,, n .-e

r

.zfigxm “5.8%“.
in («an
v

.a.’ r.
.

.47“

3.0..» n4." .
$1.740 .

. .
:Kr.h§ .
.22»; in“ u.

«
... t E.
.mfiezafi
z? ..a.w........:s...n ,

. n 3."
‘5 . t 5

kg

.

.

. 2.1!...sth P1
JerzrfldflumrL u :15
z...

9M! 108,1... .
. f. 11%; 2.
O ..-31. '3. . .
m.

.........s... n.»

t. i105
.54.).

L... .m. f. a :3
fifrvm... r.~. fig

!

 

... . av... .11.:
, . . . 1'2.
£25fi 3 ,
. . i. ,
.1. .3. b... _ .

‘III‘

 

fig.” .mfizpfixmafifi.

. . 31.2: x &.

 

 

m... LIBRARY
Michigan State

2003 University

 

 

 

This is to certify that the
dissertation entitled

RELATIONS AMONG LEISURE AS TIME, ACTIVITY, AND
EXPERIENCE IN AFTER-SCHOOL PROGRAMS: INDIVIDUAL AND
PROGRAMMATIC FACTORS

presented by

HENG-CHIEH WU

has been accepted towards fulfillment
of the requirements for the

Community, Agriculture,

Ph-D- degree m Recreation and Resource Studies

 

 

ale/mg

Major Professor’s Signature

{4‘3”} 02’] 2M5)

Date

MSU is an affirmative-action, equal-opportunity employer

o-u---.--.-c-—<-------o--—-—-a---o--n—n--—-.-.-.-o-.-.-.--n---.--.-.-.-.-.-.-.-.-.-v-.-—.—-n-u-n-.-.-u-

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 DATE DUE

 

we» > We

090413

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K IProj/Acc8-Pres/ClRC/DateDue indd

 

RELATIONS AMONG LEISURE AS TIME, ACTIVITY, AND EXPERIENCE IN
AFTER-SCHOOL PROGRAMS: INDIVIDUAL AND PROGRAMMATIC FACTORS

BY

Heng-Chieh Wu

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Community, Agriculture, Recreation and Resource Studies

2008

ABSTRACT

RELATIONS AMONG LEISURE AS TIME, ACTIVITY, AND EXPERIENCE IN
AFT ER-SCHOOL PROGRAMS: INDIVIDUAL AND PROGRAMMATIC FACTORS

By

Heng-Chieh Wu

The federal government funded the 21$t Century Community Learning Centers

(21St CCLC) under the Improving America’s Schools Act of 1994; the program has since

emerged as the largest after-school program (ASP) initiative in the United States. The
reauthorization under the No Child Lefi Behind Act (N CLB) in 2002 shifted the focus

from community education to a specific emphasis on academic success.

Using data from the state evaluation of Michigan 21St CCLC programs, this

study examined how different program characteristics, experiences, and actual

participation in leisure activities were associated with participants’ leisure
. . . . . . t
expenences (enjoyment and voluntary part1c1patton) 1n the 21S CCLC ASP context.

[Two overarching areas were addressed in this study. The first focused on
characteristics of programs that were associated with a greater proportion of leisure
participation, as well as how participation in different types of leisure activities (e. g.,
“relaxed leisure”--- free-play/social time and “transitional leisure”--- sports, arts, and
youth development activities) was distributed as a function of program characteristics.
The second targeted leisure activity participation, both overall and in conjunction with
students’ perceptions of program quality and programmatic characteristics, that may be

associated with student leisure experiences. In both areas, the results of this study can

contribute not only theoretical understanding of ASP processes, especially for

academically oriented ASPs such as 21St CCLC, but also practical guidance that has

implications for retaining students in such programs.

For [Mom

I Love You.

iv

'0

ACKNOWLEDGEMENTS

I know I will sound like a broken record, but I really have to say this one more
time -- Thank You Laurie, for your demonstration of great leadership, diligence,
intelligence, generosity, kindness, humor, and most importantly, for your taking me to
be part of your research team at MSU Outreach and Engagement. Working with you

,f‘”

for the last five plus years is just a blessing and a life-changing experience; an

W—u’

 

r—i .—

experience that has allowed me to grow to be the kind of person I want to be and to get

into a profession that has true meaning in my life.

Celeste, Laura, Beth, Megan, andNai-Iguan -- Thank you for all these years’

A, ,,_, __,_ ___- . in..- .r
’ __ .__._ “-1 - tn.-

 

teamwork and friendship. Many times you made me wonder how people can be so
different and yet work so well together. Many times, I am overwhelmed by the
warmth you give me and, and you guys never fail to amaze me with your wonderful

qualities as individuals as well as your admirable work ethics as my teammates.
t

r. Paulsen «37 You are my mentor in all aspects, my “Lau-Ban” who treats me

.u. .h —-

the other way around, and my major advisor who helps megain the professional . . .

... W1 need through my involvement in national and international conferences

and my trips to China, Mexico and South Africa. The countless meetings we have had
have consistently reminded me of how lucky I am and how I should best represent you,
my precious mentor and my closest fatherffigure _i»n,,lif_e,._,,,._ i I I -
Dr. Villarruel, Dr. Rosenbaum, and dear Dr. Bristor in Heaven -- I have to
WM

reiterate my gratitude to you for being such a supportive committee who responded to

my work promptly, reflectively, and insightfully.

Pavlina, my crazy Czech colleague in the past, present, and I am sure in the
future -- you and Victoria are my extended family from Mars and your positive energy
and compassion toward other people have inspired me to care for the poor and to
improve human rights.

_, f Nai-UK’uan, your name is listed twice here because we play, work, eat, and do
everything together, including the Edewood and Never Land. We share families and
the love you, Jessica, and Martina have generously put in my heart will last forever.

For my ‘partner’ Ariel, ‘sister’ Lei, ‘first-Chinese-fiiend’ Mike, ‘tennis and life

v—--"--" .

o-AH wM

 

coach’ MyungSun, my ‘family-back-home’ uncle Mike, and my ‘very special fi'iend’

Chih-Hsiung -- Thank you for being so patient and loving with me through the

 

 

numerous moments of laughter and tears, happiness and grief, excitement and
disappointment, and achievements and hardships. I appreciate your presence in my life

and the quality of time we spend together. You warm me and make me laugh.

vi

TABLE OF CONTENTS

 

 

 

 

 

LIST OF TABLES ~ -- ~ ‘ ‘ ‘ s “ ‘ ix
LIST OF FIGURES vvvvvvv - ~ - - ~ — — - - — - ~ —xi
Introduction vvvvv - - -- -~ vvvvvvvvvvv = ----- - ‘ 1
Literature Review “ “ = 8
What iS Leisure? ............................................................................................................ 8
Leisure in the ASP Context ......................................................................................... 13
ASPS as Contexts for Leisure Time .................................................................... 13
ASPS as COUtCXtS for Leisure Activities ............................................................. 14
ASPS as COHtCXtS for Leisure Experiences ......................................................... 17
Program-Level Characteristics of ASPs: Different Leisure Contexts ------------------------- 19
Organizational Type ............................................................................................ 20
Academic Participation Policy ............................................................................ 20
Leisure Activities Offered ................................................................................... 21
Breadth 0f Activities Offered .............................................................................. 21
Program Improvement FOCUS .............................................................................. 22
Youth Experiences in the Program .............................................................................. 22
YOUth Perceptions 0f Program Quality ............................................................... 22
YOUth Participation in Activities ........................................................................ 23
Research Questions and Hypotheses ........................................................................... 27
Method ‘ ‘ “ = 33
Sample ......................................................................................................................... 33
Missing Data Irnputation ............................................................................................. 35
Measures and Procedure .............................................................................................. 36
EZReports ........................................................................................................... 37
Program Satisfaction Surveys ............................................................................. 38
Site Annual Report FOI'InS .................................................................................. 4O
StUdent'LeVCI Variables .............................................................................................. 4O
Demographics ..................................................................................................... 40
Dosage ................................................................................................................. 40
Perceptions 0f Program Quality .......................................................................... 42 .
Leisure Experiences ............................................................................................ 42
Site-Level Variables .................................................................................................... 43
Organization Type .............................................................................................. 43
Total Leisure Activities Offered ......................................................................... 43
Types Of Leisure ACtiVIty Offered ...................................................................... 43
Activity Breadth Offered .................................................................................... 44

vii

 

Resu is — - ~ ..................................... - - - ----45

Descriptive Statistics .................................................................................................... 45
Analytic Strategy ......................................................................................................... 50
Results for Research Question 1: What Influences Leisure Participation? ----------------- 55
Results for Research Question 2: What Contributes to Leisure Experiences? ------------ 65
Discussion - -- - 89

 

Research Question 1: What Characteristics Are Associated with Leisure Dosage? "-90
Research Question 2: What Characteristics Are Associated with Leisure

 

Experiences? ................................................................................................................ 92
Limitations ................................................................................................................. 100
Recommendations for FUturC Research ..................................................................... 103
Recommendations for ASP Planners ......................................................................... 105
References “““ “ ‘ 109

viii

Table 4-1:
Table 4-2:
Table 4-3:
Table 4-4:
Table 4-5:
Table 4-6:

Table 4-7: Final Models for Programmatic Predictors of Leisure Dosage (Sample

Table 4-8:

Table 4-9:

LIST OF TABLES

Sample 1 Descriptives ................................................................ 47
Sample 2 Descriptives ................................................................ 49
Results from Null Models Predicting Leisure Dosage ....................... 55
Final Models for Programmatic Predictors of Leisure Dosage ............ .59
Variances Explained in the Leisure Dosages Models.........................6O

Results from Null Models Predicting Leisure Dosages (Sample 2) ...... 61

2) .......................................................................................... 64
Variances Explain in the Leisure Dosages Models (Sample 2)........ 63

Results from Unconditional Models Predicting Leisure Experiences....66

Table 4-10: Results from Overall Leisure Dosage Models Predicting Leisure

Experiences ............................................................................ 68

Table 4-11: Results from Program Perceptions Models Predicting Leisure

Experiences ............................................................................ 69

Table 4-12: Results from Program Perceptions and Overall Leisure Dosage

Models Predicting Leisure Experiences)........................................ 71

Table 4-13: Results from Program Perceptions and Relaxed Leisure Dosage

Model Predicting Leisure Experience (Enjoyment) ............................ 72

Table 4-14: Results from Program Perceptions and Arts Dosage Model

Predicting Leisure Experience (Enjoyment) ........................................ 74

Table 4-15: Results from Program Perceptions and Youth Development Dosage

Model Predicting Leisure Experience (Enjoyment) ........................ 75

Table 4-16: Results from Program Characteristics Model Predicting Leisure

Experiences ............................................................................ 78

Table 4-17: Results from Program Characteristics, Program Perceptions and

Overall Leisure Dosage Model Predicting Leisure Experiences .......... 81

ix

Table 4-18: Results from Program Characteristics, Program Perceptions and Free
Play/ Social Time Dosage Model Predicting Leisure Experience
(Enjoyment) .......................................................................... 83

Table 4-19: Results from Program Characteristics, Program Perceptions and
Sports Dosage Model Predicting Leisure Experience (Enjoyment) ..... 85

Table 4-20: Results from Program Characteristics, Program Perceptions and Arts
Dosage Model Predicting Leisure Experience (Enjoyment) ................ 86

Table 4-21: Variances Explained in Proportion to the Enjoyment Baseline Model 87

Table 4-22: Variances Explained in Proportion to the Enjoyment Baseline
Model ................................................................................... 88

LIST OF FIGURES

Figure 2—1: Study Framework .................................................................... 32

Figure 4-1: The Interaction Effect between Relaxed Leisure and Peer Support on
Students’ Overall Enjoyment ...................................................... 73

Figure 4-2: The Interaction Effect between Youth Development Dosage and
Governance Opportunities on Students’ Overall Enjoyment .............. 75

Figure 4-3: The Cross-level Interaction Effect between Sports Dosage and
Operating Organization Type on Students’ Overall Enjoyment .......... 84

xi

Introduction
After-school programs (ASPs) first emerged with the intention of providing a

refuge and diversion from the streets in the immigrant neighborhoods of major cities in

the late 19th century (Halpern, 2003). At that time, local storefronts, settlement houses,

and churches were the major facilities that provided childcare services afier school. Many
of these local entities later organized themselves into national youth organizations (e.g.,
Boys Clubs, YMCA, Scouts and 4—H). Before the mid-1990s, most responsibility for
ASPs was left to the non—govemmental organizations (N GOs) due to the lack of
investment and formal involvement of the federal government (Gayl, 2004). Many of
these NGOs are still in operation today, and the ASPs that are offered by these national or
local NGOs usually utilize, integrate, or maximize neighborhood resources with an
ultimate vision to promote positive youth development in a broad sense (e. g., the Search
Institute’s 40 assets).

A report from the National Center for Education Statistics indicated that the
percentage of public schools offering extended-day programs for school-aged youth
tripled between 1987 and 1999 (Kleiner, Nolin, & Chapman, 2004). Halpern (1999)
observed that the growing interest in ASPs is driven by four major factors: (a) the belief
that public space is no longer safe for children; (b) the feeling that it is stressful and
unstimulating when children care for themselves; (c) the concern that many children need
academic remediation; and (d) the conviction that low-income children deserve the same
opportunities for enrichment activities, such as arts and sports, as their more advantaged

peers.

Faced with the challenges and needs of many families in arranging after-school

childcare, in 1994, the federal government funded the 21St Century Community Learning

Centers (21St CCLC) program, which is the largest ASP initiative in existence. 215t

CCLC was originally created by the Improving America’s Schools Act of 1994 (PL.

103-3 82) and was designed for the broader use of school resources by local communities

(J ames-Burdumy et al., 2005). Federal funding for 21St CCLC grew from a $1 million

demonstration effort in 1997 to a $40 million program in 1998, expanding to $1 billion in
2002; funding has generally remained stable since. Currently, funding comes from the
federal government and is implemented through Departments of Education at the state
level.

The reauthorization under the No Child Lefi Behind Act (NCLB) in 2002 shifted

the focus of the 21St CCLC programs from community education to a specific emphasis

on academic success. The NCLB Act requires states to provide supplemental services to
low-income students in schools that fail to achieve adequate yearly progress; because the
supplemental services must occur outside the regular school day, the ftmded ASPs are

held responsible for producing academic improvement outcomes (J ames-Burdumy et al.,

2005). 21St CCLC’s mission is to provide academic and other enrichment activities that

reinforce and complement the regular academic program (N aftzger et al., 2007). In fact,
the federal performance targets exclusively focus on academic achievement and
classroom behavior (e.g., improved grades and state assessment scores in
reading/language arts and mathematics; improved homework completion and classroom

attendance according to teacher surveys) (N afizger et al., 2007). As of December 2006,

about 66% of national grant recipients were school districts, and 89% of funded programs
were operated in schools (N afizger et al., 2007). Compared to a national study in 1991
that indicated that less than 10% of the ASP directors identified “improving academic

skills of all children” as the most important program purpose (Vandell & Shumow, 1999),

the current expectations for 21St CCLC programs to improve students’ academic
performance distinguishes 21St CCLC programs from other models. In fact, a recent
federal report on 21St CCLC program offerings across 22 states during the 2005-06

school year indicated that more than half of the 2,142 21St CCLC centers (53%) either

provided mostly tutoring or homework help activities or mostly academic enrichment
activities designed to enhance reading, math, and writing skills; only 20% of centers
provided mostly recreational activities and only 27% offered a wide variety of activities

across the aforementioned categories (N aftzger et al., 2007). Although activity offerings

varied substantially across 21St CCLC programs and not all program directors practiced

the academic-only approach--in practice, the majority of 21St CCLC programs in

Michigan offer a mixture of activities that includes both academic and non-academic

curriculum for youth—-21St CCLC remains well known as a billion-dollar ASP initiative

that specifically targets school success (McCallion, 2003).

Mounting evidence has shown that hi gh-quality ASPs or school-based
extracurricular activities have the potential to produce beneficial youth outcomes such as
enhanced safety and prevention of j uvenile crime (Hirsch, 2005; National Institute on

Out-of-School Time, 2007), lower rates of school drop-out and delinquency (Davalos,

Chavez, & Guardiola, 1999; Eccles & Barber, 1999; Mahoney & Stattin, 2000),
improved academic performance and school behaviors (Darling, Caldwell, & Smith,
2005; Eccles & Barber, 1999; Mahoney, Lord, & Carryl, 2005) and personal and
social-emotional skills (Barber, Eccles, & Stone, 2001; Fredricks et al., 2002; Hansen,
Larson, & Dworking, 2003; Hirsch, 2005). However, less attention has been given to
understanding participation preferences and motives of the youth themselves.
Researchers recognize that recruiting for and sustaining ASP participation can be
challenging (Kirshner, O'Doonoghue, & McLaughlin, 2002; Lauver & Little, 2005).
Parents may express concerns about ASPs being too structured, too chaotic, or too boring
(V andell & Shumow, 1999), and school-age youth, especially older adolescents, may
be uninterested in program offerings, responsible for family duties, or confronted with
competing interests after school such as hanging out with friends, part-time jobs, or
extracurricular activities (Belle, 1999). As Noam (2005) noted, a prerequisite to showing
any benefits of program participation is that youth attend; and this can be especially
challenging with older youth who can decide for themselves how to spend their time after
school. In addition, because, unlike school, participation in ASPs is usually not mandated,
youth who choose to participate in an ASP are essentially taking more initiative in
engaging in learning, whether that learning is focused on academic or non-academic
skills. As a result, it is imperative for program administrators to look closely at program
design and context in order to attract and sustain youth’s participation.

In response to an increase in juvenile crimes and youth issues that emerged in the
late 19808, policy makers called for greater utilization of the resources available through

parks and recreation agencies in order to reach youth (Witt & Crompton, 1996). Since

then, leisure studies has become one of the major foci for youth development work
(Eccles & Templeton, 2002). Researchers and theorists believe that leisure is an
important domain of adolescents’ life because it can provide “a wide variety of
opportunities for exploration of possibilities, expression of interests, and experimentation
with action alternatives " (Kleiber, 1999). During their leisure time, adolescents are given
opportunities to practice personal control over their environments, act autonomously,
experience competence and enjoyment, explore identity, access social networks, and
acquire skills that can help them successfully transition into adulthood (Brown &
Theobald, 1998; Darling, Caldwell, & Smith, 2005; Eccles & Barber, 1999; Kleiber,
Larson, & Csikszentrnihalyi, 1986).

