answered. For coaches hesitant to take the survey, the
administrator assured the participants that the process would take no more than 7

minutes.

Treatment of Data
In order to answer the ﬁrst question, “Can the factor structure of the CES be
replicated with a sample of volunteer youth sport coaches?” a CFA utilizing maximum

likelihood procedures with the covariance matrix as input was performed using AMOS

3O

(Arbuckle, 1996). As proposed by F eltz et a1. (1999), hierarchical-factor model
representing four ﬁrst-order factors and a second-order general factor was tested with all
of the factors speciﬁed as correlated with one another. The ﬁt of the model to the data
was evaluated using the chi-square statistic (12 ), the NNFI, the CFI, and the RMSEA.
These indices were chosen because they are appropriate for CFAs with maximum-
likelihood procedures (Tabachnick & Fidel], 2001) and have been used previously with
the CES.

The second question, “What is the strongest dimension of coaching efﬁcacy
among volunteer male and female youth-sport coaches?” could not be addressed because
of the low number of female respondents. The third question, “What are the strongest
predictors of coaching efﬁcacy among the sources measured?” was examined using a
multivariate multiple regression (MMR) analysis with eight sources of coaching efﬁcacy
(perceived team ability, perceived team improvement, social support, years as a volunteer
coach, and years of experience as an athlete as predictor variables and the four
dimensions of coaching efﬁcacy as dependent variables. In the case of a signiﬁcant
overall multivariate effect, follow-up univariate multiple regression analyses were
conducted on each efﬁcacy dimension.

A t-test was conducted between ice hockey coaches and all other sport coaches.
Due to the high number of ice hockey coaches (n=289) compared to all other sports
coaches (n=105) there may be a signiﬁcant difference between the two groups and an

independent sample t-test would best assess this.

31

CHAPTER 4
RESULTS

Conﬁrmatory Factor Analysis

To test the factorial validity of the CES (the ﬁrst research question of this study) a
CF A was performed utilizing maximum likelihood procedures with the covariance matrix
as input using AMOS (Arbuckle, 1996). As proposed by Feltz et a1. (1999), a hierarchical
factor model representing four ﬁrst-order factors and a second-order general factor
representing coaching efﬁcacy, was tested with all the factors speciﬁed as correlated with
one another. In terms of the global indices of ﬁt, the chi-square was large, 12 (248) =
929.22, p < .001 and lZ/df= 929.22/248 = 3.75. The NNFI = 0. 86 CFI = 0. 89 RMSEA=
0. 084. This model did not ﬁt the data as well as the ﬁrst-order model representing the
just the ﬁrst-order factors, 12 (246) = 809.22 , p < .001, lZ/df= 3.29, NNFI = 0.89, CFI
= 0.91, RMSEA = 0.076. Although the indices were considered only marginally
acceptable according to Hu and Benler (1999) and Browne and Cudeck (1993), they were
in line with what others have found (Feltz et al., 1999; Lee et al., 2002; Myers et al.,
2005; Sullivan & Kent, 2003). The means, standard deviations, and frequency
distribution for each item are presented in Tablel and means and standard deviations for

the CES subscales are presented in Table 2.

Sources of Coaching Eﬂicacy
The sources of coaching efﬁcacy included coaching experience, athlete ability,
improvement of athletes over time, years of playing experience, perceived support from

the organization, athletes, and community support. Pearson correlations between the

32

sources and the CES subscales are contained in Table3. To determine the predictive
strength of the sources on the subscales of coaching efﬁcacy, the sources and the CES
subscales were entered into a MMR analysis. Results indicated that the overall
multivariate test was signiﬁcant Wilk’s; A = 0.46, F (4, 329) = 96.48, p < 0.001. The
univariate regression analyses showed that all four equations were signiﬁcant (p < .001),
for motivation, strategy, technique, and character building [F s (8, 332) = 10.34, 17.46,
20.71 , 5.25, respectively]. A summary of the MMR is presented in Table 4 along with
the standardized beta coefﬁcients for the predictor variables. As can be seen from the
table, perceived ability, parent support, and community support did not predict any of the
dimensions, and organizational support was a weak predictor of technique efﬁcacy only.
In contrast, perceived team improvement and athlete support were signiﬁcant predictors
of motivation, strategy, and technique efﬁcacy. Years of playing experience was a
signiﬁcant predictor of strategy and technique efﬁcacy, but not motivation and character
building. The only signiﬁcant predictor of character building efﬁcacy was athlete
support.

In addition to the regression procedures, the relationship between the two
perceptual sets (i.e. sources and dimensions of coaching efﬁcacy) was examined via
canonical correlation analysis (see Table 5). Canonical correlation was used because the
coaching efﬁcacy model speciﬁes that a set of sources are related to a set of coaching
efﬁcacies (ﬁrst-order dimensions). Two signiﬁcant canonical relationships were observed
between the two perceptual sets (Rc, = .59, Ref = .35; RC; = .37, Ref = .13): However,
the redundancy index that measured the amount of variance predicted in the coaching

efﬁcacy dimensions from the set of predictors was 19.52% for the ﬁrst canonical

33

relationship and only 3.24% for the second canonical relationship. Therefore, only the
ﬁrst canonical relationship was interpreted.

Canonical loadings were examined to determine the variables that most
contributed to the multivariate relationship between the two sets. Inspection of the
canonical loadings in Table 5 shows that for the sources of coaching efﬁcacy, all of them
were important contributors to the relationship (i.e., loadings were greater than .30) with
playing and coaching experience contributing most to the relationship. The loadings for
the coaching efﬁcacy dimensions also indicate that all dimensions were important
contributors to the canonical set and that technique and strategy dimensions were the
most important. Thus coaches who perceived higher ability and improvement in their
athletes, who perceived greater support, and who had more experience in playing and
coaching were more conﬁdent in their coaching, especially in terms of the technique and
strategy aspects.