Previous studies have shown that students tend to perceive school-related
activities as coercive, constraining, and boring (Csikszentrnihalyi & Larson, 1984;

Csikszentrnihalyi, Larson, & Prescott, 1977) even if they are doing the activities in ASPs

(Shemoff & Vandell, 2007). As the current govemment-funded 21St CCLC ASPs are

confronted with tremendous pressure to supply academic enrichment and produce
improved test scores (Gayl, 2004; Halpem, 1999; Hirsch, 2005; US Department of
Education, 2007), ensuring that ASPs are not merely an extension of the same school—day
experience and providing youth with enjoyable and engaging learning experiences can be
a real challenge (Shemoff & Vandell, 2008).

The field of leisure studies has made some effort to investigate youth program
participation and its potential association with positive outcomes (Caldwell & Smith,
2006; Darling, Caldwell, & Smith, 2005; West & Crompton, 2001; Witt & Caldwell,

2005), with an assumption that youth program participation is a form of leisure. However,

participation in a youth program does not automatically mean that it will be perceived as
leisure. Furthermore, researchers have suggested using recreation and leisure activities as
incentives to encourage participation in ASPs (Hirsch, 2005; Lauver & Little, 2005);
empirical studies are still needed to indicate how participation in leisure activity can play
a role in increasing participants’ perception of engagement in the ASPs as leisure and
eventually recruit and sustain participation. As youth cannot receive the expected
program benefits if they do not attend the program, more work is required to understand
what factors influence participation (Borden et al., 2006; Eccles, 2005; Perkins et al.,
2007; Wimer et al., 2006), especially in a more academic-oriented ASP context.

Thus far, the majority of studies about ASPs have examined participation
globally without distinguishing among the kinds of activities in which the youth
participates; only a few studies have examined ASP participants’ experiences with
different types of the activities, and these have tended to use the Experience Sampling
Method (ESM; see Larson, Hansen, & Moneta, 2006; Shemoff & Vandell, 2007;
Vandell et al., 2005). Although these studies have provided essential information on
immediate experiences resulting from different activity participation, how different
combinations of activity participation relate to participants’ overall experiences—for
example, as reflected in program satisfaction, a common way for program
administrators to assess their program success—remains unknown. Studying youth’s
perceptions of participation in ASPs as leisure experiences—namely, their enjoyment
and voluntary participation—is important because it not only reflects program quality

from the participant’s perspective, but also provides the groundwork for

understanding how to sustain youth’s participation to convey program benefits

(Vandell & Shumow, 1999).

Using data from the state evaluation of Michigan 21St CCLC programs, this

study examined how different program characteristics, experiences, and actual
participation in recreation and non-academic enrichment activities were associated
with participants’ perceptions of leisure in the structured, school-based ASP context.
A main focus was on how different types of the leisure activity participation could be
related to leisure experiences, while taking into account different program
characteristics and experiences.

Because data from adolescent participants were nested in program sites and
programs varied significantly in organizational type and foci, multilevel modeling
(HLM, v.6.20; Raudenbush & Bryk, 2002) was used to serve the research purposes.

The literature review for this study was conducted following the themes below:

1. What aspects of ASP participation can be considered as leisure?

2. What site-level characteristics of ASPs are associated with greater participation in
ASP leisure activities (e. g., recreation and non-academic enrichment)?

3. When youth participate in higher levels of leisure activities (e. g., recreation and
non—academic enrichment), are they more likely to perceive their experiences in
the program as leisure?

4. Do the relationships between leisure participation and leisure experiences differ as

a function of program quality or programmatic characteristics?

Literature Review
What is Leisure?

Leisure researchers have been challenged in the process of trying to define
leisure. Edginton, C. R., Jordan, DeGraaf & Edginton, SR. (2002) summarized seven
ways to define leisure: Leisure as time, as an activity, as a state of mind, as a symbol
of social status, as a social instrument, as an anti-utilitarian concept, and as a part of a
holistic process. However, for the most part, studies have conceptualized leisure as
time, activity, and/or experience (Hazel, Langenau, & Levine, 1990; Murphy, 2000;
Parrish, Hikido, & Fowler, 1998; Ruskin et al., 2001).

Leisure time. To many, leisure is best identified as "flee or discretionary time"
(Meyer, 1964), referring to time during people’s lives when they can act without
obligation (Durnazedier; Hamilton-Smith, 1991). The idea of viewing leisure as time
occurred primarily as a result of the Industrial Revolution, when one’s free time was
used to rest and recuperate in order to go back to work (J urriu, 2000). Iso-Ahola (1999)
exarrrined contemporary society and argued that although people might refer to their
nonworking hours as free time, only a small proportion of this time can indeed be
considered “free” in the sense that we now use the word—free from obligation for
subsistence (addressing physical needs) or work activities.

Csikszentrrrihalyi, Larson and Prescott (1977) conducted the first empirical
study specifically about adolescents’ leisure time. Twenty-five Chicago-area
adolescents aged 13 to 18 years carried pocket-sized electronic paging devices and
received several signals each day for a week at random times between 8 A.M. and 11

PM. The researchers found that adolescents were most likely to spend time (one-third)

in conversation with peers. Additionally, although adolescents considered homework
and class to be challenging and valuable skill-building activities, they also felt higher
levels of constraint, coercion, and boredom in comparison to talking with peers and
participating in games and sports. Adolescent (leisure) time use has continued to be of
interest (Caldwell, Smith, & Weissinger, 1992), supported by literature focused on
understanding the out-of-school lives of youth (Belle, 1999; Darling, Caldwell, &
Smith, 2005; Gordon & Caltabiano, 1996; Hirsch, 2005; Mahoney, Larson, & Eccles,
2005)

Leisure activities. Another definition views leisure as a unique and discrete set
of activities. Kraus (1990) suggested that this view of leisure is closely aligned with
that of recreation in that most people would consider recreation activities to overlap
substantially with leisure activities. Some researchers also note that leisure and
recreation seem to share a great deal in common based on the verbal meanings
adolescents assigned to the terms and that many recreational activities are considered
leisure (Mobily, 1989; Shaw, 1986). Leisure and recreation activities can include
sports, outdoor recreation, performing arts (dance, drama, music), visual arts, travel
and tourism, crafts, self-improvement, hobbies, social recreation, and voluntary
services (Edginton, Jordan, DeGraaf, & Edginton, 2002). However, participating in a
leisure or recreation activity does not always promise the experience of leisure; an
activity that is “leisure” for one person may not be for another person. For example,
many students would consider playing softball to be a leisure or recreation activity,
but for a student who does not want to participate or feels awkward and unskilled at

sports, it can seem like anything but. As a result, it is impossible to obtain a set of

activities fully encompassed by the term “leisure” (Roadburg, 1983; Shaw, 1985).
Nonetheless, recreation activities are often employed as one way to help define and
understand leisure (Edginton, Jordan, DeGraaf, & Edginton, 2002).

Leisure experiences. Perhaps the most common and widely accepted approach
is to define leisure as an experience—a function of an individual’s state of mind.
DeGrazia (1962) proposed that leisure is “a state of being in which activity is
performed for its own sake” (p. 13). Others have emphasized perceived freedom and
intrinsic motivation as critical components of the leisure experience (Iso-Ahola,
1999), with the former being most important (N eulinger & Breit, 1969). Theorists
who stress the experiential dimension of leisure believe that leisure cannot be
experienced unless it is based on one’s choice, freedom, and self-determination.

However, some researchers argue that no practice, including leisure, is free
from the power structure formed by the inequalities of the society. Put differently, the
choices that one can make about how to allot nonworking time are often constrained
by outside conditions, including time, money, and information accessibility.
Therefore, under those conditions, claiming leisure is based on one’s “free choice”
overlooks the socially constructed reality and, within this definition, is falsely
considered leisure (Fox, 2000; J uniu & Henderson, 2001). However, Rojek, in his
book “Leisure and Culture, ” emphasized the role of human agency in pursuing
leisure. He noted that all of those who had conceptualized the experiential nature of
leisure centered on an ultimate notion, freedom. Even if full freedom is not obtainable,

human beings still make decisions about their leisure within the constraints that exist.

10

Therefore, he argued that overlooking all considerations of choice in one’s leisure is
equal to reducing human beings to structural and cultural automatons.

Aware of the ambiguity in using “choice” or “freedom” to define
experiential leisure, Stebbins (2001) proposed that the distinguishing factor is “lack of
coercion” rather than “free choice”. In his view, leisure is “uncoerced activity
undertaken during free time where such activity is something people want to do and,
at a personally satisfying level using their abilities and resources, they succeed in
doing” (p. 350). He emphasized the “agreeable obligation” as an attitude and form of
behavior that is very much a part of leisure, as such obligation accompanies positive
attachment to an activity that brings the participant pleasant memories and meets the
person’s expectation. The concepts of choice and freely chosen, in this context,
should not be the definers of leisure, but utilized as one of the important aspects to
understand human beings’ leisure experience (Blumer, 1969; Ruskin et al., 2001).

Thus, for some, adolescent leisure experiences can be conceptualized as a
pleasurable state of mind in conjunction with the perception that their activity or time
is occurring within a context of freedom, choice, or self-determination (Iso-Ahola,
1980; Raymore, Godbey, & Crawford, 1994). From a theoretical standpoint, students
who feel that they have they have chosen what they do with their time or have choice
about how they use that time should experience greater competence, intrinsic
motivation, and enjoyment from their leisure experiences than those who perceive
less freedom; moreover, these experiences have the potential to carry over to other

domains of their lives and help shape behaviors and attitudes for successful transition

11

into adulthood (Guinn, Semper, & J orgensen, 1996; Hultsman & Kaufman, 1990;
Kleiber, Larson, & Csikszentrnihalyi, 1986).

Kleiber, Larson, and Csikszenmihalyi (1986) found that adolescents’ leisure
activities in general are associated with higher levels of perceived freedom, intrinsic
motivation and positive affect than what they termed productive (e. g., studying) and
maintenance (e. g., eating and resting) activities. They also argued that leisure
activities can be divided into two different categories: relaxed leisure and transitional
leisure. Relaxed leisure refers to activities that require less concentration and have
relatively few intellectual or physical challenges, such as socializing with friends,
playing, and watching television. On the other hand, transitional leisure, such as
sports, arts, and games, present relatively high levels of challenge and demand
concentration. The authors suggest that transitional leisure has special significance for
adolescents, because in those activities, adolescents have opportunities to experience
higher levels of freedom and intrinsic motivation than doing other activities such as
productive or maintenance work. Meanwhile, these activities, in contrast to relaxed
leisure, also require adolescents to use discipline and to extend their ways of knowing
and experiences. As transitional leisure elicits enjoyment through the exertion of
effort and by providing opportunities for mastery, adolescents who participate in
transitional activities grow better equipped to enjoy obligatory activities and settings
such as school and work.

Other researchers have examined the psychological properties that tend to be
associated with adolescent leisure experiences. For example, Mobily (1989)

conducted a qualitative study to examine the meanings of recreation and leisure

12

among adolescents by studying high school students who were unfamiliar with the
definitional concept of leisure. Analysis of the students’ answers revealed three
subjective aspects of leisure that resonated with the adolescents: the absence of
self-evaluation, enjoyment and choice. Caldwell, Smith, and Weissinger (1992)
developed the Leisure Experience Battery for Adolescents (LEBA) intending to
capture four dimensions of adolescent leisure experience that had been widely
discussed in the literature: awareness, boredom, challenge, and distress/anxiety. The
results confirmed that adolescents often reported these experiences during their
leisure time and activities and suggested that examining these psychological
properties may be a way to characterize the quality of adolescents’ leisure
experiences.
Leisure in the ASP Context

Within the framework of time, activity, and experience, ASPs often, wholly or
in part, provide a leisure context for participating youth.
ASPs as Contexts for Leisure Time

From the perspective of “leisure time,” programs usually operate after the
regular school day. This fosters a natural context for youth to experience leisure,
given that, for most, school life is their primary “work” or obligatory task, consuming
one-quarter of a youth’s waking hours (Larson & Verma, 1999). Several reports
indicate that youth in the United States have substantial discretionary time; but also
that much of this time is unstructured, unsupervised, and unproductive (Carnegie
Council on Adolescent Development, 1992; Larson & Verrna, 1999; United Nations,

2007). As researchers are concerned that many youth do not know how to make

13

their discretionary time meaningful and developmentally appropriate (Caldwell,
Baldwin, & Walls, 2004), much of the literature has focused on the positive aspects
of structured leisure time and negative aspects of unstructured leisure time (Caldwell
& Smith, 2006). ASP participation, in this context, is usually considered a positive
use of youth’s leisure time and one of the most popular forms of structured leisure.
ASPs as Contexts for Leisure Activities

From the perspective of “leisure activities,” ASP participation can be
reviewed as participants’ leisure because it often offers a variety of leisure activities.
Practical reports from the ASP field indicated that recreation activities offered in
ASPs can include, but are not limited to, sports, free-play, board games, social events,
arts, music, dance, theater, computer and technology, and youth development
curriculums such as career development, character development, community
service/service learning, conflict resolution, cultural activities, leadership
development, life skills, resistance programs, and healthy living (George, Cusick,
Wasserman, & Gladden, 2007; Naftzger et al., 2007; Reisner, White, Russell, &
Birmingham, 2004; Van Egeren & Reed, 2007). Leisure activities offered in ASPs
can also be categorized into the two major types of leisure described by Kleiber and
colleagues (1986): transitional and relaxed leisure.

Transitional leisure. Transitional leisure activities are often referred to as
“skill-building” or “hi gh-investment” activities, and some researchers term them
“structured activities” because they are typically organized and facilitated by adults
(Caldwell & Smith, 2006; Kleiber, 1999; Mahoney, Larson, Eccles, & Lord, 2005;

Osgood, Anderson, & Shaffer, 2005; Ruskin et al., 2001). ASPs can be a critical

l4

context for enabling access to transitional leisure activities. Researchers have found
that youth spend more time doing sports, arts, community service, academic
enrichment, homework and other learning opportunities at school-based ASPs versus
elsewhere (Posner & Vandell, 1999; Vandell et al., 2005). In fact, increasing youth’s
accessibility to enrichment activities, especially for low-income and at-risk youth, is
one of the major forces that have initiated and sustained public support for ASPs
(Halpern, 1999; Mahoney, Larson, & Eccles, 2005; McGillis, 1996).

Relaxed leisure. Although some researchers are concerned that unstructured
youth leisure may be associated with behavioral problems such as aggression,
juvenile crime and gang activities, low academic performance and delinquency, and
experimentation with alcohol, tobacco, substance abuse and teen pregnancy (Colwell,
Pettit, Meece, Bates, & Dodge, 2001; Osgood, Anderson, & Shaffer, 2005; Sherman
et al., 1998), others advocate the importance of providing youth with environments
that cultivate unstructured social interactions and allow them to “hang out” during the
non-school hours (Devereux, 1976; Rivkin, 1995; Walker & Arbreton, 2001). For
many youth, social contexts that are not led by adults provide opportunities to
experience personal control, practice autonomy (Silbereisen & Eyferth, 1986),
experiment with social roles, behaviors, and ideas that may support the transition to
adulthood (Caldwell & Darling, 1999), and develop skills such as fiiendship
formation, playfulness, and negotiation (Kleiber, 1999; Larson & Verrna, 1999;
Sutton-Smith, 1997). Sutton-Smith (1994) further argued that the significant amount

of time that contemporary Western preadolescents and adolescents spend in sports

15

and other organized leisure activities constitutes an historical trend of adults
supervising, controlling, and rationing youth free time.

It is notable that some ASPs offer periods of free play and social time, similar
to school recess, during which students do activities on their own with little
curriculum or structure involved. These free—play-based recreation activities can form
a major component of an after-school program, although this was more likely in the
past when most after-school programs functioned as school-age child care rather than
enrichment programs (Halpern, 2002). Although conducted within the context of
ASPs, these types of activities may be experienced by youth as unstructured leisure,
or the “relaxed leisure” defined by Kleiber, Larson, & Csikszentmihalyi (1986).

Transitional vs. relaxed leisure. To date, ahnost no published work has
examined the differential impacts of transitional and relaxed leisure participation in
ASPs. In one of the few studies, Mahoney and Stattin (2000) surveyed 703
14-year-old Swedish youth to examine whether participation in structured leisure or
non-structured leisure was associated with different levels of antisocial behavior.
Non-structured leisure activities were operationalized as participation in Swedish
Youth Recreation Centers (YRCs), which provided activities such as billiards,
ping-pong, field trips and basketball with little adult supervision. Structured leisure
activities were conceived as participation in non-YRC community programs, where
youth got to participate in activities such as sports, music, theater, and fine arts that
were typically highly structured and closely supervised by adults. Results indicated

that youth who participated in structured leisure activities (at non-YRCs) were less

16

likely to evidence antisocial behavior than youth who participated in non-structured
leisure (at YRCs) (Mahoney & Stattin, 2000).

Referencing these results, Osgood, Anderson and Shaffer (2005) suggested
that youth programs should offer structured activities and that providing unstructured
activities may increase rather than decrease problem behavior. However, results from
the Mahoney and Stattin’s study are inconclusive; the authors cautioned the potential
effect of selection bias as their study was not based on an experimental design.
Furthermore, structured versus non-structured activity participation was a
program-level factor-that is, all the structured leisure activities were located at
non-YRC programs, and all the non-structured leisure activities were located at YRC
programs. The results, therefore, may be due to the overall program quality (YRCs’
exhibiting overall lower quality compared to non-YRCs) rather than the effects of
leisure activity types. In fact, this possibility is supported by the authors’ description
of the centers’ lack of well-organized activities, little emphasis on skill-building, and
staff whose background and education varied substantially. In addition, the study
operationalized participation as attended or did not attend without taking into account
variations in individual dosage, which would allow examination of the extent to
which more or less participation was linked to more or less problem behavior. While
an interesting descriptive study, it says little about the differential impact of
structured and non-structured leisure activity participation on youth behavior.

ASPs as Contexts for Leisure Experiences
Finally, the experiential approach to leisure defines it as uncoerced behavior

(Ruskin et al., 2001) in conjunction with a pleasurable state of mind (Iso-Ahola, 1980;

17

Raymore, Godbey, & Crawford, 1994). Although adolescence is a critical period for
the development of autonomy (Eccles, Early, Fraser, Belansky, & McCartby, 1997;
Ryan & Deci, 2000), little literature is available regarding the predictors or
consequences when youth participate in ASPs out of choice as opposed to when they
are compelled by others.

Self-deterrrrination theory (Deci & Ryan, 1985; 2002) proposes that healthy
human psyches gravitate toward situations that provide autonomy, competence, and
relatedness. Moreover, human beings have a natural tendency to master and integrate
their experiences into a coherent sense of self. This internalization and integration process
is contingent upon social factors that either facilitate or impede it. Intrinsic motivation
arises when individuals perceive free choice, participating an activity based on positive
feelings, beliefs, and attitudes (Ryan & Deci, 2000). When youth are intrinsically
motivated, their sense of agency will emerge (Larson, 2000). In addition, the fulfillment of
the autonomy, competence and relatedness can elicit a sense of engagement, which is
defined by Connell & Wellbom (1991) as a combination of participation, attention, tasks,
and positive affect. As young people's perceptions, values, and skills are influenced by
their relationships with others and the contextual constraints or opportunities available to
them (Leffert et al., 1998), it is imperative to examine how ASPs as unique social,
cultural, educational, and physical ecologies could foster youth’s self-determination to
attend the ASPs (Mahoney, Larson, & Eccles, 2005; Noam, 2004; Simpkins, Ripke,

Huston, & Eccles, 2005).

18

Program—Level Characteristics of ASPS: Different Leisure Contexts

Researchers have two general approaches to characterizing ASPs. One defines
ASPs as any after-school or out-of-school-time activity, including both traditional
extracurricular activities such as clubs, sports teams, church and youth groups, and
lessons (e. g., dance, music) and more comprehensive programs such as 21st CCLC
(Mahoney, Larson, & Eccles, 2005). The other approach is to focus on ASPs like let
CCLC (e. g., Vandell & Shumow, 1999) which provide more comprehensive curriculum
that span the breadth of homework and tutoring to sports, arts, youth development, and
community service, with youth taking part in a range of activities (N aftzger et al., 2007).
This study focuses on the latter type of ASP.

Researchers in general have stressed the importance of utilizing ASPs as part of
ecosystem that connects family, schools, and community together to provide youth with
essential learning opportunities and a safe environment to stay after school (Baldwin,
Caldwell, & Witt, 2005; Eccles & Gootrnan, 2002; Hirsch, 2005; Villarruel,
Montero-Sieburth, Dunbar, & Outley, 2005; Villarruel, Perkins, Borden, & Keith, 2003).
ASPs are supervised by adults and convene regularly in the two to four hours after school
at the school site or at a community-based organization, and share a common goal of
providing a safe environment for school-age youth. ASPs vary greatly on programmatic
factors such as the grade level or ethnic backgrounds of the major service population, the
number of youth served, and number of days of operation in a year or a week (Eccles &
Templeton, 2002; Halpem, 1999; Vandell & Shumow, 1999). Program-level factors that

might be related to youth participation in ASPs are introduced below:

19

Organizational Type
The type of organization that oversees the ASP can be a major factor in the

development of program priorities and activities offered. A federal report on let CCLC
revealed that most grant recipients were school districts and located in school buildings
(Naftzger et al., 2007). As schools tend to be skilled and familiar with academics and
academic enrichment approaches, and given that the 2lst CCLC initiative emphasizes a
need to show improved school outcomes, school-based ASPs have shown a strong
academic focus with curricula most exclusively targeting academic-oriented activities
(Chang-Rios, 2007; Naftzger et al., 2007). On the other hand, ASPs offered by NGOs,
such as faith—based, community-based, and non-profit organizations, often have a long
history of youth programming and are more likely to use a youth development
philosophy (George, Cusick, Wasserman, & Gladden, 2007; Hirsch, 2005). Although
studies are not available that compare how different operating types of the ASPs might
prioritize their activities differently, it is reasonable to propose that the different
organizational philosophies and resources might lead to different emphases or approaches
to after-school programming.
Academic Participation Policy

In practice, some ASPs have specific policies in place regarding student
participation in academic activities. Attendance policies are usually employed to ensure
that students receive sufficient academic instruction. Some programs may impose the
policy on all attendees, while others only apply it to students identified as in need of
academic remediation. The designation and implementation of the policy can largely

reflect the program foci and priorities on academic subjects. As previous literature have

20

stressed that students tend to perceive low levels of enjoyment and engagement while
doing school-related activities (Csikszentmihalyi & Larson, 1984; Csikszentmihalyi,
Larson, & Prescott, 1977) even within the ASP context (Shemoff & Vandell, 2007),
programs that impose a stricter attendance policy on academic activity participation may
result in less enjoyment and desire to attend compared to those programs with looser
policies.
Leisure Activities Offered
Similarly, programs vary in the extent to which they offer leisure activities vs.

academic activities. The proportion of leisure activity offerings out of all offerings is
likely to represent the program’s overall philosophy of what makes an effective youth
program and what contributes to successful youth development. Because leisure activities
have been suggested by many researchers and practitioners to be appealing to youth and
encourage their participation in ASPs (George, Cusick, Wasserman, & Gladden, 2007;
Hirsch, 2005; Lauver & Little, 2005), higher proportions of leisure activities offered by
sites may be associated with more motivation to participate and greater enjoyment in the
program.
Breadth of Activities Offered

Researchers have discussed the benefits of programs offering a wide variety of
activities, including fulfilling the needs of different populations, attracting and sustaining
participation, and allowing youth to have choices and practice autonomy (Lauver & Little,
2005; Mahoney, Larson, & Eccles, 2005). As a result, programs that offer a variety of
different types of leisure activities should have youth who express higher levels of

enjoyment and willingness to participate.

21

Program Improvement Focus

Research has indicated that when programs are of low quality, students may be
physically present but cognitively and emotionally unengaged (Vandell, Shumow, &
Posner, 2005). Therefore, both researchers and practitioners have highlighted the
importance of getting feedback from participants to ensure that programs are seen as
interesting, desirable places for youth to go (Lauver & Little, 2005; McKenzie &
Smeltzer, 2001). The utilization of data, such as youth and parent perceptions and
feedback about the program, in the development of program improvement strategies,
should result in hi gher-quality programming and, presumably, greater interest on the part
of participating youth.

Youth Experiences in the Program

Numerous studies have identified experiences associated with high—quality youth
programs, which are characterized here by two dimensions: youth perceptions of program
quality and youth activity participation.
Youth Perceptions of Program Quality

Youth need adults who are caring, supportive, and able to model positive
behaviors. In the ASP literature, significant emphasis has been directed toward studying
the quality of staff, specifically in terms of their ability to manage behavior, model
prosocial norms, and provide appropriate supervision, guidance, and support (Calvert,
Zeldin, & Weisenbach, 2002; Hirsch, 2005 ; Vandell, Shumow, & Posner, 2005; Vandell
& Shumow, 1999). In addition, positive peer interactions and healthy social dynamics are
important factors to ensure that youth enjoy and want to attend the ASPs (Darling,

Caldwell, & Smith, 2005; Vandell & Ramanan, 1991).

22

Researchers have also stressed the importance of providing opportunities for
youth to engage in activities that matter to them and set a context for autonomy and
initiative-taking; those are the experiences that entail enjoying obligatory activities over a
long period and are essential to a successfirl transition to the adulthood (Kleiber, Larson,
& Csikszentrnihalyi, 1986; Larson, 2000). Therefore, ASP activities that help promote
youth governance and leadership experiences or provide youth with hands-on,
challenging and fun activities are highly recommended in the current literature
(Innovation by Design and Center for Teen Empowerment, 2000; Lauver & Little, 2005;
Quinn, 1999; Villarruel & Lerner, 1994).

Youth Participation in Activities

Participation at all: Coming to the program. Program activities are often the most
important vehicle to recruit for and retain youth in ASPs (McLaughlin, Irby, & Langman,
1994). Although empirical studies are lacking that examine the impacts of leisure activity
participation on participants’ program enjoyment and desire to attend, previous studies
and evaluation reports about ASPs have documented some notable activity preferences
and participation patterns.

For example, Hirsch’s (2005) qualitative study in six low-income, minority, urban
Boys & Girls Clubs over a four-year period indicated that physical and recreational
activities were by far the youth’s favorite activities. In general, sports is particularly of
interest to many ASP participants (Lauver & Little, 2005; Oulette, Hutchinson, & F rant,
2005), and this is especially true for boys (Hirsch, 2005; Lerner, 2005) whose
participation in structured leisure declines overall with age due to competing interests

(Theokas, Lerner, Phelps, & Lerner, 2006). Coatsworth and Conroy (2007) suggested that

23

integrating sport as part of ASP activities could produce enhanced program effects on
youth physical, socio-emotional, and academic well-being. However, sport participation
has also found associated with negative outcomes, such as age-inappropriate behaviors and
negative peer interactions (Hansen, Larson, & Dworking, 2003) and drinking activity
(Barber, Eccles, & Stone, 2001). Arts have been found to appeal to many program
participants, especially girls (Hirsch, 2005; Lerner, 2005).

Previous studies have shown that older youth’s involvement in ASPs or
extracurricular activities decreases as they reached higher grade levels (McNeal, 1999;
Vandell & Shumow, 1999). In addition, older youth may experience higher levels of
boredom in ASPs because activities can be mostly designed for younger participants
(Belle, 1999). However, although participation in youth development activities (such as
community service, conflict resolution, and leadership training) has been found to be
lower overall than in sports, arts, and music (Theokas, Lerner, Phelps, & Lerner, 2006),
youth development activities can be of particular interest to many older youth (Lauver &
Little, 2005). In review of more than 60 evaluations of ASPs, Lauver and Little (2005)
found that older youth were drawn into youth development activities because they might
have the potential to help them apply for colleges or obtain a job in the future. One
unpublished report describing results from a large study of ASPs based in Chicago
revealed that participants enjoyed learning opportunities they were given through arts,
sports, technology, and communications activities (Goerge, Cusick, Wasserman, &
Gladden, 2007).

Although fi'ee play and social time has not been a major focus in the ASP

literature, some researchers feel that ASPs should provide a safe space for students to

24

“hang out” and just “relax” so that they can be away from outside stresses and have an
opportunity to restore themselves or make friends (Hirsch, 2005; Walker & Arbreton,
2001). Studies of school recess, which can be considered a corollary of ASP free time,
claim that recess is “a form of human or animal behavior, self-motivated and carried on
for intrinsic purposes” (Kraus, 1990, p.41), and has important educational and
developmental implications (Pellegrini & Smith, 1993; Sutton-Smith, 1990). Furthermore,
some researchers are concerned that structured/transitional leisure may actually have
negative impacts on youth because it can be overly confined by societal rules, destroying
youth’s initiative and creativity, setting a context for competition and creating stressful
experiences (Caldwell, Smith, & Weissinger, 1992; Devereux, 1976; Rivkin, 1995 ;
Sutton-Smith, 1994).

Degree of participation within the program: Dosage. Weiss, Little and Bouffard
(2005) proposed that program participation can take several forms, including enrollment
(i.e., getting youth in the door), attendance (i.e., the frequency/intensity of time spent in
the program), and engagement (i.e., the extent to which the youth is motivated to
participate). The majority of studies have measured participation as enrollment, or
“absolute attendance” (Fiester, Simpkins, & 'Bouffard, 2005), which classifies youth as
participant or non-participant without considering the types and intensity of participation.
Over the past few years, researchers have begun to move beyond this global approach to
a more sensitive measure: continuous “dosage” (e. g., number of years, days or hours of
attendance) (Eccles & Templeton, 2002; Eccles & Barber, 1999; Hansen, Larson, &
Dworking, 2003; Lauer et al., 2006). In addition, Van Egeren, Wu, & Reed (2007)

proposed the use of the proportion of participation by type of activity in order to better

25

capture what each student references when reporting about the program overall (e. g., did
they mostly participate in homework help? In arts activities?) The authors found that
while students’ academic outcomes were associated with the number of hours of program
attendance, students’ feelings about the program were related to the proportion of
academic, recreational, arts, or youth development activities in which they participated
while there.

In addition, a handful of studies have also suggested that participation breadth,
which refers to the variety of types of activities youth participated within or across
programs, can motivate youth to participate in ASPs (Lauver & Little, 2005; Mahoney,
Larson, & Eccles, 2005), and is associated with more beneficial outcomes, including
academic achievement (Baker & Witt, 1996; Gerber, 1996; Marsh, 1992), higher life
satisfaction (Gilrnan, 2001 ), and positive attitudes toward their plans for education,
college admission, and future occupation (Marsh, 1992; Marsh & Kleitrnan, 2002).
Therefore, youth who participate in a wide variety of different types of activities should
find the program more enjoyable and should be more willing to participate. As of today,

empirical studies are unavailable that have examined whether greater breadth of activities
in comprehensive ASPs, especially programs like 21$t CCLC ASPs that emphasize

academic outcomes, result in high levels of enjoyment or desire to participate.

26

Research Questions and Hypotheses
Based on self-determination theory, it was important to acknowledge factors that
encourage youth participation in ASPs, and leisure time, activity, and experience may play
a critical part in sustaining and strengthening that engagement. The review of the related
literature suggested several program-level and individual-level characteristics that could

influence leisure activity participation and experiences. The purpose of this study was to

. . . . . . . . . st
rdentrfy factors that influence students’ lersure partrcrpatron and experrences 1n the 21

CCLC ASP context.

The study framework consisted of four constructs: Program characteristics, leisure
activity dosages, program perceptions, and leisure experiences (see Figure 2-1). In this
study, leisure activity dosage was defined as (a) overall proportion of leisure participation;
(b) proportion of the separate activity categories that constitute the relaxed leisure
activities (i.e., free play/social time), and transitional leisure activities (i.e., sports, arts,
and youth development); and (c) breadth of leisure activity. Leisure experiences were
defined as overall enjoyment and voluntary participation. Leisure as time was embedded
in the framework in two different ways: (a) the leisure activity dosages capture the use of
free (after-school) time; and (b) the relaxed leisure that allows students unstructured time
within the program was differentiated from more structured transitional leisure.

It was proposed that leisure activity dosages would be varied based on different
program characteristics. Also, students’ leisure dosages would influence their leisure
experiences, and the relationships might be differed by different program perceptions and
characteristics. A detailed description of the research questions and hypotheses was listed

below:

27

Research Question 1: What influence leisure participation?

Research Question 1.]: Are program characteristics related to leisure dosages?

I hypothesize that programs with the following characteristics will have
students who report higher leisure dosages:

- Community-based organizations

Greater breadth of activity offerings

- Higher proportion of hours of leisure activities offered

Looser requirements for academic activity participation

More use of evaluation data for program improvement

Research Question 2: What contributes to leisure experiences?

Research Question 2.1: Are leisure activity dosages related to leisure experiences?

I hypothesize that higher overall leisure dosage (in contrast to academic
dosage) will result in higher levels of enjoyment and greater voluntary
participation rates.

I hypothesize that greater leisure activity breadth will result in higher levels of
enjoyment and greater voluntary participation rates.

In addition, because the previous literature did not present sufficient
information to predict which types of the leisure dosages might provide more
positive leisure experiences, each of the following types of the leisure dosages
will be explored with regard to its relationship to students’ perceptions of
leisure and voluntary participation.

- free play/social time (this is also relaxed leisure in contrast to transitional

leisure), or

28

- youth development

Research Question 2.2: Are students who perceive the program to be of higher
quality more likely to report more positive leisure
experiences ?

- I hypothesize that students who report higher levels of the following program
qualities will be more likely to view their participation as leisure experiences:

- Staff supportiveness
- Lack of staff injustice
- Opportunities for governance
- Positive peer relationships

Research Question 2.3: Do relationships between leisure dosages and leisure

experiences differ as a function of perceived program quality?
- I hypothesize that the relationships between leisure dosages (overall leisure

dosage; dosage by types (i.e., free play/social time, sports, arts, youth
development); and activity breadth) and leisure experiences will moderated by

students’ program perceptions around:

- Staff supportiveness
- Staffinjustice

Opportunities for governance

29

- Positive peer relationships

Research Question 2.4: Are program characteristics associated with students ' leisure
experiences?
- I hypothesize that programs with the following characteristics will have
students who report more positive leisure experiences:

- Community-based organizations

Higher proportion of leisure activity provision

Greater breadth of leisure activity offerings

Looser requirements for academic activity participation

Greater utilization of evaluative data on leisure activity improvement

Research Question 2.5: After accounting for student level factors, do relationships
between leisure dosages and leisure experiences difl'er as a
function of program characteristics?

. Because there is a lack of literature addressing how different program
characteristics might influence the relationships between leisure dosages
(overall leisure dosage; dosage by types (i.e., free play/social time, sports, arts,
youth development); and activity breadth) and leisure experiences, each of the
following program characteristics will be explored as moderators of these

relationships:

30

Community-based organizations

Higher proportion of leisure activity provision

Greater breadth of leisure activity offerings

Looser requirements for academic activity participation

Greater utilization of evaluative data on leisure activity improvement

31

 

 

Program Characteristics

Operating organization
Leisure activity offered (predictor; control variable for the level-one
proportion of hours of leisure activity)

0 Activity breadth offered (predictor; control variable for the level-one
recreation activity breadth)

Academic participation policy
0 Utilization of evaluative data for improvement

 

 

 

Leisure Activity Dosages

Proportion of hours in general leisure
Proportion of hours in different types
of leisure activities
- Relaxed leisure

0 Free play and socials
-* Transitional leisure

 

V

 

 

0 Sports
0 Arts Leisure Experiences
0 Youth development

0 Recreation activity breadth 0 Overall enjoyment

 

 

0 Voluntary

 

 

 

>< Participa‘i“

 

 

Program Perceptions

Staff supportiveness

Staff injustice

Opportunities for governance
Peer relationships

 

 

*Note: Level-1 control variables: Gender, grade level, race/ethnicity, and total
days attended.

Figure 2-1: Study Framework

32

Method

Sample

The data used in this study were collected as part of the Michigan 21St CCLC

Statewide Evaluation conducted by Michigan State University’s (MSU) Community

Evaluation and Research Center (CERC), a department of University Outreach and

Engagement. The Michigan Department of Education (MDE) requires 21St CCLC grant

applicants to serve schools in which 30% or more of the students enrolled are eligible for
free or reduced price meals (most sites far exceed this requirement), with priority given to
schools that are eligible for Title I school-wide programs, serve students from
high-priority schools, or are middle-school students (Michigan let Century Community

Learning Centers, 2008).

During the 2006-07 school year, 20,519 students were enrolled in 188 21St CCLC

programs coordinated by 32 grantees. Approximately 92% of the sites (n = 172) were

operated by public school districts. Sites operated an average of 125 days in a year,

ranging from 42 to 214 days. Although 21st CCLC programs served students from Pre-K
to 12th grade as well as adult family members of participating students, this study focused

th th . . . . . .
on 4 - to 12 —grade program partrcrpants rn order to utrlrze student surveys grven to that

age group.

Among the 20,519 students, 14,481 students were at 4th to 12th grade levels.

About 1% of students (203 students out of 14,481) had attendance recorded only for meal

times or only attended special events, field trips, or less frequently offered activities such

33

as computer/technology and health sessions (accounted for 3% and 1.6%, respectively,
within the total amount of activity offerings). Because these activities were not offered on
a regular basis or the length and the availability of the activities varied significantly
across programs, students who only attended these activities were excluded from the
present study. For the remaining 14,278 students, their time spent in these activities was
also eliminated from the dosage calculation because they did not fall in the primary
activity categories being studied. .