T -test results revealed a signiﬁcant difference between ice hockey coaches and all
other sports coaches on most efﬁcacy sources (Table 8) and all efﬁcacy dimensions
(Table 9). Efﬁcacy sources of parental support t(388)= -1.62, p < .05, community
support, t(384) = -5.23 p < .05, organization support, t(387) = -1 .79, p < .05, and athlete
support (t(386) = -1.32, p < .05) were found to be signiﬁcantly different. For efﬁcacy
dimensions, signiﬁcance was found in all four dimensions; motivation efﬁcacy (t(382) =
-5.83, p < .05), game strategy efﬁcacy (t(384) = -7.40, p < .05), character building
efﬁcacy (t(389) = -5.12, p < .05), and teaching technique (t(388) = -7.88, p < .05).

Results for both efﬁcacy sources and efﬁcacy dimensions ice hockey coaches had lower

34

results than all other coaches. These results indicate the ﬁndings in this study may only

generalize to ice hockey coaches.

35

CHAPTER 5
DISCUSSION

The purpose of this study was to assess validity of the CBS for use with youth
sport coaches and to examine the inﬂuence of proposed sources of coaching efﬁcacy on
dimensions of coaching efﬁcacy beliefs. Considering the results as a whole, there is
evidence to support the use of the CES as a multidimensional scale to assess the coaching
efﬁcacy of youth sport coaches. The ﬁt indices, though considered marginal by some
psychometricians (Hu & Bentler, 1999), are slightly better than what previous studies
have found and are consistent with those studies (F eltz et al., 1999; Lee et al., 2002;
Myers et al., 2005; Sullivan & Kent, 2003). Furthermore, Marsh, Hau, and Wen (2004)
have argued that Hu and Bentler’s proposed cut-off values for testing goodness of ﬁt
indices were intended to be guidelines rather than ‘golden rules’ such as those used in
traditional hypothesis testing. Given the consistency of ﬁt indices across studies, different
sample characteristics, and sample sizes, and given the predictive validity of the measure,
continued use of the CES is reasonable with youth sport coaches as well as high school
and Division II and HI collegiate coaches.

In addition, Myers et al. (2005) recommended the use of a condensed scale of the
CES because their analysis of the original rating scale structure indicated that coaches
were being asked to distinguish between too many levels of coaching efﬁcacy. The
present study was the ﬁrst to test the factor structure of the CES using a condensed, 5-
category scale (i.e., no, low, moderate, high, and complete conﬁdence). Based on Linacre
(2002), Myers et al. (2005) indicated that a category (e.g., no or low) should attract at

least 10 observations at the item level for testing a model’s ﬁt against the data. Results

36

showed, however, that very few coaches reported having no conﬁdence on any of the
items. In speculating, once coaches have made the decision to take a coaching position,
even a volunteer one, they perceive that they have at least some conﬁdence to do the job.
As Feltz et al. noted, individuals would probably not enter or remain in coaching if they
had little or no conﬁdence in their coaching ability. Thus, a 4-category scale that excludes
a “no conﬁdence” category may be more appropriate for future research using the CES
with youth sport coaches.

In addition to few responses in no conﬁdence category, the mean values for each
of the four dimensions were around the ‘high conﬁdent’ category. These values are
consistent with previous research (F eltz et al., 1999). The high conﬁdence responses do
not mean that these coaches were necessarily highly qualiﬁed to coach, just that they

were conﬁdent.

Sources of Coaching Eﬂicacy

In terms of the most salient sources of efﬁcacy information for youth sport
coaches, the canonical correlation analysis revealed that coaching and playing experience
were most strongly associated with technique and strategy efﬁcacy beliefs. In terms of
sources of coaching efﬁcacy, the results support previous ﬁndings regarding coaching
experience as a strong and consistent predictor of coaching efﬁcacy across all levels of
coaching (Feltz et al., 1999; Marback, Short, Short, & Sullivan, 2004; Myers et al., 2005;
Sullivan et al., in press).

Playing experience also was a strong predictor of coaching efﬁcacy. Prior to

Sullivan et al. (in press), studies had not included playing experience as a possible source

37

of coaching efﬁcacy. Sullivan et al. found that playing experience (controlling for
coaching experience) was a signiﬁcant predictor of game strategy efﬁcacy but not other
efﬁcacy dimensions. For youth sport coaches, who, on average, have relatively less
coaching experience than high school and college coaches, playing experience may
provide the sport-speciﬁc knowledge of the skills, rules, vocabulary, and strategy of how
the game is played as a basis for coaching efﬁcacy beliefs. Unfortunately, playing
experience may also be masked as expertise in coaching and coaches may rely on
“accepted knowledge” of coaching from their own experience rather than knowledge
based on pedagogy (Rushall, 2004). As stated previously, the youth sport coaches in this
study considered themselves, on average, to be highly conﬁdent on all four dimensions
even though 38% had less than 2 years of coaching experience. It is possible that as
coaches gain more coaching experience, playing experience becomes less important.
Future longitudinal research is needed to help tease out the strength of these predictors.
Future research also should examine the relationship between coaching efﬁcacy and
actual coaching competence to determine if coaches become more accurate in their self-
assessments as they gain coaching experience.

Although the results support previous ﬁndings (e.g., Feltz et al., 1999; Myers et
al., 2005) in that all predictors and efﬁcacy dimensions were strong contributors to the
relationship, perceived team improvement and athlete support were unique signiﬁcant
predictors of coaching efﬁcacy among youth sport coaches. Myer et al. found that, with
collegiate coaches, the strongest predictor of coaching efﬁcacy was perceived team
ability. As Myer et al. noted, perhaps the major inﬂuence of perceived team ability has to

do with the more competitive nature of collegiate sports compared to youth sports. In

38

youth sports, seeing one’s athletes improve in their playing ability appears to provide
success information that the coach is doing a good job of teaching the skills and strategies
of the game.