There were essentially two samples for this study. For the questions about

dosages and program characteristics (Research Question 1), the sample (Sample 1)

consisted of 14,278 4th-12th graders from 182 programs, with an average grade level of

6th grade and a fairly even gender proportion. Participants were mostly Afiican American

(73%), with 18% white, 6% Hispanic, 2% Arabic, and 1% other race/ethnicity. The
average attendance for this population was 40 days (SD = 35.65), ranging from 1 to 159
days in the school year. The majority of the programs were run by schools (86%), with
the remainder administered by non-profit organization such as community-based
organizations (CBOs).

For questions that included students’ perceptions of program quality and their
leisure experiences (Research Questions 2 series), the sample (Sample 2) only included
students who completed the Program Satisfaction Survey (N=4,160 from 165 programs).
The 17 programs that were not included in Sample 2 showed similar program
characteristics as the 165 included programs, evidenced by the similar data distribution

and no statistical difference resulting from t-test analyses. The only exception was

34

program type; all the excluded programs were school-based programs as opposed to the

mixed program types in Sample 2.
The average grade level of this second set of the sample was 6th grade, and the

majority were females (53%) and African American students (66%), leaving the rest
white (23%), Hispanic (7%), Arabic (3%) and other race/ethnicities (1%). Because
students had to participate till the end of the programs to receive the surveys, the average
attendance was 64 days, which was considerably higher than the Sample 1 population. In
fact, Sample 2 could be considered as a sampling product of Sample 1, with Sample 1
being the true population that consisted of all eligible students.
Missing Data Imputation

Sample 2 was defined by the existence of Program Improvement Survey data.
From that survey, four program perception measures and the two outcome variables were
formed. At the preliminary data cleaning stage, missing values were detected on the six
variables, with 1% ~ 2% of the data missing on the overall enjoyment and program
perceptions measures, and 13% of the data missing on the voluntary participation
outcome. The relatively large difference in the missing data proportion between the
voluntary participation and the rest of the measures might be due to the order of the
questions listed in the survey, in which the voluntary participation question was the first
question. Students may have overlooked this question because it was listed on the very
top of the survey or because they had not become oriented to the surveying process. The
difference in the missing data prOportions may also be due to the fact that this measure
was only made by one single question, whereas the rest of the measures were formed by

multiple questions (number of items ranged from 3 to 8). The computation of the mean

35

score for each measure was based on a completion rate of 75% or above for all
corresponding responses. As a result, measures that were consisted of multiple questions
were more likely to be assigned a valid value than a single-question measure.

To eliminate the impact of missing data and account for the nested effects, the
statistical program R version 2.7.0 (R Development Core Team, 2008) was used to
impute the missing data. Because the six measures that had missing data included both
continuous and categorical variables, an add-on package called ‘mix’ in addition to the
original R program (Schafer, 2007) was also employed. Students’ demographic variables,
total attendance, and the information pertaining which programs they participated in were
included in the imputation process along with all the six measures. By utilizing the R
(mixed) program, the missing data within each measure were replaced by the plausible
imputed values assigned by the expectation-maximization (EM) algorithm. Irnputed
values that fell slightly outside of the acceptable range for a given scale were rounded to
the nearest acceptable values as suggested by Peugh & Enders (2004). This process
enabled the missing data to be estimated using not only individual-level data, but also
program-level data, and preserved the nested nature of the dataset.

Measures and Procedure

The primary purpose of the Michigan 21St CCLC Statewide Evaluation Project is

to assess student academic outcomes for federal reporting and to identify characteristics
of more or less successful programs. Thus, the choice of measures used in this study was
constrained by the data available from the overall evaluation project. Through this project,
three major data sources were used in this study: (a) a demographics and attendance

tracking system called EZReports, (b) Program Satisfaction Surveys, and (c) Site Annual

36

Report Forms. For clarity, measures and procedures will be described here, and student-
and site-level variables will be described in separate sections.
liZReporur

EZReports is an online data tracking system in which program staff report on site
information and individual demographic and attendance, which is entered by date and
activity. Each site has staff enter the data for that site, and training, technical assistance,
and monitoring are provided by MSU. Sites enter each activity in their schedule into
EZReports, including its objective, description, activity category, session length, and
dates and times offered.

Because staff tend not to be consistent in classifying the activity types and
because activity types may be categorized by administrative support staff who are
responsible for data entry rather than by experienced youth providers, all program
activities were reviewed and further recoded by the author and checked by the MSU team
into a variety of 25 sub-categories based on the activity’s name, descriptions and
objectives provided by the program staff. The sub-categories included: General academic
enrichment, homework help, tutoring, cultural studies, team sports, non-team sports, free
play, games, social activities, career development, character education, community
services, conflict resolution, leadership development, independent living skills, personal
and psychosocial development, mentoring, resistance skills, computers, video/media, arts,
music, dance, theater, and health/nutrition. Because some activities might include
multiple sub-categories, a coding strategy was developed by assigning the primary
sub-category for each activity, with an additional secondary sub-category to be identified

if needed.

37

For the purposes of this study, all activities were then grouped into five major
categories: academic activities (including general academic enrichment, homework help,
tutoring, and cultural studies), free play/social time (including free play and social
activities), sports (including team and non-team sports), arts (including visual arts, music,
dance and theater) and youth development activities (including career development,
character education, community services, conflict resolution, leadership development,
independent living skills, personal and psychosocial development, mentoring, and
resistance skills). Students’ dosages were calculated based on the hours of participation
recorded in these five categories. If a student attended an activity that fell into multiple
categories, the participating time was given a value of 1 if it fell in the primary category
and .5 if it fell in the secondary category‘. For example, if a student spent an hour in an
activity that “allows students to learn how to say no to drugs and to work on an art piece
at the end of the class to promote what they have learned,” then the dosage for this
student was 1 hour for youth development activities (resistance skills) and 0.5 hour for art
(visual art).

Program Satisfaction Surveys
The Program Satisfaction Survey was developed by the 21St CCLC Statewide

Evaluation team in collaboration with program administrators, local evaluators, and MDE.
It was designed to assess student’s perception of program quality and experiences.

For the study sample (Sample 2) in this study, paper surveys were generated by

. . . . . . th
MSU for a total of 9,679 students across 177 srtes; rnclusron crrterra were 4 - to

 

The author also calculated the dosage based on only primary categories. The differences were fairly

small and results from the study did not vary by the two methods. The inclusion of the secondary category
in the dosage calculation process fit in better with the study focus.

38

12th-grade students who attended program activities that were eligible for the study for at

least one day of attendance within three months prior to the survey administration period,
which is during the mid to late April in 2007. Surveys were printed with bar codes that
included the student’s EZReports ID and site and grantee information. A first page listed
the student’s name, grade level, and the site and grantee names. A waiver of consent has
been approved for this survey; however, the first page also describes the voluntary nature
of the survey and how confidentiality will be maintained. Program staff were responsible
for administering the surveys to eligible students during a time allotted for academics or
quiet play. After completing the survey, students were asked to remove the first page with
the identifying information in order to maintain confidentiality of responses from

program staff. Program staff collected the de-identified surveys and sent them back to the

21St CCLC statewide evaluation team at MSU.

Surveys were formatted to facilitate scanning and automatic upload of responses
using Remark OCR software. The ID embedded in the barcode printed on the survey was
used to link student attendance and demographic information with the survey responses.
During the survey administration period, a total of 3,115 students to whom the surveys
were addressed turned out to be no longer attending and were removed from the
calculation of response rate. The final survey sample was comprised of 4,160 valid
responses from 163 programs, with an overall return rate of 63% after removing the
non-attendees from the calculation. The primary reason for non-response was staff not
administering the survey as requested. Only data that were relevant to the current study

will be described below.

39

Site Annual Report Forms

The Annual Report Form (ARF) was an online reporting form in which Michigan
let CCLC grantees reported on their annual progress for 2006-07 to MDE. Two forms
were available, one at the grantee or organizational level and one about each specific site.
Data from the site-level ARF were used in this study. Program administrators complete
the forms.

Student-Level Variables

Demographics

Individual-level data retrieved from EZReports includes demographic variables:
gender, race/ethnicity, and grade level. Gender were coded as 1=male and O=female.
Race/ethnicity were coded as a series of dummy variables (l=yes, 0=no) for the
following racial/ethnic groups: white, black, Hispanics and Middle Eastern. Grade level
was a continuous variable ranging from 4 to 12, with the numbers reflecting the actual
grade levels.
Dosage

Overall, this study uses the total days of attendance as an indicator of intensity,
the number of activity categories attended as a reference to breadth, and proposes an
inventive way of featuring attendance, which is to consider the proportion of attendance
in certain types of activities in order to capture the experience that youth reference when
considering their perceptions of the program regardless of the number of hours they
attended (Van Egeren, Wu, & Reed, 2007).

Total days of attendance. The total number of days that a student attended during

the 06-07 school year.

40

Overall Leisure Dosage. In this study, the phrase “recreation activity” and

“leisure activity” were interchangeable. Leisure activity was defined in its broadest sense,
which includes any activities that were not targeted on academic enrichment or
enhancement (such as general academic enrichment, homework help and tutoring). The
overall leisure dosage was a participation proportion variable that calculated based on the
sum of participating hours in all non-academic activities divided by the overall hours of
participation. The values ranged from 0 to 1, with 0 representing no participation in
recreation (or, say, participation was all in academic activities), and 1 representing all
participation in recreation activities (that is, no participation was in academic activities).

Leisure Dosage Types. Participation in different types of leisure activities was
conceptualized on the following four categories: free play/social time, sports, arts, and
youth development activities. The dosage for each leisure activity type was calculated
based on the total participating hours in each category divided by the total leisure activity
participating hours. Values ranged from O to 1, with 0 meaning no participation in the
category and 1 meaning all participation in leisure activity was in this specific category.
In addition, the free play/social time dosage also represented the ‘relaxed leisure,’ with
the opposite of that being the ‘transitional leisure.’ The value of the free play/social time
equaled to 1 indicated that all leisure participation fell into the ‘relaxed leisure’ category,
while 0 meant that all the leisure time was spent in ‘transitional leisure.’

Leisure Activity Breadth. The leisure activity breadth referred to the total numbers
of different leisure categories attended, which was based on the count of whether the
student was participating in the following four categories: free play/social time, sports,

arts, and youth development. The possible value ranged from 0 to 4, with 0 meaning none

41

of the aforementioned categories was ever attended during the 06-07 school year, while 4
indicating the full participation breadth.
Perceptions of Program Quality

Program perceptions that pertain to participants’ perceptions of program quality
were assessed in four domains: Stajfsupportiveness (5 items, Cronbach’s Alpha=0.84)
assessed students’ perceptions of the degree of staff warmth, caring, and interest in
students (e.g., “Staff care about me” and “staff really try to make activities interesting
and fun”; staff injustice (3 items, Cronbach’s Alpha=0.66), assessed students’ perceptions
of the extent to which staff are unfair (e. g., “Staff punish kids without even knowing what
really happened” and “Staff don’t care what I think”; opportunities for governance and
decision-making (8 items, Cronbach’s Alpha=0.82) got at students’ perceptions of the
degree to which the program has opportunities for choice and autonomy (e.g., “Staff and
kids decide the rules together,” “Kids get to choose their activities,” and “Staff listen to
kids’ ideas about the program”); and peer relationships (7 items, Cronbach’s Alpha=0.83)
addressed students’ perceptions of the degree to which students in the program support
one another (e. g., “Kids help each other out” and “Kids treat each other with respect”).
Leisure Experiences

Leisure experiences were represented by two variables: Overall program
enjoyment and voluntary participation. Overall program enjoyment were assessed by
four questions: “I look forward to coming to the program,” “When I don't come to this
program, I miss it,” “I get bored at this program,” and “I would tell other kids to come to
this program for fun activities” (N of items—=4; Cronbach’s alpha=0.67). Voluntary

participation was assessed by one question: “What is the MOST IMPORTANT reason

42

that you come to this program?” Responses were: “I want to come,” “My parents want
me to come,” and “A teacher, principle, or counselor wants me to come.” The answers
were dummy-coded as 1 =“voluntary participation” (“1 want to come”) and 0 = “not
voluntary participation” (the other two options).
Site-Level Variables
Organization Type
Each program’s operating organization referred to the type of grantee who

administered the grant, and was dummy-coded as l = school vs. 0 = non-profit
organization. The type of school included public schools, regional/intermediate education
agencies, college/universities, and charter schools. Non-profit organizations included
community-based organizations or other not-for-profit organizations.
Total Leisure Activities Oflered

Students were constrained in their ability to participate in leisure activities by the
degree to which sites offer these types of activities. Overall leisure activity provision was
calculated based on the total hours of leisure activities of any type that programs offered
in proportion to all the hours of activities that the program offered in their whole
curriculum. Data were retrieved from EZReports.
Types of Leisure Activity Oflered

The time programs offered in each type of the leisure activities (free play/social
time, sports, arts, youth development) was also calculated based on the offering hours of
all activities in that category in proportion to the total leisure activity offering hours. Data

were retrieved from EZReports.

43

Activity Breadth Offered
This measure referred to the total number of leisure activity types (free play/social
time, sports, arts, and youth development) offered by a site. Data were retrieved from

EZReports.

44

Results
Descriptive Statistics

To answer the Research Question 1: “What influence leisure dosages?,” data fi'om
Sample 1 was analyzed and presented in Table 4-1. The average overall leisure dosage
was 54% (SD = 32%), meaning that about 54% of students’ time was spent in leisure
activities, leaving about 46% (SD = 32%) of the time spent in acaderrrics. Within the time
students’ spent in leisure activities, an average of about 19% was spent in relaxed leisure
(free play/social time, SD = 32%) as opposed to in transitional leisure (81%, SD = 32%).
Within transitional leisure, students spent an average of 25% of their time in sports (SD =
34%), 26% in arts (SD = 32%), 19% in free play/social time (SD = 32%), and 16% in
youth development activities (SD = 27%). The average activity breadth was 1.83 (SD =
1.22), meaning that students on average participated in less than two leisure activities
during their ASP participation. At the site level, programs on average set up 65% (SD =
18%) of their curriculum in leisure activities, leaving 35% of the time in academics (SD =
18%). Within the leisure curriculum, arts activities were offered the most (33%, SD =
21%), followed by sports (24%, SD = 21%), youth development (23%, SD = 20%) and
free play/social times (19%, SD = 24%). The average site-level breadth of activity
offerings was 3.08 (SD = 1.01), which was much higher than what students actually
attended. Programs in general had some formal form of policy in place for academic
activity participation (evidenced by M = 2.44, SD = .65, possible range from 0 to 3), and
a low utilization of evaluative data for leisure curriculum improvement (M = .95, possible

range from 0 to 2).

45

Study participants in Sample 2 had similar participation patterns as Sample 1 (see
Table 4-2), with the overall leisure dosage at 55% (SD = 30%). They also participated
most in sports and arts (both at 25%, SD = 31%, 29%, respectively), followed by the free
play/social time (23 %, SD = 33%) and youth development activities (17%, SD = 24%).
Sample 2 students participated in a greater breadth of activities than Sample 1 (N=2.19,
SD = .13), most likely due to their greater participation (M = 64 and SD = 38 days in
Sample 2 vs. M = 40 and SD = 36 days in Sample 1). In general, students rated the four
program quality questions at a medium to high level, (M = 2.67 to 3.33, possible range
from 1 to 4), and about 71% of the students reported that their participation was voluntary
as opposed to impelled by adults (29%). Students rated their participation at a relatively
high level of enjoyment (M = 3.2, SD = .74). The characteristics of the Sample 2
programs were almost identical to the Sample 1 programs, which was reasonable given
that Sample 2 was comprised of 165 programs from within Sample 1’5 182 programs.

Table 4-2 presented the descriptives of Sample 2.

46

Table 4-1: Sample 1 Descriptives

 

 

 

 

 

 

Variable Name Description M Median SD Min Max
LEVEL ONE (N=l4,278)
Quftcome Variables
Leisure Dosage Proportion of Participating Hours in
Leisure Activities in General 0'54 0'5 5 0'32 0‘00 1'00
Free Play/ Social Proportion of Participating Hours in
Tune D°Sage F“? Play 0’ 30ml Tm"? 0.19 0.00 0.32 0.00 1.00
Actrvrtres among All Leisure
Activities
Sports Proportion of Participating Hours in
Sports Activities among All 0.25 0.00 0.34 0.00 1.00
Leisure Activities
Arts Proportion of Participating Hours in
Arts Activities among All Leisure 0.26 0.10 0.32 0.00 1.00
Activities
Youth Proportion of Participating Hours in
Development Youth Development Activities 0.16 0.00 0.27 0.00 1.00
among All Leisure Activities
Leisure Activity Number of Types of Leisure
Breadth Activities Participated (Ranged 1.83 2.00 1.22 0.00 4.00
from 0 to 4)
[n_dependent Variables
Gender Student 3 Gender (l=Male, 0.49 0.00 0.50 0.00 1.00
O-Female)
Grade Level Student’s Grade Level 6.46 6.00 1.71 4.00 12.00
White White (l=Yes, 0=No) 0.18 0.00 0.39 0.00 1.00
Black Black (l=Yes, 0=No) 0.73 1.00 0.44 0.00 1.00
Hispanics Hispanics (l=Yes, 0=No) 0.06 0.00 0.23 0.00 1.00
Middle Eastern Middle Eastern (l=Yes, O=No) 0.02 0.00 0.13 0.00 1.00
Total Days of Student’s Total Days Attended in the
Attendance 06-07 Regular School Year 4021 ”'00 35'65 1'00 ”900
LEVEL TWO (N2182)
Organization The Types of Operating Organization
Type (l=School-based, 0=NGO/CBO) 0'86 1'00 0'35 0'00 1'00
Leisure Activity Proportion of Hours Offered in
Offered Leisure Activities in General 0'65 0'67 0'18 0'00 1'00
Free Play/ Social Proportion of Hours Offered in Free
Time Activity Play/ Social Time Activities
Offered among All Leisure Activities 0'19 0'11 0'24 0'00 1'00
Offered
Sports Activity Proportion of Hours Offered in
Offered Sports Activities among All 0.24 0.22 0.21 0.00 1.00
Leisure Activities Offered
Arts Activity Proportion of Hours Offered in Arts
Offered Activities among All Leisure 0.33 0.33 0.21 0.00 1.00
Activities Offered
Youth Proportion of Hours Offered in
Development Youth Development Activities
Activity among All Leisure Activities 0’23 0'20 0'20 0'00 1'00
Offered Offered

 

47

Table 4-1: Sample 1 Descriptives (Continued)

 

 

Variable Name Description M Median SD Min Max
Leisure Activity Number of Types of Leisure
Breadth Offered Activities Offered (Ranged from 3.08 3.00 1.01 0.00 4.00
0 to 4)
Academic Academic Participation Policy
Participation (Ranged from O to 3) 2.44 2.67 0.65 0.33 3.00
Policy
Utilization of Utilization of evaluative data for
evaluative data improvement (Ranged from 0 to 0.95 1.00 0.54 0.00 2.00
for improvement 2)

 

48

Table 4-2: Sample 2 Descriptives

 

 

 

 

 

 

Variable Name Description Mean Median SD Min Max
LEVEL ONE fN=4+160)
_O_utcome Variables
Voluntary Voluntary Participation
participation (l=Yes, 0=No) 0.71 1.00 0.46 0.00 1.00
Enjoyment Enjoyment (Ranged from 1 to 4) 3.02 3.25 0.74 1.00 4.00
Independent Variables
Gender Student 5 Gender (l=Male, 0.47 0.00 0.50 000 1.00
0—Female)
Grade Level Student’s Grade Level 6.20 6.00 1.72 4.00 12.00
White White (l=Yes, 0=No) 0.23 0.00 0.42 0.00 1.00
Black Black (l=Yes, 0=No) 0.66 1.00 0.48 0.00 1.00
Hispanics Hispanics (l=Yes, 0=No) 0.07 0.00 0.25 0.00 1.00
Middle Eastern Middle Eastern (l=Yes, 0=No) 0.03 0.00 0.16 0.00 1.00
Total Days of Student’s Total Days Attended in the 159.0
Attendance 06-07 Regular School Year 63'64 “'00 38°26 1'00 0
Staff . Staff Supportrveness (Ranged from 3.33 3.60 0.73 1.00 4.00
Supportrveness l to 4)
Staff Injustice Staff Injustice (Ranged from 1 to 4) 2.06 2.00 0.87 1.00 4.00
Governance Opportunities for Governance
(Ranged from 1 to 4) 2.72 2.75 0.72 1.00 4.00
Peer . . Peer Relatronshrps (Ranged from 1 2. 67 2.71 0.69 1.00 4.00
Relatronshrps to 4)
Leisure Dosage Proportion of Partrcrpatrng Hours m 0.55 0. 5 5 0. 30 0.00 1.00
Leisure Actrvrtres rn General
Free Play/Social Proportion of Participating Hours in
Tm" D°Sage F“? may °’ S°°"1T‘me. 0.23 0.03 0.33 0.00 1.00
Actrvrtres among All Leisure
Activities
Sports Proportion of Participating Hours in
Sports Activities among All 0.25 0.10 0.31 0.00 1.00
Leisure Activities
Arts Proportion of Participating Hours in
Arts Activities among All Leisure 0.25 0.14 0.29 0.00 1.00
Activities
Youth Proportion of Participating Hours in
Development Youth Development Activities 0.17 0.04 0.24 0.00 1.00
among All Leisure Activities
Leisure Activity Number of Types of Leisure
Breadth Activities Participated (Ranged 2.19 2.00 1.30 0.00 4.00
from 0 to 4)
LEVEL TWO (N=165)
Organization The Types of Operating
Type Organization (l=School-based, 0.85 1.00 0.36 0.00 1.00
0=NGO/CBO)
Leisure Activity Proportion of Hours Offered in
Offered Leisure Activities in General 0'64 0'67 0'18 0'00 1'00
Free Play/Social Proportion of Hours Offered in Free
Time Activity Play/Social Time Activities
Offered among All Leisure Activities 0'19 0'12 0'24 0'00 1'00

Offered

49

Table 4-2: Sample 2 Descriptives (Continued)

 

 

 

Variable Name Description Mean Median SD Min Max
Sports Activity Proportion of Hours Offered in
Offered Sports Activities among All 0.24 0.22 0.22 0.00 1.00
Leisure Activities Offered
Arts Activity Proportion of Hours Offered in
Offered Arts Activities among All 0.33 0.33 0.21 0.00 1.00
Leisure Activities Offered
Youth Proportion of Hours Offered in
Development Youth Development Activities
Activity Offered among All Leisure Activities 0'23 0'18 0'20 0'00 1'00
Offered
Leisure Activity Number of Types of Leisure
Breadth Activities Offered (Ranged from 3.07 3.00 1.03 0.00 4.00
Offered 0 to 4)
Academic Academic Participation Policy
Participation (Ranged from 0 to 3) 2.47 2.67 0.63 0.33 3.00
Policy
Utilization of Utilization of evaluative data for
evaluative data improvement (Ranged from 0 to 0.96 1.00 0.55 0.00 2.00
for improvement 2)
Analytic Strategy

Given the nature of student data nested within programs, hierarchical linear

modeling (HLM) was employed to address variations across program contexts (HLM

v.6.20; Raudenbush & Bryk, 2002). A series of two-level HLM models were conducted,

with student-level variables at Level 1 and program-level characteristics at Level 2.

Based on the concept of multilevel modeling, the Level-1 model was designed to produce

estimates within each Level-2 program j. The Level-1 model could be presented as:

Yij: B0} + l3!) le+ 8:]

Where Y ,j is the observation for the ith student in Level-2 program j, [301- is the Level-1

intercept within program j, Bl] is a Level-l slope within program j, and Bij is error for

student i in programj. Assuming a Level-1 model with one predictor, X1, the Level-2

models would appear as:

50

Bo]: 1’00 + Y01 Zj + 140}
1311:100 +YHZj + ulj

where BOj is the Level-l intercept in Level-2 program j; 1’00 is the mean value of the
Level-1 outcome, controlling for the Level-2 predictor, Z]; 701 is the slope for the Level-2
variable Zj; uoj is the error for program j; Blj is the Level-1 slope in Level-2 program j;
1’10 is the mean value of the Level-l slope, controlling for the Level-2 predictor, Z -; 711 is

the slope for the Level-2 variable Z}; and u1 J- is the error for program j.

Because normality tests from the residual files showed that the normality
assumption was not met for most of models and the employment of data transformation
techniques failed at correcting the distribution, all the reports were relied on the robust
standard errors, which have been known for its relative insensitivity to misspecification at
the levels of the model and the distributional assumptions at each level. This strategy is
recommended for analyzing multilevel data when the number of the higher level units
(clusters) is moderately large relative to the number of explanatory variables at a higher
level (Luke, 2004).

Null models. Model-building procedures began by testing the null models with no
predictors at both levels in order to examine whether variance existed across programs
and a multi-level analysis was necessitated (if no program-level variance were evidenced,
regular hierarchical multiple regression could be conducted). All variables were
grand-mean centered for the ease of interpretation and the fixed effects were assumed for

all Level-1 variables. In addition to statistical test results from the null models, the author

51

also considered the design effect for the proposed models. L. K. Muthen and B. Muthen
(1999) suggested that when a design effect is greater than 2, researchers would need to
take into account the effect of the clustering structure in the data. Statisticians in general
agreed that the estimated design effect for the multi-level modeling is approximately
equal to: l + (average cluster size - l)*intraclass correlation (Carlin & Hocking, 1999;
Elley, Kerse, Chondros, & Robinson, 2005; Muthen & Muthen, 1999).

Unconditional models. After the need for a multi-level model was confirmed, an
unconditional model for each outcome variable was tested. Students’ demographic
variables (i.e., race, gender and grade level) were tested first to evaluate their potential
effects on the outcomes, followed by the Level-1 variables proposed by each research
question and hypothesis.

T wo-Level HLM models. After the unconditional models (models with no Level-2
predictors) were built, Level-2 variables were introduced into the model. Because the
primary focus of the study was to examine how leisure dosage influences leisure
experiences after taking into account students’ demographic differences, experiences in
the program, and program characteristics, Level-2 program effects were tested only on
the intercept (the average score of the Level-1 outcomes) and the dosage slopes (effects
of the dosages on the Level-1 outcomes). For example, to test the effect of organization
type, the dummy-coded Level-2 organization type variable was tested on the Level-l
intercept in order to examine whether students’ average leisure experience score varied
by the two types, and on the slope to examine whether students with greater leisure
dosage reported more positive leisure experiences when they were in CBO-based

programs as opposed to school-administered programs. Level-2 variables were not tested

52

on the slopes of student demographic and total attendance variables, nor on the Level-1
program perceptions, meaning that the Level-2 programmatic variables were not tested to
examine their moderating effects on relations between demographic, attendance, or
perceived program quality and the outcomes. While these are interesting questions, they
were not the focus of the current study.

Effect sizes2 and the statistical test results for the coefficients were considered
throughout the model testing and selection processes. Although there was a lack of
documented studies that presented what a reasonable effect size should be for studying
ASP leisure participation and experiences, a report evaluating 28 ASP studies suggested
an average effect size of 0. 14 for the outcome of school bonding (Durlak & Weissberg,
2007). This might be an appropriate reference for the current study as most of the ASPs
in this study were operated by schools or implemented in the school buildings.

In accordance with the method prescribed by Raudenbush and Bryk (2002), coefficients
that were trivially small were removed before proceeding with analyses. The final models
therefore included only statistically significant variables. Two exceptions to this rule
were total leisure activity dosage at Level 1 and total leisure activity offering at Level 2.
The total leisure dosage/offering was calculated based on the total leisure hours in
proportion to total activity dosage/offering (including academics), and the types of leisure
dosage/offering entered were calculated in proportion to the leisure dosage/offering only
(excluding academics). Therefore, the total leisure dosage at Level 1 and the total leisure

offering at Level 2 allowed evaluation of the impact of different types of leisure

2
I t
2 =
The effect size calculation was based on the Cohen’s r farrrily: r (2 +df . It is commonly

used in social science studies to help explain the direction and magnitude of effects (Cohen, 1988;
McCartney & Rosenthal. 2000).

 

 

53

dosage/offering while controlling for the overall leisure dosage/offering. For example,
two students might have the same amount of dosage in arts (say, 10%), but for one, the
remaining 90% attendance was in homework help, while for the other, the remaining
90% attendance was in sports activities. According to the previous literature and based on
the study hypotheses, the student whose remaining proportion of time was spent in
homework help might report less enjoyment than the student who spent the rest of his
program time in sports. This relationship was unlikely to be identified unless how
students spent their overall program time was controlled. The same rationale could be
applied to the program-level offering, where two programs that both offered 10% of their
leisure curriculum in arts might have students with very different experiences if one
program designed the remaining curriculum to focus on sports while the other focused
the rest of the time on homework help. In addition, because students’ participation
emphasis was constrained by what was available to them (that is, students could not
participate in activities unless they were offered by the program), the Level-2 variables of
the types of leisure activity offerings and activity breadth variables were always
employed as control variables for corresponding dosage outcomes listed in Research

Question 1.

54

Results for Research Question 1: What Influences Leisure Participation?
Research Question 1.1: Are program characteristics related to leisure dosages?

Sample I

Null models. Results from a series of six analyses testing the null models with
leisure dosages as the outcome variables rejected the null hypothesis that the mean
dosages of the programs were equal and indicated significant variability across programs.
The intraclass correlations (ICCs) ranged fi'om 0.37 to 0.64, meaning roughly 37% to
64% of the total variance in students’ participation dosages was associated with programs
as opposed to individual students (see Table 4-3). Activity breadth had the most variance,
while dosages in sports, arts, and youth development had the least variance; nonetheless,
in all cases, substantial variance in students’ dosage of leisure activities was due to
program characteristics. Altogether, results from the tests of the null models suggested
that multilevel analyses were warranted. In addition, the within-program reliability
estimates ranged from 0.97 to 0.99, indicating the estimated differences could be

accepted as reliable indicators of real differences among program’s population means.

Table 4-3: Results from Null Models Predicting Leisure Dosage

 

 

. 2 .
Outcomes (33:22:22: ts X value ICC 11-33:;th Reliability
(df=l 81 )
Overall Leisure Dosage 0.05 l3650.29*** 0.55 43.60 0.98
Free Play/ Social Dosage 0.07 22314.02*** 0.64 50.81 0.99
Sports Dosage 0.04 9559.74*** 0.38 30.43 0.97
Arts Dosage 0.04 8190.96*** 0.37 29.66 0.97
Youth Development Dosage 0.04 10564.84*** 0.48 38.18 0.98
Activity Breadth 0.98 20217.06*** 0.61 48.24 0.99

 

"‘ p <.05;**p <.01;***p <.001.
Note. The design effect is based on average cluster size = 78.45.

55

Unconditional models. Level-1 demographic variables and total days of
attendance were tested prior to the introduction of the Level-2 program characteristics in
order to examine the true effects of the program characteristics after controlling for
individual factors. Table 4-4 summarizes the results. Significant intercepts indicate the
difference of the average level of participation from 0. For students’ overall leisure
participation, no Level-1 demographic and total attendance variables were found to be
significant predictors. For students’ flee play/social time dosage, males and younger
students had higher levels of participation in social/free play than female and older
students. For students’ participation in m, gender and total days of attendance were
significant predictors, meaning male students spent more time in sports, as did students
who participated in the program for a longer period of time. Participation in a_rts was
negatively associated with gender and grade level, and positively related to total days of
attendance, indicating females, younger students, and students with greater attendance in
general participated more in arts activities. Similar to participation patterns for arts
activities, females and students with greater attendance were found to participate more in
youth development activities. In addition, racial minority students were found to
participate more in youth development activities than white students. Finally, the breadth
of students’ leisure participation breadth was found to be negatively associated with
grade level and positively related to total attendance, meaning younger students and
students with greater attendance participated in a greater variety of leisure activities than

older students and students who attended fewer days.

56

T wo-Level HLM models. After removing the insignificant Level-1 variables,
Level-2 program characteristics were introduced into the model and tested for effects on
the intercept (average) of each outcome variable. For overall leisure participation, after
controlling for the extent to which the sites offered more leisure activities in general, the
results indicated that programs that offered more activity breadth had students’ who had
greater participation in leisure activities as a whole. Alternatively, programs that
employed more concrete and stringent attendance policies had students who participated
less in leisure activities (reflecting a higher dosage in academic activity participation).
Among the specific types of leisure activities, no Level-2 predictors were found to predict
dosage in ee la /social time dosage, a_rts; dosage, or youth developing; dosage except
the extent to which the site offered that type of activity, which was included to control for
the opportunities students had to participate in those activities. Only students’
participation in m was associated with Level-2 characteristics; sites that offered more
breadth, indicating that a greater variety of different activity types were available for
students to participate in, was related to more individual students’ participation in sports.

For individual students’ participation in a greater breadth of activities, it was
found that after controlling for site-level breadth, programs that offered a larger
proportion of leisure activities within the whole curriculum and programs that employed
a more stringent policy for students’ attendance in academic activities had students
participate in a greater breadth of different types of leisure activities. Thus, while students
at sites with stringent attendance policies participated less in leisure activities overall,

they participated in a greater variety of leisure activities.

57

Notably, two proposed program characteristics, organization type and utilization
of evaluative data for improvement, were not significantly related to any of the dosage
outcomes, suggesting that these programmatic factors did not carry over to affect
participations at the student level. It is also noteworthy that although many of the
coefficients that reached statistical significance were small (e. g., coefficients found on
slopes of total days of attendance for predicting arts and youth development dosages were
positive values that rounded up to zero), the approximate eflect sizes reached values

around .20, presenting an acceptable magnitude.

58

.on_m “onto a Peon—co .1. mm Hobo 38:8,.” u 8 ”836508 conmvuavgmcs n ~80 .802 :oov a... .1.” 8V minnov a...

 

 

 

 

 

 

 

 

59

 

 

 

 

mad 8.22 _ wt 5.36 53on 33:2
cod 863 m cm: $.86 Basic—goo 58>
36 3.33 cm: 1.86 32
3d 5.8: a: 2:86 38%
36 823 cm: 1.86 .33 8cm
Ed Edens a: .3486 2334 =Eo>o
“5:09:00 -
Essex an :0 859$ . seem 58,5
. . . .............................. -.- .-..- w ............... . . . - 3:8
M: o So *2 o _ M: o 8o :moo oogeo=<
.................... . . . . . . mace-5
m «mo So «38? 3 o So 13o 53on
who mutate
36 mod 1.... and aod 1.306 and mod steed and mod .5..qu .86 mod 4.....on de mod timed mo
ugh.“- 08am
. . . . . . . . . . . . . . . ween-e0
ONO _mo .33— 2o woo 2o 2: So 2o 30 woo So 3o 80 So ............... 8:2
.3
3.0 mod :33 and 86 1.35 36 Ed I...w~.¢ wed 8.0 .3335 and 86 .1336 mod 56 itemd .3285
slam-n
. . . _ . . . . . . ............... m .............................. 8:883.
to 2; rise 2 o 8o .23 «3 8o taco message-_-
.......... --- :6 So .86- 223
N3 :3 $.85- ..... ..... 25 8d :85- ............... 25 5o .86- --------------- .93 035
uuuuuuuuuu ll.- w — .O _Od {Neda de ~O.o *iuwcdr 0N6 N66 5.2.06.0 NN.O —O.C **N0.0 1....- --l- ll- “06:00
m we“ “-010.01 m w “0in H Wm- .«30 w wm .300 m w 980 m % woou Tl—ofl
ommwoD omamom owamom oMmmoQ $3.5 wamofl ,
. . woo-mm noxi
5305 5553‘ Eocene—gon— Eso> wt< 38am .85 been 2364 =Eu>0 .

 

 

 

 

owmwoa 2:qu mo $328.5 2358““on com 2252 RE”— nvé vim-_-

The final dosage models took into account effects from the Level-1 controlling
variables and Level-2 programmatic factors. Compared to the initial null models, the
pr0portion of variance explained in average dosages was ranged from 40% to 57%. This
indicated that the final models accounted for about 40% to 57% of the initial

between-program variance in dosages (see Table 4-5 for details).

Table 4-5: Variances Explained in the Leisure Dosages Models

 

 

 

NUII MOGCI Final MOCICI % of Variance
Types of Dosages _ _ .
(Total Variance) (Total Variance) Explarned
Overall Leisure Dosage 0.05 0.03 40.0%
Free Play Dosage 0.07 0.03 57.1%
Sports Dosage 0.04 0.02 50.0%
Arts Dosage 0.04 0.02 50.0%
Youth Development Dosage 0.04 0.02 50.0%
Activity Breadth 0.98 0.44 55.1%
Sample 2

Sample 1, the full sample, was used to assess Research Question 1 in order to
include the broadest array of programs possible in evaluating the hypotheses about the
relation between program characteristics and students’ leisure dosage. However, some
sites within Sample 1 did not have all measures needed to examine the research questions
beyond Research Question 1. Therefore, a subsample of the sites in Sample 1 that had
data available to address the subsequent research questions were included in Sample 2
(165 sites out of 182 Sample 1 sites). Therefore, technically speaking, Sample 2 is a
sample of Sample 1, with Sample 1 considered the population. In order to get a better
sense about how the Sample 2 data were distributed and to describe whether and how the
two samples differed, the same procedures for Research Question 1 analyses were

employed using the Sample 2 data. Consistent with the results from the Samplel , the

60

testing of the null models confirmed that there were sufficient variation across programs
and the proceedings of the multilevel analyses were needed for each dosage variable (see
Table 4-6 for details).