As for athlete support, it was the strongest form of social support in predicting
coaching efﬁcacy, especially motivation and character building efﬁcacy. Surprisingly,
support from parents was not as strong in informing coaches’ efﬁcacy beliefs. Perhaps
the support from athletes is more visible, immediate, and positive with athletes at these
ages. Young athletes are more apt to display their affections toward their teammates and
coaches (i.e. hugging, laughing, etc). When they display positive actions coaches might
infer their athletes are learning positive character building skills. If an athlete displays
negative actions (i.e. yelling, hitting, etc.) the coach might then infer the athlete is not
developing character building skills. Additionally, because the emphasis is more on fun
and skill development and less on winning and elite performance, parents may not be as
vocal about the coach’s ability as they might be with high school, elite, or college-level
play.

Furthermore, in terms of salient dimensions of coaching efﬁcacy, it makes sense
that volunteer youth sport coaches would ﬁnd technique and game strategy efﬁcacy to be
more inﬂuenced by efﬁcacy sources because these dimensions represent the basics of the
sport. Motivation and character building are efﬁcacy dimensions that may become more
salient with more experience and higher level coaching. As Feltz, Short, and Sullivan (in
press) noted, the relationships between the sources of coaching efﬁcacy and the
dimensions may vary by coaching level. The relationship between years of

coaching/playing and Technique and Game strategy may be more important at the youth

39

level; years in coaching and Game strategy and Motivation efﬁcacy may be more
important at the high school level; and, the association between perceived team ability
and Motivation and Character building efﬁcacy may be more important at the college
level. A systematic study of these relationships by coaching level within the same sports
awaits further research.

Although the results suggest that youth sport coaches rely most strongly their own
playing and coaching experience for their conﬁdence, the best methods to improve their
conﬁdence and competence is through coaching education (Woodman, 1993). The
empirical literature on coaching education as a source of coaching efﬁcacy is small, but
consistent. Coaches who have participated in a coaching education program are more
conﬁdent in all aspects of coaching efﬁcacy compared to either a control group of
coaches who have not taken a course and/or their pre-course conﬁdence levels (Campbell
& Sullivan, 2005; Lee et al., 2002; Malete & Feltz, 2000). Coaching education programs
may be especially important for improving motivation and character building competence
and conﬁdence because coaches may not get this information as readily from playing and
coaching experience alone. However, the coaching education backgrounds of the coaches
in this study were not assessed.

Intuitively it is important for ﬁmdamental skills of the sport to be taught to young
athletes but neglecting skills that build character in young athletes is a detriment to their
overall development. Sport can be miss-characterized as an activity that innately teaches
character to all participants. Sport should be looked at as a tool that, when used properly,
can be potent and effective to teach basic character principles. Unfortunately beginning

level coaches may not take into consideration the need for conscious modiﬁcation of their

40

coaching to use sport as a tool rather than assuming that the sport in and by itself will
teach the young athletes character. Coaching education can bring awareness to coaches of
this problem and assist them in becoming better coaches in this area.

Limitations

A few limitations were present in this study. Due to convenience sampling there
was an over sampling of ice hockey coaches. About 73% of the participants in this study
were coaches recruited at a USA Ice Hockey coach’s clinic. This was a concern about
oversampling because USA Ice Hockey typically has more education requirements of its
coaches than other youth sports programs. The over representation of hockey coaches
may have inﬂated the efﬁcacy results. However, although there were signiﬁcant
differences between the hockey and all other coaches on most of the descriptives and
CBS dimensions, the hockey coaches had signiﬁcantly less playing experience, had
signiﬁcantly less social support from parent, community, the organization, and athletes,
and had signiﬁcantly lower ratings of efﬁcacy on CBS dimensions than did all other sport
coaches.

Also, because ice hockey traditionally attracts male coaches very few respondents
were females. One of the questions this study sought to answer was what dimensions of
the CES were strongest amongst male and female coaches? This question was not
answered due to the underwhelming number of female participants.

Another limitation was the non-inquiry of coaching education background of the
participants. Many of these coaches may have gone through coaching education courses

giving them a solid foundation of awareness and knowledge of coaching issues.

41

Compared to others without similar coaching education backgrounds these coaches

would be predicted to have higher coaching efﬁcacy.

Future Directions

Although this thesis examined the antecedents of the F eltz et al. (1999) model
with youth sport coaches, it did not investigated the consequences of the coaching
efﬁcacy dimensions. F eltz et al. hypothesized that coaches with higher efﬁcacy scores
would use more positive coaching, have more players satisﬁed with their playing
experience, have higher winning percentages, and higher efﬁcacy levels among athletes
and teams. Most of these outcomes would be appropriate to study at the youth level as
well as high school and college levels. However, given that winning is not usually the
most emphasized aspect of successful coaching at the youth level, future research should
examine additional outcomes of coaching efﬁcacy that are more appropriate to youth
sports. For instance, do young athletes of high efﬁcacy coaches improve more in their
skills over the course of a season than athletes who play for low efﬁcacy coaches? Do
young athletes of high efﬁcacy coaches self-report having more fun than those who play
for low efﬁcacy coaches?

Secondly, because sportsmanship and fair play are usually at the forefront of
purposes of youth sports, future research should examine the explicit relationships
between sources and consequences of CBE. For instance, coaches’ character-building
efﬁcacy should predict their emphasis on sportsmanship in their instructions to athletes

and their role modeling with referees, which should, in turn, predict their athletes’

42

sportsmanship attitudes and behavior. Youth sport coaches also might be interviewed to

explore what they perceive as the most potent sources of CBE.