Null models. Using the same procedure and criteria for the model selection, a
two-level model was identified for each dosage variable using the Sample 2 data. A
summary of the results for the Sample 2 final null models is presented in Table 4-6. Note
that results fiom the Sample 2 null models are very similar to the Sample 1 null models
presented in Table 4-3. The significance tests indicated that there was significant
variability across programs. The intraclass correlations (ICCs) ranged from 0.47 to 0.73,
meaning roughly 47% to 73% of the total variance in students’ participation dosages was
associated with programs as opposed to individual students (see Table 4-6). Activity
breadth had the most variance, while dosages in youth development had the least variance;
nonetheless, in all cases, substantial variance in students’ dosage of leisure activities was
due to program characteristics. The within-program reliability estimates ranged from 0.93
to 0.97, indicating the estimated differences could be accepted as reliable indicators of

real differences among program’s population means.

Table 4-6: Results from Null Models Predicting Leisure Dosages (Sample 2)

 

 

. 2 .
Outcomes och/$322: ts X value ICC EDESIgtn Reliability
P (df=164) 6° 8
Overall Leisure Dosage 0.05 4381 .24*** 0.51 13.25 0.93
Free Play/Social Dosage 0.08 8562.46*** 0.67 17.31 0.96
Sports Dosage 0.05 3647.56*** 0.47 12.38 0.93
Arts Dosage 0.04 3861 .28*** 0.49 12.93 0.93
Youth Development Dosage 0.03 4454.84*** 0.54 14.09 0.94
Activity Breadth 1.27 10689.61 *** 0.73 18.67 0.97

 

* p <.05;**p <.01;***p <.001.
Note. The design effect is based on average cluster size = 25.21.

61

Unconditional models. Level-1 factors were included as control variables in
examining the effects of Level-2 programmatic factors. The significance tests on Level-1
factors showed some variations between the two samples. Significance test results that
were different from the Sample 1 are bolded in Table 4-8 for reading convenience.
Specifically, in Sample 1, there were gender differences on youth development dosages,
but in Sample 2, there were gender differences in overall leisure activities participation
dosages but not in youth development dosage. For overall leisure activities, females were
found to participate more in overall leisure activities than male students. Similarly,
findings from the Sample 1 data indicated that younger students participated more in free
play activities, while this was not the case for the Sample 2 participants. Lastly, because
Sample 2 consisted of survey participants who had stayed in the programs long enough to
be able to receive and complete the surveys, it was expected that the relationship between
the total days of attendance and participation dosages would vary between the two
samples. In the Sample 2 data, students with greater attendance spent more time in free
play and social activities; the original patterns found between the outcomes and arts and
youth development dosages using the Sample 1 data were no longer obtainable.

Two-Level HLM models. Compared to results fiom Sample 1, the sports dosage
model and activity breadth model remained unchanged at both levels using the Sample 2
data. For the rest of the models, most of the changes occurred in Level 1 as described in
the last paragraph. The only difference in the Level-2 predictors was the attendance
policy for the leisure dosage model. A prelirrrinary test on the inclusion of this variable
using Sample 2 data showed a trend for statistical significance with coefficient value at

004, standard error at 0.02 and df=161. Although this factor was not statistically

62

significant(p=0.06), the effect size of 0.15 was close to the original model supported by
the Sample 1 data (see Table 4-7 for details).

Comparing the initial null models to the final models using the Sample 2 data, the
inclusion of Level-1 controlling variables and Level-2 programmatic predictors helped
explain a considerable proportion of the total variance, ranging from 33% 63%. A

summary of the results is listed in Table 4-8.

Table 4-8: Variances Explained in the Leisure Dosages Models (Sample 2)

 

 

Null MOdCI Final MOdCI % of Variance
Types of Dosages _ _ .
(Total Variance) (Total Vanance) Explained

Overall Leisure Dosage 0.05 0.03 40.0%
Free Play Dosage 0.08 0.03 62.5%
Sports Dosage 0.05 0.02 60.0%
Arts Dosage 0.04 0.02 50.0%
Youth Development Dosage 0.03 0.02 33.3%
Activity Breadth 1.27 0.50 60.6%

 

Again, the reason for utilizing Sample 2 data for Research Question 1 was not to
replace the findings obtained from Sample 1, but rather was to help better understand the
nature of Sample 2 as the sampling product of Sample 1 (the true population) and to
ensure that the two samples were not substantially different when interpreting the Sample
2 results when addressing subsequent research questions. While for the most part the
results of the two samples were similar, some differences in the student—level variables of
gender, grade level and total attendance in relation to certain types of the dosages
emerged and are likely to be linked to the characteristics of the students who were more
likely to complete the student survey as well as the characteristics of sites that ensured

that students received the survey.

63

9.6 N 268... we... 66866.66 666:2 8 2866.6 6.3 .6 66266686 ._66.v 62..” av 6366.0. a...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

66.6 63666 _6_ ...666 .6685 66366..

66.6 66662 62 2:66 66656235 58>

66.6 66.668 62 $.86 32

56 666,62 ~6_ 2.386 6:666

_66 2.266 62 1.666 as 8.6

66.6 66.. _ SN N2 .1666 8663 6826
66:36:96 mx 66 65666860 8:65.» 68am E66636

. . . --------------- M -------------------- .1. E: --------------- l .i i .8666
6N6 666 .NN6 M 8:66:63.
--------------- u EL . . . . . . 66666.60

_66 _66 ...666 ~66 _66 £666 £32m

. . . . . . . . . . . . . . . . . . 6566.5
666 666 ...:-6 N66 666 1.666 666 666 ...666 666 666 ...666 666 666 ..._66 666 666 ...6666ou§.$§m

. . . . . . . . . . . . . . . 666660
N66 666 ..66_ 6.6 666 26 .26 666 26 26 666 666 666 666 N66 ----- ----- 2:63
666 66.6 ...:.~ _66 _66 ...26 66.6 _66 .163 R6 _66 ....-.6 66.6 8.6 1.666 666 _66 .1666 668.35

M N-_o>o.-_

. . . II II II II II It . ............... . . . ............... 8:66:22
moo coo Era—co m mu: :6: 262:. Hogan— 360,—.
--------------- W 2.6 _66 :86- .2- 823
:6 66.6 .666; ----- 6.6 _66 166-. ----- .l !. r: ---------- .33 6680
--------------- 1- l. i 66.6 _66 ...66.6- 6N6 _66 .666 26 66.6 .666 36 _66 «6.6. .6650
6 mm. a W 6 66 6'80 6 w 686 6 6|... 680 M .6. w 680 6 66 Hoolo 364,164

“@6600 owamofl w owmmoQ owamoQ N owmwom owamofl
“ I I . “com-6m .632..—
5635 b.53o< M Each—£059 53> L 3.2. mtg—m _ >65 ooh-m 956$.— .

 

AN 26ng owmmofl 8:664 60 66866on 266888on 8.6 22602 RE”— Ké 6366.-

64

Results for Research Question 2: What Contributes to Leisure Experiences?
Research Questions 2.1, 2.2, and 2.3 have the same null and unconditional models. I
first present the results of the null and unconditional models, and then proceed to
analyses examining each research question.

Null models. Two null models tested whether students’ leisure experiences
(program enjoyment and voluntary participation) varied as a function of site. The results
indicated that there was sufficient variability across programs and that proceeding to
multilevel analyses was warranted. For the program enjoyment model, a variance
component of .08 was obtained, with x2 (164, N=4,160) = 932.58, p<001, an average
within program reliability of .75 and an ICC of .15. This indicated that about 15% of the
total variance in students’ enjoyment was associated with programs. The corresponding
design effect was 4.64, with an average cluster size equaled to 25.21. For the model that
captured students’ voluntary participation, a Bernoulli model was employed because
the dependent variable was dichotomous. A variance component of .40 was obtained,
with x2 (164, N=4,160) = 463.24, p<.001, and an average within program reliability at
0.60. This indicated that there was sufficient variation across programs on students’
voluntary participation and the employment of multilevel analyses were necessitated as
well.

Unconditional models. Level-1 demographic variables and total days of
attendance were introduced into each model prior to the final analyses. Results are
presented in Table 4-9.For the program enjoyment model, students with fewer days of
attendance and younger students reported higher levels of program enjoyment. No

differences were found for gender and race. For the voluntary participation model, male

65

students, younger students, and students with more days of attendance were less likely
than their counterparts to report that they were attending because they wanted to rather
than because someone made them. Race was not associated with voluntary participation.
While the coefficients for total days of attendance were small negative numbers
that rounded up to zero in the both models, the effect sizes were.27 for the enjoyment
and.19 for the voluntary participation model, suggesting that the relationship was not
trivial. In addition, the effects of students’ grade level on the two outcomes were in
different directions; younger students reported enjoying the program more than older

students, but were less likely to report voluntary participation than the older graders.

Table 4-9: Results from Unconditional Models Predicting Leisure Experiences

 

 

 

 

 

Fixed Effect Enjoyment Voluntary participation

Level-l M s_e_ 1 Odds Ratio _Co_ef g r

Intercept 3.00*** 0.03 0.99 2.83 1.04*** 0.08 0.72

Gender ---- ---- -—-- 0.79 -0.24** 0.08 0.22

Grade Level -0.05** 0.01 0.26 1.11 0.10” 0.03 0.23

Total Days of Attendance —0.00** 0.00 0.27 0.99 -0.00* 0.00 0.19

Random Effect Varrance df x2 Reliability
Component

Program Enjoyment 0.08*** 164 931.92 0.74

Voluntary participation 033*" 162 413.09 0.55

 

* p <.05;*"‘p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Research Question 2.]: Are leisure activity dosages related to leisure experiences?
Using the unconditional models identified in the previous step, the analyses
designed to answer Research Question 2.1 then introduced the leisure dosages variables
into the models predicting program enjoyment and voluntary participation. Analyses for
this section were separated by the following types of dosages: (a) overall leisure dosage

(as opposed to academic dosage), (b) the four types of the dosages within leisure

66

activities (after controlling for overall leisure dosage): free play/social time, sports, arts,
youth development; and (c) activity breadth.

Overall leisure dosage. Final model results are presented in Table 4-10. The
analysis of leisure dosage as a predictor of students’ program enjoyment revealed a
positive relationship; the greater the proportion of time spent participating in leisure
activities, the more enjoyment that students reported in the program. Conversely, because
the operationalization of overall leisure dosage was generally characterized as students’
participation in leisure activities as opposed to in academic activities, these results can
also be interpreted as supporting a negative relationship between the extent of
participation in academic activities and enjoyment in the program. Put differently, the
results suggested that students who spent more time in the program participating in
leisure activities reported greater enjoyment than those who spent more time in academic
activities. Similarly, results from voluntary participation model showed the same positive
relationship between leisure dosage and voluntary participation, evidenced by students
with one standard deviation higher than the average in their leisure dosage reporting
roughly twice (odds ratio=2.09) more likely to participate voluntarily than those they had
average leisure dosage.

Leisure activity types and breadth. After accounting for overall leisure dosage, the
four activity types and activity breadth did not significantly predict pro gram enjoyment or
voluntary participation outcomes. A trend approaching significance emerged between
free play/social time dosage and students’ perception of enjoyment, with a coefficient
value of 012, standard error at .06, T-ratio at -1.95 and p-value at .05. This trend

suggests that participation in relaxed leisure (free play/social time activities) may be a

67

factor in influencing student’ overall program enjoyment. This was not included in the
final model for this research question, but was considered in analyses for the next

research questions.

Table 4-10: Results from Overall Leisure Dosage Models Predicting Leisure Experiences

 

l
r

 

 

 

 

Fixed Effect Enjoyment Voluntary participation
Level—1 Co_ef s_e _r Odds Ratio M sg r
Intercept 3.00*** 0.03 0.99 2.86 1.05*** 0.08 0.72
Gender --- --- --- 0.80 -0.22** 0.08 0.20
Grade Level -0.05** 0.01 0.26 1.11 0.10“ 0.03 0.23
Total Days of Attendance —0.00*** 0.00 0.29 1.00 0.00* 0.00 0.20
Leisure Dosage 0.22“ 0.07 0.24 2.09 0.74*** 0.20 0.28
Random Effect Variance Df x2 Reliability
Component
Enjoyment 0.08*** 164 937.62 0.74
Voluntary participation 0.31*** 164 397.56 0.54

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Research Question 2.2: Are students who perceive the program to be of higher
quality more likely to report more positive leisure
experiences?

Students’ perceptions of the program (e.g., staff supportiveness, governance
opportunities, and peer relationships) were introduced into the original models described
previously. The results indicated that all the variables representing students’ perceptions
of the program significantly predicted the two outcomes. Specifically, students’
experiences with staff supportiveness, opportunities for governance, and peer
relationships positively predicted their level of enjoyment in the program and whether
they reported voluntarily participating, while their perceptions of staff injustice were

negatively related to these two outcomes (see Table 4-11).

68

Interestingly, students’ total days of attendance, initially a significant factor for
both outcomes, was no longer significant after taking into account students’ perceptions
of program qualities. This suggests that it was actually not the total days of attendance
that mattered in developing their leisure experiences, but the program qualities that
students perceived during their participation. The same pattern was found between gender
and voluntary participation. At this stage of the analysis, the gender differences that used
to be found in the unconditional model became insignificant. Students’ grade level was
the only demographic variable that remained significant in the models, suggesting that
even after controlling for experiences with staff and peers, older students still reported
lower levels of enjoyment in the program but higher voluntary participation rates than the

younger students.

Table 411: Results from Program Perceptions Models Predicting Leisure Experiences

 

 

 

 

 

Fixed Effect Enjoyment Voluntary Participation
Level-l Odds
C_oe_f at: r L360 QM g r

Intercept 302*" 0.02 0.99 2.67 0.98*** 0.06 0.79
Grade Level -0.04*** 0.01 0.34 1.39 0.13*** 0.04 0.28
Staff Supportiveness 0.19*** 0.02 0.63 0.87 0.33*** 0.07 0.36
Staff Injustice -0.07*** 0.01 0.39 1.38 -0.14** 0.05 0.23
Governance 032*" 0.02 0.82 1.18 0.32*** 0.07 0.32
Peer Relationships 017*“ 0.02 0.60 1.14 0.17" 0.06 0.21
Random Effect Variance df x2 Reliability

Component
Enjoyment 0.03*** 164 522.38 0.62
Voluntary participation 0.3 l *** 164 391.04 0.53

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

69

Research Question 2.3: Do relationships between leisure dosages and leisure
experiences differ as a function of perceived program quality?

To examine whether interaction effects exist between leisure dosages and
program perceptions and to test the effects of leisure dosages and program perceptions
simultaneously, a series of analyses were conducted by introducing the main and
interaction effects between the two sets of the factors onto the outcomes. Results are
presented in Table 4-12.

Overall leisure. For the overgll leisure dosage model, no significant interaction
effect was found between dosage and program perceptions on the two outcomes. All the
main effects of program perceptions remained even after overall leisure dosage was
included, suggesting that greater dosage in leisure activities (in contrast to academic
activity participation) and more positive programmatic experiences such as greater staff
supportiveness, fewer perceptions of staff injustice, more opportunities for governance
and better peer relationships were associated with higher levels of enjoyment and higher
voluntary participation rates. The results could also be interpreted as, holding all the
program satisfaction aspects constant, the more participation in leisure activities still
yielded higher levels of enjoyment and voluntary participation rates compared to
participation in academic activities. Students’ grade level remained significant in the

models.

70

Table 4-12: Results from Program Perceptions and Overall Leisure Dosage Models
Predicting Leisure Experiences

 

 

 

 

 

Fixed Effect Enjoyment Voluntary Participation
Level-1 Odds
9.96.1” as r _Ratio _Co_ef. s_e r

Intercept 3.02*** 0.02 0.99 2.72 1.00*** 0.06 0.80
Grade Level -0.04*** 0.01 0.35 1.39 0.13*** 0.04 0.28
Staff Supportiveness 0.19*** 0.02 0.63 1.37 0.33*** 0.07 0.36
Staff Injustice -0.07*** 0.01 0.40 1.19 -0.14** 0.05 0.24
Governance 0.3 l *** 0.02 0.82 1.14 0.31*** 0.07 0.32
Peer Relationships 017*” 0.02 0.59 2.10 0.17** 0.06 0.21
Leisure Dosage 0.18*** 0.05 0.29 0.87 0.74*** 0.19 0.29
Random Effect CVarrance df x2 Reliability

omponent
Enjoyment 0.03*** 164 510.71 0.61
Voluntary Participation 028*" 164 370.51 0.50

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Leisure activity types. Previous analyses including free play/social time dosage
revealed a trend between this activity type and students’ program enjoyment. However,
that analysis did not include students’ program perceptions. Here, dosage was
re-examined with the inclusion of the program perceptions variables and the interaction
effects between the each program perception scale and fiee play/social time dosage. This
analysis found that one interaction term, between perceived peer relationships and free
play/social time dosage, significantly predicted program enjoyment. Controlling for
students’ grade level, total leisure activity participation, and other program quality
experiences, students with greater peer supports revealed high level of enjoyment
regardless of their free play/social time dosage (see Table 4—13 for details). However, for
students who reported poorer peer relationships, more participation in relaxed leisure
(free play/social time) was significantly related to lower levels of enjoyment in the

program. The interaction results are presented in Figure 4-1. Based on the study design,

71

low dosages of relaxed leisure reflect high dosages of the transitional leisure—that is, arts,
sports, and youth development activities. As a result, the results can also be interpreted as
indicating that students with low peer support in the program reported low program
enjoyment when they participated in relaxed leisure as opposed to transitional leisure. No
other interaction effects for free-play/social time dosage were found related to students’

overall enjoyment and voluntary participation outcomes.

Table 4-13: Results from Program Perceptions and Relaxed Leisure Dosage Model
Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-1 Q)_e_f s_e r
Intercept 3.02*** 0.02 0.99
Grade Level -0.04*** 0.01 0.35
Staff Supportiveness 0.19*** 0.02 0.63
Staff Injustice -0.07*** 0.01 0.40
Governance 0.31*** 0.02 0.81
Peer Relationships (PR) 0. 14*" 0.02 0.44
Leisure Dosage 0.19*** 0.05 0.28
Free Play/Social Time Dosage
(FP) -0.43*** 0.11 0.28
Interaction Term: PR x FP 0.13*** 0.04 0.28
Variance . . .
Random Effect Component x2 Relrabrlrty
Enjoyment 0.03*** 164 489.12 0.60

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

72

 

 

 

 

 

 

 

 

 

 

 

4
3.5 i I ........................ .