43

APPENDICES

44

APPENDIX A

Coaches Background Information Questionnaire
Please check only one answer per question-
1. Your sex: ____(1) Male, ___(2) Female
2. Primary sex of your team: __(1) Male , _(2) Female _(3) co-ed

3. Level/Age group of team coaching. (e.g. U-19, Pee-wee):

 

4. Your ethnic afﬁliation: __ (l) Caucasian,
_ (2) African American
__ (3) Native North American Indian
_ (4) Asian American
_ (5) Hispanic
_ (6) Other

 

5.Your Age:

6. Educational Background: (check highest level competed)

(1) Did not complete High School _(2) High School graduate

(3) Less than 2 years college/tech _(4) 2 or more years college
(5) Bachelor’s degree _(6) Some Master Level work
(7) Master’s degree __(8) Some Doctoral level work

(9) Completed Doctorate

7. Ifattended college, what was undergraduate major? , Master’s
major

 

 

Doctorate degree?

 

8. Check the primary sport you coach. (Please check only one)

(1) Basketball _(2) Hockey_(3) Football

(4) Tennis _(5) Baseball _(6) Softball
(7 Swimming __(8) Golf _(9) Volleyball
(10)Track and Field _(11) Wrestling __(12) Soccer
(13) Cross Country _(14) Other

9. Total numbers of years coaching this sport

 

10. Present position. (1) Head, (2) Assistant

11. Number of years of playing experience in this sport at each level
( 1) Youth (2)High School
(3)College (4) Professional

45

12. Approximately how many hours per week do you spend involved in fulﬁlling your coaching

 

 

duties, planning etc. ? In season Out of season
13. Are you paid for your present coaching services? (1)NO (2) Yes
Poor Excellent
14. How would you rate the overall ability of the athletes
on your team this year? 0 1 2 3 4 5 6 7 8 9

15. How would you rate the overall improvement of your athletes over the course of the season?
0 1 2 3 4 5 6 7 8 9

16. In comparison with your perception of the ideal youth sports program, how would you rate

the support given to you by the parents of your athletes?
0 1 2 3 4 5 6 7 8 9

17. In comparison with your perception of the ideal youth sports program How would you rate
the community support for your team? 0 1 2 3 4 5 6 7 8 9

18. In comparison with your perception of the ideal youth sports program How would you rate

the support given to you by the organization that runs your program?
0 l 2 3 4 5 6 7 8 9

19. In comparison with your perception of the ideal youth sports program, how would you rate

the support given to you by your athletes?
0 1 2 3 4 5 6 7 8 9

46

I"

PWNQMPP’

ll.
12.
13.

14.
15.
l6.
17.

18.
19.
20.
21.

22.

APPENDIX B

Coaching Confidence Questionnaire

Coaching conﬁdence refers to the extent to which coaches believe that they have the capacity to
affect the learning and performance of their athletes. Think about how conﬁdent you are as a
coach. Rate your conﬁdence for each of the items below. Your answers will be kept completely

conﬁdential.

How conﬁdent are you in your ability to--

help athletes maintain conﬁdence in
themselves?

recognize opposing team’s strengths during
competition?

mentally prepare athletes for game strategies?

understand competitive strategies?

instill an attitude of good moral character?
build the self-esteem of your athletes?
demonstrate the skills of your sport?
adapt to different game/meet situations?

recognize opposing team’s weakness during
competition?
motivate your athletes?

make critical decisions during competition?
build team cohesion?

instill an attitude of fair play among your
athletes?
coach individual athletes on technique?

build the self-conﬁdence of your athletes?
develop athletes’ abilities?

maximize your team’s strengths during
competition?
recognize talent in athletes?

promote good sportsmanship?
detect skill errors?

adjust your game strategy to ﬁt your team’s
talent?
teach the skills of your sport?

No

Conﬁdence

47

1

Low

NNNN NNNN NNNNNNN

NNNN

Moderate High

wwwwwwwwwwwwwmw

umber»

«bk-bli-h-h-bh-b-h-hbbbhb

h-B-b-b

Complete

MMMM MMMMMMMMMMM

MMMM

23. build team conﬁdence?

24. instill an attitude of respect for others?

48

APPENDIX C
INFORMED CONSENT

Coaching-Efficacy Study Coaches Consent Form

You are being asked to participate in a study conducted by graduate student Nathan Roman under
the supervision of Deborah Feltz, Ph.D., from Michigan State University, title “EXTENDIN G
VALIDITY EVIDENCE FOR THE COACHING EFFICACY SCALE WITH VOLUNTEER
YOUTH SPORT COACEHS.” The purpose of this study are to examine structural validity of the
CES (F eltz et al., 1999) on youth sport coaches using a condensed rating scale structure. It is
believed that the project will have practical applications within coaching education.

As part of this research, you will be asked to complete a questionnaire about your coaching
conﬁdence. The questionnaires (approximately 10 minutes to complete) will contain items
speciﬁc to coaching conﬁdence and behaviors.

Your participation in this study will remain private, conﬁdential, and anonymous, no one except
the principal investigators will have access to these responses or to participation records. After
participating, you will no be able to be identiﬁed. At the end of the project, responses will be
presented at the group level to ensure the conﬁdentiality of individual responses, and coaches will
not be identiﬁed. Group-based ﬁndings will be made available to those who are interested. Your
privacy will be protected to the maximum extent allowable by law.

Your participation in this study would be greatly appreciated. However, please know that you
may refuse to participate or withdraw from the project at any time and without penalty. You may
also refuse to answer any speciﬁc question. If you would like to participate, please sign this form
and return it to the investigators.

If you have any question concerning this study, please contact Dr. Deborah Feltz, at
517.355.4732 [dfeltz@msu.edu] or Nathan Roman at 517.432.7121 [romannat@msu.edu].
Additionally, if you have any questions or concerns regarding your rights as a study participant,
or are dissatisﬁed at any time with any aspects of this study, you may contact-anonymously, if
you wish- Peter Vasilenko, Ph.D., Chair of the University Committee on Research Involving
Human Subjects (UCRIHS) by phone: 517.355.2180, fax: 517.432.4503, email address:
ucrihs@msu.edu, or regular mail: 202 Olds Hall, East Lansing,.MI 48824.

Thank you for your time and cooperation,

 

 

 

Dr. Deborah Feltz, Principal Investigator Date
Nathan Roman, Graduate Student Date
I , have been informed of and voluntary agree to participate in the above-

 

mentioned study.

 

 

Signature Date

49

APPENDIX D
IRB APPROVAL

50

APPENDIX E
TABLES

51

Table 1
Means, standard deviations and frequency of response for each item.

 

Frequency per Category of Conﬁdence Response

Item M SD No Zero Low Moderate High Complete
Response

 

ME 1 3.99 0.62
MB 3 3.53 0.77
MB 6 4.28 0.62
MEIO 4.03 0.71
ME12 3.89 0.76
ME15 4.08 0.66

1 94 308 89
25 232 180 51
1 47 274 169
10 99 271 111
11 141 239 98
5 75 292 116

0 0

1 3

1 O

l 0

2 1

4 O
ME23 4.09 0.71 2 1 4 85 267 133
G82 3.67 0.84 3 2 29 171 212 75
GS4 3.65 0.85 2 l 38 172 200 79
GS8 3.56 0.84 3 l 37 212 174 65
GS9 3.68 0.85 2 2 28 174 200 86
GSll 3.61 0.81 l 2 33 181 213 62
GS 1 7 3.64 0.79 0 2 22 196 206 66
G821 3.59 0.83 1 2 35 194 193 67
TE7 3.79 0.98 0 1 42 156 151 142
TE14 3.79 0.87 0 l 31 140 214 106
TE16 3.79 0.74 l l 10 152 256 72
TE18 4.13 0.71 0 O 4 77 266 145
TE20 3.70 0.80 4 l 22 171 213 81
TE22 3.84 0.90 2 2 29 136 206 117
CBS 4.36 0.63 1 0 0 44 238 209
CB13 4.29 0.68 0 0 5 46 241 200
CB19 4.52 0.59 2 0 O 25 195 270
CB24 4.41 0.63 0 0 2 38 219 233

 

52

Table 2
Descriptive statistics for the proposed sources of coaching efﬁcacy, coaching efﬁcacy
dimensions, and team variables. (N = 394)

 

 

M SD Minimum Maximum
Total coaching efﬁcacy 3.92 0.53 1.92 5.00
Motivation efﬁcacy 3.98 0.54 2.00 5.00
Character building efﬁcacy 4.39 0.52 3.00 5.00
Game strategy efﬁcacy 3.63 0.68 1.14 5.00
Technique efﬁcacy 3.84 0.67 2.00 5.00
Support form community 6.34 1.83 1.00 9.00
Support from parents 6.67 1.68 2.00 9.00
Support from organization 6.75 1.83 1.00 9.00
Support ﬁom athelets 7.25 1.29 2.00 9.00
Perceived athlete ability 5.55 1.50 1.00 9.00
Perceived athlete improvement 6.69 1.34 2.00 9.00
Years coaching 3.60 4.11 0.00 30.00
Years playing experience 5.06 5.46 0.00 20.00

 

53

Table 3
Correlations of sources with total coaching efﬁcacy and dimensions of coaching efﬁcacy

 

TCE ME CBE GSE TE

Support from community .24 .23 .22 .17 .22
Support from parents .30 .30 .23 .23 .23
Support from organization .27 .25 .21 .21 .22
Support from athletes .39 .40 .32 .31 .32
Perceived athlete ability .25 .26 .13 .24 .13
Perceived team improvement .31 .31 .20 .25 .20
Years playing experience .32 .09 .10 .34 .10
Years coaching experience .33 .23 .12 .35 .12

 

54

Table 4

Summary of the MMR Analysis and Standardized Beta Coefﬁcients for Predictors CES

Efﬁcacy (N = 394)

 

Standardized Beta Coefﬁcients for

 

Coaching dimensions

 

 

Predictor Variable Motivation Strategy Technique Character
Perceived ability -.004 .07 -.02 -.09
Team improvement .16** .14* .19M .11
Parental support .09 -.003 .05 .03
Community support -.36 -.04 -.008 .06
Organizational support .07 .07 . 12* .05
Athlete support .22M .1 l .04 .21M
Years of coaching .15** .27** .19" .04
Years of playing .03 .30" .41 ** .06
Univariate Fs 10.34" 17.46“ "‘ 20.71* * 5.25 *"'
R2 .20 .30 .33 .11
Adjusted R2 .18 .28 .32 .09

 

*p < .05. **p < .01

55

Table 5
Summary of the Canonical Correlation Analysis Between Sources of Coaching Efﬁcacy
and Coaching Efﬁcacy Subscales

 

Standardized
Canonical
Variables Loadings
Sources
Community Support .31
Parental Support .39
Organizational Support .37
Team Support .47
Perceived Team Ability .31
Perceived Team Improvement .43
Years Coaching .58
Years Playing Sport .75
Percentage of Variance 22.32%
Redundancy 7.89%
Coaching efﬁcacy subscales
Motivation .56
Game strategy .90
Technique .97
Character Building .40
Percentage of Variance 55.27%
Redundancy 19.52%
Canonical correlation .59