E ‘\
4)
E’ 3 “ —0-— Low Peer Support
6
I: ' ' 'l- - - High Peer Support

2.5 -

2 l
Transitional Leisure Relaxed leisure

 

 

 

Figure 4-1: The Interaction Effect between Relaxed Leisure and Peer Support on
Students’ Overall Enjoyment

No interaction effects between dosage of m activities and program
perceptions on the two outcomes. Dosage of M activities did not have a significant
main effect on either of the two outcomes when entered alone. However, after controlling
for students’ experiences with the program, a relatively small but significant positive
effect was found on for arts dosage and program enjoyment (see Table 4-14). No

interaction effects were significant between arts dosage and program qualities in

predicting either of the outcomes.

73

Table 4-14: Results fiom Program Perceptions and Arts Dosage Model Predicting Leisure

 

 

 

 

Experience (Enjoyment)
Fixed Effect Enjoyment
Level-l Coef s; r
Intercept 3.02*** 0.02 0.99
Grade Level -0.04*** 0.01 0.34
Staff Supportiveness 0.19“” 0.02 0.63
Staff Injustice -0.07*** 0.01 0.39
Governance 0.31*** 0.02 0.82
Peer Relationships (PR) O.l7*** 0.02 0.59
Leisure Dosage 0.17" 0.05 0.27
Arts Dosag 0.06* 0.03 0.15

Variance . . .

Random Effect Component x2 Rellability
Enjoment 0.03*** 164 502.41 0.61

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; I = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Similarly, dosage of youth development activities was not a significant factor in
predicting leisure experience outcomes when entered alone. After controlling for program
perceptions, both main and second-order effects were found on youth development
dosage and its interaction with students’ perceptions of opportunities for governance (see
Table 4-15 for details). That is, holding all other factors consistent, students who
perceived greater opportunities for governance reported higher level of enjoyment than
those who didn’t. Students who perceived high level of governance opportunities
reported a slightly lower level of enjoyment as their participation dosage in youth
development activities increased; however, among students who reported fewer
opportunities for governance, enjoyment in the program increased as they participated in
more youth development activities (illustrated in Figure 4-2). The youth development
dosage did not predict voluntary participation still, after controlling for program

perceptions.

74

No main or interaction effects were evidenced for activity breadth in predicting

program enjoyment or voluntary participation after including perceptions of program

quality.

Table 4-15: Results from Program Perceptions and Youth Development Dosage Model
Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-1 (_Cng E r
Intercept 3.02*** 0.02 0.99
Grade Level -0.04*** 0.01 0.35
Staff Supportiveness 0.19*** 0.02 0.63
Staff Injustice -0.07*** 0.01 0.40
Governance (GV) 0.34*** 0.02 0.81
Peer Relationships 0.17*** 0.02 0.59
Leisure Dosage O.18*** 0.05 0.29
Youth Development Dosage (YD) 039* 0.19 0.16
Interaction Term: GV x YD -O. 15* 0.06 0.19
Random Effect Variance x2 Reliability
Commnent
Enjoyment 0.03*** 164 502.83 0.61

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

 

 

 

 

 

 

 

 

 

 

 

4.5
I .........................
.............. .
4 a
g, 3.5 —
is. —o— Low Governance
'E’
El 3“ -------HighGovemance
2.5 —
2 . A
Youth Development Youth Development
(Low Dosage) (High Dosage)

 

 

Figure 4-2: The Interaction Effect between Youth Development Dosage and
Governance Opportunities on Students’ Overall Enjoyment

75

Research Question 2.4: Are program characteristics associated with students’ leisure
experiences?

Previous analyses for research question 2.1 to 2.3 identified a series of factors
related to leisure dosage and program perceptions that were associated with students’
leisure experience outcomes. Because students’ participation and experiences occurred
within the program contexts, it was of interest to further investigate whether and how
program characteristics might also play a role in influencing students’ leisure experiences
in the program. The program characteristics examined in this study included: operating
organization type, leisure activity and type offering, activity breadth offering, academic
participation policy, and utilization of evaluative data for improvement.

To begin the Level-2 model building process, each of the proposed Level-2
factors was tested to assess whether they significantly predicted the intercepts (average)
of the Level-1 baseline leisure experience models (i.e., the models including only
significant demographic and total attendance variables). Level-1 leisure dosages and
program quality perceptions were not included at this stage in order to obtain a baseline
model for the Level-2 effects.

As shown in Table 4—16 of the site-level offerings, the proportion of site-level
leisure activity offerings (in contrast to academic activity offerings) was not significantly
related to students’ perceptions of program enjoyment. However, after controlling for
site-level leisure activity offerings, the more fi'ee play/social time offerings significantly
predicted less program enjoyment. Site-level activity breadth (i.e., the number of
different leisure activity types offered by the site) was also positively related to student’s

level of enjoyment, meaning the greater breadth of the leisure activity offerings by the

76

site, the more likely students were to report their program experiences being enjoyable.
Students’ grade level and total days of attendance remained negatively associated with
the level of program enjoyment after including the program effects. The free-play/social
time and activity breadth offerings were the only two significant Level-2 factors
associated with students’ perception of enjoyment.

The same pattern was found between students’ reports of voluntary participation
and the extent of programs’ free play/social time activity offerings. Although the amount
of leisure activity offerings (in contrast to academic activity) was not associated with
students’ reports of their voluntary participation either, program offerings in free
play/social time activities were negatively associated with students’ voluntary
participation rates after controlling for the overall leisure provision. No other program
characteristic was found to contribute significantly to the two outcomes. The Level-1
significant demographic factors remained unchanged.

Because the study design resulted in fi'ee play/social time activity being the only
type of the leisure activity that captured the “relaxed leisure” construct in the ASP
context, programs with a smaller proportion of this type of the activity actually
represented programs that had a greater emphasis on “transitional leisure” (i.e., sports,
arts and youth development). To illustrate, consider two programs with similar curricula
with regard to how much academic and leisure activity is provided to the students. One
program offers more in relaxed leisure activities (free play and social time), while the
other offers a more structured leisure curriculum based on transitional leisure such as
youth development activities and sports. The results of these analyses suggest that the

program that offered a greater proportion of time for transitional leisure was more likely

77

to have students report enjoyable experiences and voluntary participation than the other
one that provided the more unstructured environment with free play and social time—that
is, the hypothetical “relaxed” leisure.

To summarize, the results suggest that regardless of students’ actual extent of
participation in activities and perceptions of program quality, students tended to favor
programs that offered a stronger curriculum in “transitional leisure” than in “relaxed
leisure” and perceived greater enjoyment when programs offered a greater variety of

leisure activities.

Table 4-16: Results from Program Characteristics Model Predicting Leisure Experiences

 

 

 

 

 

 

Fixed Effect Enjoyment Voluntary Participation
Odds

Level-l Coef _S_e_ 1 Ratio Coef g 1

Gender ---- ---- ---- 0.79 -O.23 ** 0.08 0.21

Grade Level -0.05** 0.01 0.26 1.1 0.10" 0.03 0.22

Total Days of Attendance -0.00** 0.00 0.26 0.99 -0.00* 0.00 0.18

Level-2

Intercept 3.00*** 0.03 0.99 2.83 1.04*** 0.08 0.72

Leisure Activity Offering -O.28 0.17 0.13 0.78 -O.25 0.34 0.06

Free tlay/ 30““ Tune -0.27* 0.10 0.22 0.54 -O.61** 0.22 0.22

Offering

Activity Breadth Offering 0.07* 0.03 0.19 ---- ---- ---- ____

Random Effect Variance df x2 Reliability
Component

Enjoyment 0.07 161 836.22*** 0.72

Voluntary Participation 0.33 162 406.70*** 0.55

 

* p <.05;**p <.Ol;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

78

Research Question 2.5: After accounting for student level factors, do relationships
between leisure dosages and leisure experiences difler as a
function of program characteristics?

The previous analyses have identified several individual and program-level
factors that predict leisure experiences; however, individual and program-level factors
have not been tested simultaneously. The final stage of the analysis was to investigate the
significant predictors in the same models to identify what factors were most important in
predicting students’ leisure experiences in the ASP context—that is, combining the
results of Research Questions 2.1 to 2.4 into a final model to assess whether all factors
continue to have unique contributions or whether some factors drop to nonsignificance.
As addressed in the analytic plan, because the primary focus of the study was to examine
how leisure dosages influence leisure experiences after taking into account students’
demographic differences, perceptions of program quality, and program characteristics,
program-level effects were tested only on the intercept (the average scores of the
outcome variables after controlling for all other factors) and the slope of the dosages in
the Level-1 model (that is, the effects of Level-1 leisure dosages on the outcome
variables). Also, the overall leisure dosage and program-level offering were always
included for controlling purposes when individual types of the dosage/offering were
examined. For analyses that included the different types of leisure dosages (i.e., free
play/social time, sports, arts, and youth development activities), only the corresponding
type of program-level offering would be tested in the model. For example, the proportion
of activities that the program devoted to arts activities were included in the analysis that

examined students’ dosage of arts activities, but were excluded from analyses that

79

examined effects of students’ participation in sports or youth development. The rationale
behind this step was to ensure that the findings found on Level-1 dosages were not due to
the availability of the types of the activities offered at the program level. The type of
program offerings was removed from the final model if not statistically significant.

In addition, previous analyses had shown that students’ activity breadth and
individual types of the dosages were not significant predictors for the voluntary
participation outcome. Only overall dosage of leisure was relevant. The inclusion of the
Level-2 predictors further confirmed the pattern. For these reasons, analyses were not
conducted for the combination of Level-l and Level-2 effects in predicting voluntary
participation except for overall leisure dosage.

For the final overall leisure dosage models, previous results (Table 4-16) showed
that if not considering students-level leisure dosages and perceptions on program quality,
students’ leisure experiences could be positively predicted by pro gram-level activity
breadth (for program enjoyment only) and negatively predicted by relaxed leisure
offerings (for both outcomes), after controlling for site-level total leisure offering and
student-level demographics. If not considering program characteristics at all, overall
leisure dosage, alone with students’ program perceptions and grade level, predicted both
outcomes (Table 4-12). After testing the effects from the both levels simultaneously, it
was found that ahnost all proposed factors remained significantly associated with the
outcomes. The only exception was the program offering in relaxed leisure. After
considering students’ actual participation in leisure activity, the effect of site-level

relaxed leisure provision on students’ perception of program enjoyment became trivially

80

small (but the effect on voluntary participation remained significant) (see Table 4-17). No

other program characteristics were found to significantly predict the outcomes.

Table 4—17: Results from Program Characteristics, Program Perceptions and Overall
Leisure Dosage Model Predicting Leisure Experiences

 

Fixed Effect Enjoyment Voluntary Participation

 

 

 

 

Odds

L. _evel-l M S: I 1% gulf E I
Grade Level -0.04*** 0.01 0.34 1.14 0.13*** 0.03 0.28
Staff Supportiveness 0.19*** 0.02 0.63 1.39 0.33*** 0.07 0.37
Staff Injustice -0.07*** 0.01 0.40 0.86 -0.15** 0.04 0.25
Governance 0.31*** 0.02 0.81 1.36 0.31*** 0.07 0.32
Peer Relationships 017*“ 0.02 0.59 1.18 0.17M 0.06 0.20
Leisure Dosage 0.18** 0.05 0.27 2.28 0.83*** 0.21 0.30
Level-2
Intercept 3.02*** 0.02 1.00 2.74 1.01*** 0.06 0.81
Leisure Activity Offering -0.15 0.11 0.10 0.50 -0.69 0.35 0.15
Free Play/Social Time ,,
Activity Offering ---- —--- —--- 0.59 -0.53 0.21 0.20
ACt’Vlty Breadth 0.04* 0.02 o. 19
Offering
Random Effect Variance df x2 Reliability

Component
Enjoyment 0.03*** 162 495.55 0.61
Voluntary participation 0.28*** 164 370.51 0.48

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

For the final Mylar/social time dosage model, the original Level-1 results
(Table 4-13) indicated that students’ grade level, program perceptions and the interaction
effect between one of the program perception factors (peer relationships) and free
play/social time dosage were predictors of enjoyment. Previous results examining the
contributions of programmatic factors without considering students’ leisure dosages and
program perceptions (Table 4-16) showed that students’ leisure experiences were
positively predicted by Level-2 programmatic activity breadth (for both outcomes) and

negatively predicted by relaxed leisure offering (for enjoyment only), after controlling for

81

site-level total leisure offering and student-level demographics. Because students’ free
play/social time dosage was constrained by what was offered by programs, the Level-2
variable of free play/social time offerings was also entered into the model (intercept) as a
control variable.

The simultaneous test including both levels indicated that the interaction effect
between free play/social time and peer relationships remained significant after controlling
for the Level-2 effects. This indicates that even after including programmatic factors,
students’ peer relationships are associated with the way that students’ participation in
relaxed leisure relates to their overall enjoyment. At Level-2, the programmatic offerings
of free-play/social time were still negatively associated with students’ level of enjoyment.
This indicated that even when students had the same grade level, same dosage of relaxed
leisure, and same peer relationship quality, students enrolled in programs providing more
relaxed leisure still reported lower level of enjoyment than those whose programs offered
more transitional leisure (i.e., sports, arts, and youth development). The activity breadth
that was significant at Level-2 was no longer significant after students’ free play/social
time dosage and program perceptions were taken into account (see Table 4-18 for details).

No other program characteristics were found to significantly predict the outcomes.

82

Table 4-18: Results from Program Characteristics, Program Perceptions and Free
Play/ Social Time Dosage Model Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-l cg: s_e r
Grade Level -0.04*** 0.01 0.36
Staff Supportiveness O. 19*" 0.02 0.63
Staff Injustice -0.07*** 0.01 0.40
Governance 0.31*** 0.02 0.81
Peer Relationships (PR) 0. 14*“ 0.02 0.44
Leisure Dosage 0.18" 0.05 0.27
Free Play/Social Time Dosage
(FP) -0.34** 0.12 0.22
Interaction Term: PR x FP 0.12** 0.04 0.25
Level-2
Intercept 302*" 0.02 0.99
Leisure Activity Offering -0.06 0.12 0.04
Free Play/Social Time Activity
Offering -0.23*** 0.06 0.28
Variance . . .
Random Effect Component x2 Reliability
Enjoyment 0.02*** 162 473.22 0.59

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

For the final sap—rts dosage model, the previous analyses without the inclusion of
Level-2 effects did not reveal any statistically significant relations with either outcome.
The final model tested the Level-2 characteristics that were significant when no Level-l
dosages and program perceptions were included, as well as other programmatic factors
that might help present the effects of the sports dosage. This analysis including Level-2
factors revealed that the ways that sports dosage was associated with students’ program
enjoyment differed by the type of organization operating the program. Specifically, when
students were enrolled in programs run by community non-profit organizations as
opposed to schools, students with a higher dosage in sports reported higher levels of
enjoyment in the pro gram (illustrated in Figure 4-3). Students’ sports dosage revealed

little positive association with the level of enjoyment when students were in the

83

school-operated programs; evidenced by the slope of the sports dosage at 0.01
(0.22-(0.21*1)=0.0l) compared to an average slope at 0.22 for the non-school-operated
programs (see Table 4-19). Activity breadth and other programmatic factors were not

revealed any statistical significance to the outcome.

 

 

 

 

 

—o— CBO-Based Programs
----- -- School-Based Program

 

 

 

Enjoyment
DJ

 

 

 

Sports Dosage =0 Sports Dosage =1

 

 

 

 

Figure 4-3: The Cross-level Interaction Effect between Sports Dosage and
Operating Organization Type on Students’ Overall Enjoyment

84

Table 4-19: Results from Program Characteristics, Program Perceptions and Sports
Dosage Model Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-1 _Co_ef sq r
Intercept 3.02*** 0.02 0.99
Grade Level -0.04*** 0.01 0.35
Staff Supportiveness 0.19*** 0.02 0.63
Staff Injustice -0.07*** 0.01 0.39
Governance 031*" ‘ 0.02 0.82
Peer Relationships 017*“ 0.02 0.59
Leisure Dosage 0.17*** 0.04 0.29
Level-2
Sports Dosage Slope

Intercept 0.22M 0.08 0.21

Progamlype -0.21* 0.09 0.17
Variance . . .

Random Effect Component x2 Reliability
Enjoyment 0.03*** 164 510.90 0.61

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

For the final gfl dosage model, a cross-level interaction effect was found
between Level-1 arts dosage and Level-2 programs’ utilization of evaluative data for
improvement after controlling for demographics and program quality perception factors.
The main Level-1 effect of arts remained significant, meaning greater arts dosage
revealed greater enjoyment in the program. The cross—level interaction effect indicated
the degree to which students’ participation in arts activities positively contributed to their
enjoyment in the program was also moderated by the degree to which programs utilized
their evaluative data (e. g., survey feedback fiom parents and students) for improvement.
Put differently, students’ greater involvement in arts activities, in general, revealed
greater program enjoyment; however, for those who participated in programs that had a

greater emphasis on the use of evaluative data for class improvement, their greater

85

dosages in arts revealed higher levels of enjoyment than those enrolled in programs that

emphasized less the use of data for improvement (see Table 4-20).

Table 4-20: Results from Program Characteristics, Program Perceptions and Arts Dosage
Model Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-l @fi so r
Intercept 3.02*** 0.02 0.99
Grade Level -0.04*** 0.01 0.34
Staff Supportiveness 0.19*** 0.02 0.63
Staff Injustice —0.07*** 0.01 0.39
Governance 0.31*** 0.02 0.82
Peer Relationships 017*" 0.02 0.59
Leisure Dosage 0.17** 0.05 0.27
Level-2
Arts Dosage Slope

Intercept 0.07* 0.03 0.17
Utilization of evaluative data for 0.10,, 0.04 0.18

improvement

Variance . . .

Random Effect Component x2 Relrabrlrty
Enjoyment 0.03*** 164 510.82 0.61

 

* p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Lastly, for the final math development dosage model, the previously found main
and interaction effects on the youth development dosage remained significant after taking
the Level-2 factors into account (see Table 4-15 for the previous result and Figure 4-2 for
illustration). That is, the youth development dosage and its interaction effect with
students’ perception of governance opportunities were still significant factors in
predicting students’ program enjoyment after controlling for all factors (demographics,
program quality perceptions and programmatic factors). In addition, activity breadth
offering at Level-2 also contributed to students’ overall enjoyment after controlling for

the Level-l effects and Level-2 overall leisure activity provision (see Table 4-21).