 

56

Table 6

Descriptive Information Years of Coaching and Playing Experience

 

 

Yrs Coaching Yrs Playing
Sport 11 X SD Min Max X SD Min Max
Soccer 16 6.13 8.93 2 15 2.13 5.82 0 12
Volleyball 13 4.23 4.90 1 16 5.46 4.39 O 14
Football 5 5.67 3.27 1 8 2.00 1.79 0 6
Basketball 69 4.93 5.5 5 0 30 4.90 4.86 0 17
Ice Hockey 289 3.17 3.60 0 26 5.07 5.53 O 20
Softball 1 5.00 0 5 5 6.00 O 6 6
Other 1 1.00 O 1 1 6.00 0 6 6

 

57

Table 7
Descriptive Information for Hours of Preparation In-Season and Out-of-Season.

 

 

In-Season Preparation Out-of-Season Preparation
(Hours per Week) (Hours per Week)
Sport 11 X SD Min Max X SD Min Max
Soccer 16 14.50 9.13 5 30 9.29 9.19 0 28
Volleyball 13 11.92 6.30 4 20 1.15 3.00 0 10
Football 5 13.80 1.64 12 15 4.00 1.00 3 5
Basketball 69 6.99 8.40 0 50 1.99 1.99 0 30
Ice Hockey 289 6.45 10.89 0 40 0.84 2.20 4 20
Softball 1 15.00 0.00 15 15 10.00 0.00 10 10
Other 1 10.00 0.00 10 10 5 0.00 5 5

 

58

Table 8
Descriptive Comparisons Of Efﬁcacy Sources Between Ice Hockey Coaches and Other
Sport Coaches

 

 

N X SD t df p< .05

In-Season Prep

(Hours/Week)
Hockey 288 6.45 10.89 -2.24 387 *
Other Sports 101 9.11 8.34
Out-Season Prep

(Hours/Week)
Hockey 289 0.84 2.20 -5.41 389 *
Other Sports 102 3.09 6.06
Years Coached
Hockey 288 3.17 3.60 0.01 392
Other Sports 104 4.67 5.1 1
Years Played
Hockey 289 5.07 5.53 0.18 392
Other Sports 105 4.29 4.91
Perceived Athlete Ability
Hockey 288 5.51 1.45 -0.69 387
Other Sports 101 5.63 1.65
Perceived Athlete Improvement
Hockey 284 6.63 1.29 -1.40 382
Other Sports 100 6.85 1.44
Perceived Parental Support
Hockey 289 6.45 1.59 -1.62 388 *
Other Sports 101 7.31 1.62
Perceived Community Support
Hockey 285 6.06 1.74 -5.23 384 *
Other Sports 101 7.12 1.75
Perceived Organization Support
Hockey 288 6.50 1.76 -1.79 387 *
Other Sports 101 7.47 1.72
Perceived Athlete Support
Ice Hockey 287 7.08 1.25 -1.32 386 "‘
Other Sports 101 7.71 1.28

 

59

Table 9
Descriptive Comparisons of Efﬁcacy Dimensions Between Ice Hockey Coaches and
Other Sport Coaches

 

 

N X SD t (if p < .05
Motivation Efﬁcacy
Ice Hockey 280 3.89 0.52 -5.83 382 *
Other Sports 104 4.24 0.52
Game Strategy Efﬁcacy
Ice Hockey 286 3.49 0.64 -7.40 384 *
Other Sports 100 4.03 0.60

Character Building Efﬁcacy

Ice Hockey 286 4.32 0.51 -5.12 389 *
Other Sports 105 4.61 0.47

Teaching Technique Efﬁcacy

Ice Hockey 285 3.69 0.65 -7.88 388 *
Other Sports 105 4.25 0.54

 

60

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Box 7
Athltes’
’\
B 1 personal
ox characteritics
Socrocultural
Context
l
1
Box 8
BOX 4 Box 5 Athletes
9 , perceptions,
Coaches . Coaches interpretation, and
‘ expectancres, —’ behaviors /' evaluation of their
Box 2 , values, beliefs, ”3°11“, behavm’
Organizational and goals.
Climate ‘
A
1
Box 9
Athletes’ self-
perceptions,
r Box 6 beliefs, and
Box 3 Athletes’ attitudes
Coaches performance /
personal _ . and behavior
characteristics
V
Box 10
Athletes’ level
and type of
motivation

 

 

 

Figure 1. Hom’s Model of Coaching Effectiveness

61

REFERENCES
Arbuckle, J. L. (1996). Amos 4.0 [Computer Software]. Chicago: Smallwaters.

Ashton, P. T. (1984). Teacher efﬁcacy: A motivational paradigm of effective teacher
education. Journal of Teacher Education, 35 (5), 28-32.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Engelwood Cliffs, NJ: Prentice-Hall.

Bandura, A. (1997). Self-Efﬁcacy: The exercise of control. New York: W.H. Freeman
and Company.

Berman, P., McLaughlin, M., Bass, G., Pauly, E., & Zellman, G. (1977). Federal
programs supporting educational change, Vol. 11: Factors aﬂecting
implementation and continuation (Report No R-1589/7-HEW). Santa Monica,
CA: Rand Corporation.

Browne, M. W. & Cudeck, R. (1993). Altemative ways of assessing model ﬁt. In: Bollen,
K. A. & Long, J. S. (Eds.) Testing Structural Equation Models. pp. 136—162.
Beverly Hills, CA: Sage

Chase, M. A., Feltz, D. L., Hayashi, S. W., & Hepler, T. J. (2005). Sources of coaching
efﬁcacy: The coaches’ perspective. International Journal of Exercise and Sport
Psychology, 1, 7-25.

Coladarci, T. (1992). Teachers’ sense of efﬁcacy and commitment to teaching. Journal of
Experimental Education, 60(4), 323-337.

Connell, J. (1980). Multidimensional measure of children’s perceptions of control.
Denver: University of Denver.

Coopersmith, S. (1967). The antecedents of self-esteem. San Francisco: Freeman.

Corbin, C. B., Landers, D. M., Feltz, D. L., & Senior, K. (1983). Sex differences in
performance estimates: Female lack of conﬁdence vs. male boastfulness.
Research Quarterly for Exercise and Sport, 54, 407-410.

Coulson, R., & cobb, R. (1979), August). Assessment of a scale to measure generalized
expectancy of sport success. Paper presented at the annual meeting of the
American Alliance of Health, Physical Education, and Recreation, New Orleans,
LA.

62

Denham, C.H., & Michael, J .J . (1981). Teacher sense of efﬁcacy: A deﬁnition of the
construct and a model for further research. Educational Research Quarterly, 5,
39-63.

Denham, C. H., and Michael, J. J. (1981). Teacher sense of efﬁcacy: A deﬁnition of the
construct and a model for ﬁrrther research. Educational Research Quarterly, 6,
39-61.

Feltz, D. L., Chase, M. A., Moritz, S. E., & Sullivan, P.J. (1999). A conceptual model of
coaching efﬁcacy: Preliminary investigation and instrument development. Journal
of Educational Psychology, 91 , 765-776.

Feltz, D. L., & Ewing, M. E. (1987). Psychological characteristics of elite young athletes.
Medcine and Science in Sports and Exercise, 1 9, 98-105.

Feltz, D. L., & Petlichkoff, L. (1983). Perceived competence among interscholastic sport
participants and dropouts. Canadian Journal of Applied Sport Sciences, 8, 23 l -
235.

Fisher, C., Filby, N., Marliave, R., Cahen, L., Dishaw, M., Moore, J ., and Berliner, D.
(1978). Teaching behaviors: Academic Learning Time and student achievement:
Final report of Phase III-B, Beginning Teacher Evaluation Study. San Francisco:
F or West Laboratory for Educational Research and Development.

Fuller, B., Wood, K., Rappdrt, T., & Dorbusch, S. (1982). The organizational context of
individual efﬁcacy. Review of Educational Research, 52, 7-30.

Fung, L. F. (2003). Assessment: coaching efﬁcacy as indicators of coach education
program needs. Athletic Insight: The Online Journal of Sport Psychology, 5 (1).
Retrieved March 20, 2005, ﬁom
http://www.athleticinsight.comNolSlss1/Coa_rchingEfﬁcacy.htm#References

Gibson, 8., and Dembo, M. H. (1984). Teacher efﬁcacy: A construct validation. Journal
of Educational Psychology, 76 (4), 569-582.

Gould, D. (1982). Sport psychology in the 1980’s: Status,direction, and challenge in
youth sport research. Journal of Sport Pyschology, 4 (3), 203-218.

Gould, D., Hodge, K., Peterson, K., and Giannini, J. (1989). An exploratory examination
of strategies used by elite coaches to enhance self-efﬁcacy in athletes. Journal of
Sport & Exercise Psychology, 11, 128-140.

Gould, D., Hodge, K., Peterson, K., & Petlichkoff, L. (1987). Psychological foundations

of coaching: Similarities and differences among intercollegiate wrestling
coaches. The Sport Psychologist, 1, 293-308.

63

Guivemau, M., & Duda, J .L. (2002). The relationship of the moral atmosphere to

aggressive tendencies in young soccer players. Journal of Moral Education, 31
(I), 67-85.

Harter, S. (1982). The perceived competence scale for children. Child Development, 53,
87-97.

Horn, T. S. (1985). Coaches’ feedback and changes in children’s perceptions of their
physical competence. Journal of Educational Psychology, 77 (2), 174-186.

Horn, T. S. (2002). Coaching effectiveness in the sports domain. In T. S. Horn (Ed.),
Advances in sport psychology (2“. Ed.) (pp. 309-354). Champaign, IL: Human
Kinetics.

Lee, K. S., Malete, L., & Feltz, D. L. (2002). The strength of coaching efﬁcacy between
certiﬁed and noncertiﬁed Singapore coaches. International Journal of Applied
Sport Science, 14, 55-67.

Lenney, E. (1977). Women’s self-conﬁdence in achievement situations. Psychological
Bulletin, 84, 1-13.

Lesyk, J. J. and Komspan, A. S. (2000). Coaches’ expectations and beliefs regarding
beneﬁts of youth sport participation. Perceptual and Motor Skills, 90, 399-402.

Likert, R. (1932). A technique for the measurement of attitudes. Archives of Psychology,
140, 5-53.

Lirgg, C. D. (1991). Gender differences in Self-conﬁdence in physical activity: A meta-
analysis of recent studies. Journal of Sport and Exercise Psychology, 13, 294-310.

Lirgg, C. D., Dibrezzo, R., & Smith, A. N. (1994). Inﬂuence of gender of coach on
perceptions of basketball and coaching efﬁcacy and aspirations of high school
female basketball players. Women in Sport and Physical Activity Journal, 3, 1-14.

Litherland, M. S. (1996). Parents’ perceptions of participation in baseball and softball in
the developoment of boys and girls. Journal of Applied Sport Psychology, 8
(Suppl. ), 827.

Marback, T.L., Short, S. E., Short, M. & Sullivan, P.J. (2004). Coaching conﬁdence: An
examination of sources and gender differences. Journal of Sport Behavior, 28(1),
1 8-34.

Marsh, H. W., Hau, K—T, & Wen, Z. (2004). In search of golden rules: Comment on
hypothesis-testing approaches to setting cutoff values for ﬁt indexes and dangers
in overgeneralizing Hu and Bentler’s (1999) ﬁndings. Structural Equation
Modeling, 11 320-341.

64

Myers, N. D., Wolfe, E., & Feltz, D. L. (in press). An evaluation of the psychometric
properties of the Coaching Efﬁcacy Scale for American coaches. Measurement in
Physical Education and Exercise Science.

Myers, N. D., Vargas-Tonsing, T. M., & F eltz, D. L. (2005). Coaching efﬁcacy in
intercollegiate coaches: Sources, coaching behavior, and team variables.
Psychology of Sport and Exercise.

Park, J. K. (1992). Construction of the Coaching Conﬁdence Scale. Unpublished doctoral
dissertation. Michigan State University, East Lansing.

Martens, R. (1986). Youth sport in the USA. In M. R. Weiss, & D. Gould (Eds), Sport
for children and youths (pp. 27-33). Champaign, IL: Human Kinetics.

National Association for Sport and Physical Education (1995). National Standard for
Athletic Coaches. Retrieved November 23, 2004, from
http://www.aahperd.org/naspe/templatecﬁn?template=domainsStandards.htrnl

Ramey-Gassert, L., Shroyer, M. G., & Staver, J. R. (1996). A qualitative study of factors
inﬂuencing science teaching self-efﬁcacy of elementary school teachers. Science
Education, 80, 283-315.

Scanlan, T. K., & Lewthwaite, R. (1986). Social psychological aspects of competition for
male youth sport participants: IV. Predictors of enjoyment. Journal of Sport
Psychology, 8, 25-35.

Seefeldt, V., and Ewing, M. A. (1996). Youth sport in America: An overview.
President ’s Council on Physical Fitness and Sports Research Digest, Series 2
(1 1), 1-18.

Shields, D. L. L., & Bredemeier, B. J. L. (1995). Character development and physical
activity. Champaign, IL: Human Kinetics.

Shields, D. L. L., Bredemeier, B. J. L., Gardner, D. E., and Bostrom, A. (1995).
Leadership, cohesion, and team norms regarding cheating and aggression.
Sociology of Sport Journal, 12, 324-336.

Smith, R. E., Smoll, F. L., & Curtis, B. (1978). Coaching behaviors in little league
baseball. In F. L. smoll, & R. E. Smith (Eds), Psychological perspectives in
youth sports (pp. 173-201). Washington, DC: Hemisphere.

Smith, R. E., & Smoll, F. L. (1990). Self-esteem and children’s reactions to youth sport
coaching behaviors: A ﬁeld study of self-enhancement processes. Developmental
Psychology, 26, 987-993.

Smith, R. E. & Smoll, F. L. (1991). Behavioral research and intervention in youth sports.

65

Behavior Therapy, 22, 329-344.

Smith, R. E., and Smoll, F. L. (1997). Coach-mediated team building in youth sports.
Journal of Applied Sport Psychology, 9, 114-132.

Smoll, F. L., & Smith, R. E. (1993). Education youth sport coaches: An applied sport
psychology perspective. In J. M. Williams (Ed.), Applied sport psychology:
Personal grth to peak performance (2“d ed., 36-57). Mountain View, Ca:
Mayﬁeld.

Smoll, F. L., Smith, R. E., Barnett, N. P., &Everett, J. J. ( 1993). Journal of Applied
Psychology, 78 (4), 602-610.

Smoll, F. L., Smith, R. E., Curtis, B., & Hunt, E. (1978). Toward a mediational model of
coach-player relationships. Research Quarterly for Exercise and Sport, 49 (4),
528-541.

Sullivan, P. J ., Gee, C. J ., & Feltz, D. L. (in press). Playing experience: The content
knowledge source of coaching efﬁcacy beliefs. In F. Columbus (Ed.), Progress in
Educational Psychology Research. New York: Nova Science Publishers.

Sullivan, P. J ., & Kent, A. (2003). Coaching efﬁcacy as a predictor of leadership style in
intercollegiate athletics. Journal of Applied Sport Psychology, 15, 1-11.

Tabachnick, b. A., & Fidell, L. S. (2001). Using multivariate statistics. (4th ed.). Boston,
MA: Allyn and Bacon.

Vargas-Tonsing, T. M., Wamers, A. L., & Feltz, D. L. (2003). The predictability of
coaching efﬁcacy on team efﬁcacy and player efﬁcacy in volleyball. Journal of
Sport Behavior, 26, 396-406.

Weiss, M. R. (1987). Self-esteem and achievement in children's sport and physical
activity. In D. Gould & M. R. Weiss (Eds), Advances in pediatric sport sciences:
Vol. 2. Behavioral issues (pp. 87-119). Champaign, IL: Human Kinetics.

Weiss, M.R., & Bredemeirer, B]. (1990). Moral development in sport. Exercise and
Sport Science Reviews, 18, 331-378.

Wood, R., Bandura, A. & Bailey, T. (1990). Mechanisms governing organizational
performance in complex decision-making environments. Organizational Behavior

and Human Decision Processes, 46, 181-201.

Woodman, L. (1993). Coaching: A Science, an art, an emerging profession. Sport
Science Review, 2(2), 1-13.

Woolfolk, A. E., and Hoy, W. K. (1993). Teachers’ sense of efﬁcacy and the

66

organizational health of schools. ﬁre Elementary School Journal, 93 (4), 355-
371.

Zhu, W., Updyke, W. F ., & Lewandowski, C. (1997). Post-hoe Rasch analysis of optimal

categorization of an ordered-response scale. Journal of Outcome Measurement,
1(4), 286-304.

67