86

Table 4-21: Results from Program Characteristics, Program Perceptions and Youth
Development Dosage Model Predicting Leisure Experience (Enjoyment)

 

 

 

 

Fixed Effect Enjoyment
Level-l M _S_e ;
Grade Level -0.04*** 0.01 0.35
Staff Supportiveness 0.19*** 0.02 0.63
Staff Injustice -0.07*** 0.01 0.40
Governance (GV) 0.34*** 0.02 0.80
Peer Relationships 017*" 0.02 0.59
Leisure Dosage 0. 18*" 0.05 0.28
Youth Development Dosage (YD) 0.37* O. 19 0.15
Interaction Term: GV x YD -0.14* 0.06 0.19
Level-2
Intercept 3.02"” 0.02 0.99
Leisure Activity Offering -0.15 0.11 0.10
Activity Breadth Offering 004* 0.02 0.19
Random Effect vananc" df x2 Reliability
Component
Enjoyment 0.03*** 162 488.13 0.61

 

*p <.05;**p <.01;***p <.001.
Note. Coef = unstandardized coefficient; se = standard error; r = Cohen’s r effect size. Final model;
nonsignificant predictors removed.

Variance Explained In Final Models

To better illustrate how different sets of the factors (Level-1: leisure dosages and
program perceptions and Level-2: program characteristics) accounted for the total
variances for the two leisure experiences outcomes, the proportion of variance explained
by each step of the analyses for the two outcomes is presented in Table 4-22. It was
notable that the effects of dosage alone were not shown until the inclusion of the
variables representing students’ perceptions of program quality. After students’
perceptions of the program were entered, the four dosage types were able to predict
students’ enjoyment experience. Put differently, in order to examine the true effects of
leisure participation on leisure experiences, students’ perceptions of program quality
needed to be taken into account. After controlling for the perceived program quality,

students’ leisure dosages helped explain 15% of the variance for voluntary participation

87

and 63% for program enjoyment. Level-2 factors only changed the amount of variance

explained for the model assessing the contribution of sports dosage to program enjoyment

(see Table 4-19 for the previous result).

Table 4-22: Variances Explained in Proportion to the Enjoyment Baseline Model

 

g

 

 

 

 

 

 

 

 

l
l

 

‘ Voluntary .
Participation Enjoyment
Overall Overall 3 Youth

Leisure Leisure Free Play Sports Arts lDevelopment

Len/m%m%&a_r.%ygr_/m/
Baseline 0.33 0.08 0.08 0.08 0.08 0.08
Baseline + 0 o i
Dosage, 0-31 (6/o) 0.08 (on)
Baseline + ;
Dosages+ g 0.28 (15%) 0.03 (63%) 0.03 (63%) 0.03 (63%) 0.03 (63%)
Perceptions l
Baseline+ ‘ g i
Dosages+ 0 o o o o o
Perceptions+ 0.28 (l5 /o) 0.03 (75 /o) 0.02 (63 /o)g 0.03 (63 /o) 0.03 (63 /o)g 0.03 (63 A1)
Level-2 i l l

 

 

i

 

Notes. The baseline model refers to the model that had only demographics and total attendance variables at
Level 1 and no Level-2 predictors. Var.= variance; (%) = variance explained in proportion to the
baseline model. Percentages in bold indicate the point at which an increase in variance explained
occurred between steps.

88

Discussion

The purpose of this study was to identify factors that influence students’ leisure

participation and experiences in the 21St CCLC ASP context. Because ASPs can vary

significantly as a result of many factors (e. g. program goals and curricula), programmatic
factors that have the potential to influence leisure participation or experiences were taken
into account. The literature review provided insights into what different aspects of ASP
participation and experiences could be identified as dimensions of leisure, as well as the

individual and programmatic factors associated with students’ leisure in ASPs.

The ASPs examined in this study were part of the 21St CCLC initiative, , which is

well known for its mission to provide academic and other enrichment activities that

reinforce and complement the regular academic program (N afizger et al., 2007). The

. . . . t
strong emphasrs on academics and structured curriculum offerings sets 21S CCLC

programs apart from other ASP models. Because the use of the data for this study was

mainly derived from Michigan 2lst CCLC programs, researchers must be cautious in

generalizing the study results to non-21St CCLC programs.

Two overarching areas were addressed in this study. The first focused on
characteristics of programs that are associated with a greater proportion of leisure, as
opposed to academic, participation by students, as well as how participation in different
types of leisure activities is distributed as a function of program characteristics. The second
targeted leisure activities, both alone and in conjunction with students’ perceptions of the
program and programmatic characteristics, that may be associated with student leisure

experiences—that is, enjoyment in the program and their voluntary desire to attend. In both

89

areas, the results of this study can contribute not only theoretical understanding of ASP
processes, especially for academically oriented ASPs such as 21 st CCLC, but also practical
guidance that has implications for retaining students in such programs.

Research Question I .' What Characteristics Are Associated with Leisure Dosage?

The first research question was intended to identify a number of program-level
characteristics of ASPs that are associated with greater leisure participation. In this study,
participation, or so-called “dosage,” was operationalized as the proportion of time
students spent in leisure activity out of their whole participation across all program
activities. This conceptualization was designed to capture the emphasis of participation.
Although the focus of this question was examining the contributions of

program-level characteristics to leisure dosage, initial analyses examined demographic
and general attendance characteristics associated with individual student dosage that
would form control variables in all analyses. The results indicated that in general, male
students participated more in relaxed leisure (free play) and sports activities, whereas
female students had higher dosages in arts and youth development. The gender pattern in
sports and arts that emerged here confirms other findings (Hirsch, 2005; Lerner, 2005),
while the pattern for relaxed leisure and youth development has not previously been
identified. Students’ grade level was negatively associated with the dosage of relaxed
leisure and arts as well as with participation breadth, indicating that older students
participated in fewer types of leisure activities than their younger peers, especially arts
and free play. As younger students usually require more childcare services, the age

pattern in participation could be a reflection of programs viewing arts and free play

90

activities as more appropriate for the child care needs as well as the interests of younger
students.

The only racial difference found in this study was for students’ participation in
youth development activities, where racial-minority students had a higher dosage than the
white students. This difference was not a result of program offerings; that is, the
programs that served predominately minority youth did not offer more youth
development activities, thereby accounting for the effect. Youth development curricula
include a wide range of activities, such as college preparation, career skills, anger
management, leadership development, and psycho-social skills; these activities have been
identified as of interest to many ASP participants, especially older youth (George, Cusick,
Wasserman, & Gladden, 2007; Lauver & Little, 2005). It might be worth investigating
factors that contribute to the lesser participation in youth development activities by white

students as well as the youth development activities that may be most appealing to this

group in the 21St CCLC ASP context.

At the program level, site policies for students’ attendance in academic classes
unsurprisingly predicted higher dosages of academic activities—and therefore lower
dosages of overall leisure. What was interesting was that it was also positively associated

with the number of different types of leisure activities in which students actually

paiticipated. As 21St CCLC ASPs usually employ a strong academic orientation for their

programming, this pattern may result from the ways that the policies for students’

. . t . . .
academic attendance were desrgned. Annual reports from 21S CCLC srte admmrstrators

that are not part of this study suggest that some programs are designed so that students

91

are able to participate in certain leisure activities as a “reward” for their academic
participation. However, further investigation is needed to clarify this relationship. The
breadth of students’ activity participation was also found to be positively associated with
the length of their program participation, which suggested that differences found between
Sample 1 and Sample 2 in the average breadth of student participation could be a result of
the total attendance. At the program level, a greater variety of offerings—that is, more
programmatic leisure activity breadth—was not related to students’ actual participation
breadth, but was associated with greater leisure dosages in overall leisure participation
and sports. This suggests that in 2lst CCLC ASPs, although programs’ commitment to
offering different types of leisure activities do not necessarily guarantee that students’ are
actually exposed to a greater breadth of leisure activities, students do end up participating
in more leisure activities overall and in more sports activities. This phenomenon may be a
reflection of program design--prograrns may offer different types of leisure activities, but
simultaneously, with the result that actual participation breadth is unaffected, but the
overall dosage of leisure activities, particularly popular choices such as sports, are
increased. The study results also echoed the previous literature that stressed the
popularity of sports (Hirsch, 2005; Lauver & Little, 2005; Oulette, Hutchinson, & Frant,
2005); students who were offered a greater variety of leisure activities to choose from
were likely to spend more time in sports.

Research Question 2: What Characteristics Are Associated with Leisure Experiences?

The second focus of the paper was to identify factors influencing students’ leisure
experiences (program enjoyment and voluntary participation) in the 21St CCLC ASP

context, with a special interest in the dosage of leisure activities that students received.

92

Analyses were followed by the five research sub-questions that helped build the final
models with all significant predictors of students’ leisure experiences from the both
student- and program-levels.

In the initial models, which included only demographic characteristics and
attendance, students’ grade level and total days of attendance were significant predictors
of both outcomes, and students’ gender predicted voluntary participation. In the second
step, students’ perceptions of their experiences in the program with staff, peers, and
opportunities for governance were included, and all demographic and attendance
predictors apart from grade level became nonsignificant. Student’s grade level remained
positively related to voluntary participation and negatively related to students’ reports of
their experiences of program enjoyment. This contradictory effect might be due to the
fact that older students usually were given more authority and were permitted more
choice in determining their after-school time than the younger students. However, even
though older students were more likely to choose to attend, on the whole, they did not
appear to enjoy the programs as much as their younger peers. Conversely, younger
students might be impelled to participate in the program, especially because adults prefer
to ensure that younger children are housed in a safe and structured environment after
school, but nonetheless perceived greater enjoyment at the end. Previous literature has
cautioned about the decrease in ASP participation for older students (McNeal, 1999;

Vandell & Shumow, 1999). Findings from the current study showing older students have

less enjoyable experiences in the 21St CCLC ASPs than younger students and their needs

and preferences might need to be better addressed. As adolescence is a critical period for

the development of autonomy (Eccles, Early, Fraser, Belansky, & McCartby, 1997; Ryan

93

& Deci, 2000), activities that allow older youth to govern, make choices, practice
autonomy and develop meaningful relationships might to help better engage them in the

programs (Caldwell & Darling, 1999; Lauver & Little, 2005; Silbereisen & Eyferth,

1986). Despite the 21St CCLC initiative’s emphasis on academics, losing the interest of

older youth by overemphasizing academics and underemphasizing alternative enrichment
activities could undermine the ability of the programs to help older youth academically
by reducing enrollment and retention. In this case, older youth will not only not have the
opportunity to succeed academically, but also not have the opportunities to develop
positive assets for their successfirl transition into adulthood and be protected against
involvement in risky behaviors.

In the third step, the four program perceptions variables, staff supportiveness,
staff injustice, governance opportunities, and peer relationships, accounted for the
majority of the variance explained for both outcomes. Before the inclusion of these
factors, overall leisure dosage was the only dosage factor that predicted both leisure
experience outcomes, with students who participated in a greater proportion of leisure
(and therefore a lower proportion of academics) reporting more positive experiences. No
significant relationships between the four specific types of the dosages (free play/social
time, sports, arts and youth development) and leisure experience (enjoyment) were
obtained before the inclusion of the program perceptions. Voluntary participation was not
predicted by any of the four leisure activity types, and activity breadth did not predict
either leisure outcome.

The free play/social time dosage could be considered as the “relaxed leisure”

described by Kleiber and his colleagues (1986). Although there have been disputes

94

regarding whether ASPs should emphasize more structured curricula in which students
are given more enrichment opportunities that targeted on learning versus focusing on
providing a safe environment for “relaxed leisure” in which students can rest, play, or
socialize with friends and adults (Hirsch, 2005; Osgood, Anderson, & Shaffer, 2005),
little empirical research (i.e., Mahoney & Stattin, 2000) has examined the impacts of the
two types of leisure involvement. One may expect that students from a structured and
strong academic-focused ASP might prefer the “relaxed leisure” more than “transitional
leisure” during their non-academic time in the program so that they could balance out the
efforts, concentration, or even stress resulted from their academic involvement. However,
in the present study targeting such programs, the results indicated that program offerings
in relaxed leisure were negatively associated with students’ enjoyment. Put differently,
students reported higher level of enjoyment when programs offered more transitional
leisure activities (i.e., sports, arts, and youth development activities) in their
non-academic curriculum, and this was true even after controlling for their perceptions of
the quality of the program and their overall leisure activity participation. The results
suggest that youth favored transitional leisure and anticipated more opportunities to

participate in structured leisure activities rather than just relaxing and socializing with

fiiends even in academically focused programs like 21St CCLC.

At the student level, the relationship between free play/ social time dosage and
leisure experiences (enjoyment) was moderated by peer relationships. For students with
good peer support in the programs, the extent to which they engaged in free play/ social
time dosage was not related to their enjoyment experiences. However, for students with

poorer peer relationships, the greater the time spent in relaxed leisure, the less positive

95

they felt about their participation in the program as a whole, and were therefore less
likely to consider it to be leisure. As friendship is one of the most importance aspects of
adolescent life (Larson & Richards, 1994), the finding suggested the need for program
designs that help facilitate friendships during free play/social time. Furthermore, previous
literature has addressed concerns from parents and teachers with regard to negative
events that can occur during school recess, such as injuries or playground bullying
(Pellegrini & Smith, 1993; Smith, 2000; Smith & Sharp, 1994). As bullying and other
problem behaviors are most likely to occur when adults are not present, the results raised
concerns about the quality of students’ use of less structured leisure time in ASPs (in the
present study, this refers to “relaxed leisure”). Moreover, although studies in the ASP
literature have not addressed bullying events that occur during specific ASP sessions or
recess time (i.e., free play and social time), studies on “peer deviance training” have
included the contexts of ASP as environments in which youth learn deviance behaviors
by interacting with peers with behavioral problems (Henggeler, Schenwald, Borduin,
Rowland, & Cunningham, 1998). As addressed by Gold and Osgood (1992), the essential
question has been whether programs are able to establish positive youth cultures, and
results from the present study suggest that positive youth culture may be especially
important for youth participating in the relaxed leisure activities during their ASP time.
For the remaining activity types, the effect of youth development dosage was
moderated by student’s perceptions of governance opportunities. This indicated that, in
general, students’ perceptions of having governance opportunities was positively related
to their level of enjoyment in the program, and for those who had fewer opportunities for

governance, their enjoyment level increased as they participated in more youth

96

development activities. The results suggest that youth seek for governance and youth

development opportunities in the 21St CCLC ASPs, and the findings echoed previous

literature that addressed the importance of meaningfiilly engaging students in the
program and providing them with leadership and governing opportunities (Lauver &
Little, 2005; Mahoney, Larson, & Eccles, 2005).

Arts dosage was found to be a significant factor in students’ enjoyment
experience after controlling for perceptions of the program quality, with students who
attended more arts activities reporting higher levels of enjoyment in the program. After
considering programmatic factors, it was also found that programs that utilized evaluative
feedback to improve the curriculum quality had students reporting even greater overall
program enjoyment through their participation in arts. This phenomenon was expected to
be shown for all leisure activity classrooms as the importance of utilizing feedback from
the service population to improve program quality has been widely addressed by ASP
researchers (Lauver & Little, 2005; McKenzie & Smeltzer, 2001), even though empirical
studies are still needed to identify the relationship between program’s use of evaluative
data for improvement and student-level outcomes. The fact that the effect was only
presented in analyses of arts participation might be due to inadequacies of measurement;
the evaluative feedback variable was self-reported by administrators—however, it may be
that arts activities provide the most opportunity for creative change or are most frequently

examined as part of an evaluative plan.

The fourth step revealed that the type of organization that operated the 21St CCLC

program was not associated with more participation in leisure activities as a whole, but

did moderate the relationship between the sports dosage and enjoyment. Specifically,

97

when programs were run by school-based institutions, the extent to which students
participated in sports was unrelated to their enjoyment in the program. However, in
programs that were administered by CBOs and NGOs, students with greater participation
in sports reported significantly more enjoyment in the programs. No previous literature is
available that looks at the degree to which different types of organizations excel in
providing different types of activities. Moreover, in practice, many of the school-run 2lst
CCLC programs hired outside service providers to lead activities, especially for
non-academic subjects. Nonetheless, the results suggest that CBOs and NGOs are indeed
better at providing sports activities within the context of their own programs. The result
might be due to the fact that within a school-administered program, sports are more likely
to be treated similarly to free play, whereas CBOs and NGOs may have a long history of
providing sports within a youth development context, so that the activity might be truly
transitional leisure with skill-building components that permit students to set and achieve

goals and gain new knowledge. This could be an area that is ripe for further exploration,
especially within the 21St CCLC ASP context.

Last but not least, previous studies have found that students’ participation breadth
can motivate them to participate in ASPs. Participation breadth has also been shown to be
related to positive outcomes such as: academic achievement (Baker & Witt, 1996; Gerber,

1996; Marsh, 1992), higher life satisfaction (Gilrnan, 2001), and positive attitudes toward

their plans for education, college admission, and future occupation (Marsh, 1992; Marsh

& Kleitman, 2002). However, in the present study focusing on 21St CCLC ASPs, activity

breadth was not an individual- level predictor of leisure experiences; rather, it was a

significant program-level factor. The number of different types of leisure activity offered

98

to youth was not only positively associated with students’ dosages in overall leisure
participation and sports (which has been discussed in the previous section for results on
Research Question 1), but also contributed to their overall enjoyment of the program. The
later finding echoed the previous literature that suggested programs should provide a
wide variety of activities to sustain and help engage youth participation in ASPs (Lauver

& Little, 2005; Mahoney, Larson, & Eccles, 2005); and results from the present study

further confirmed the same pattern even within academically focused 21St CCLC ASPs .

In general, most of the programmatic factors did not moderate relationships
between leisure participation and leisure experiences, and the bulk of their contribution
was in predicting the extent to which students participated in leisure to a greater or lesser
extent. For example, attendance policies for academic activity participation predicted the
extent to which students participated in leisure as a whole and in different types of
activities, but not the extent to which they enjoyed the program or whether they wanted to

attend. It is likely that a balance is needed in ensuring that students participate in both

academic and le1sure act1V1t1es. If desrgned and implemented properly, 21S CCLC
program planners may use academic attendance policies to ensure students’ academic
involvement, while in the meantime allowing students to have the space and

opportunities to participate in leisure activities—particularly transitional leisure

activities—-to encourage them to come to the program.

99

Limitations
There are some limitations of this study that are noteworthy. Overall, the study

was constrained by the the availability and operationalization of data implemented as part

of the Michigan evaluation of 21St CCLC. Most of the variables used in this study were

self—report measures, which may be biased compared to, for example, outside ratings.
Also note that this study was not conducted through a randomized design, limiting the
ability to make causal attributions based on the study results.

Furthermore, while in this study, program quality perceptions were the most
influential predictors of leisure experiences, evaluation of program quality domains was
limited by the availability of the quantitative data reported by participating students. The
four program quality perception measures used in the present study represented the best
use of the available data, but it was by no means meant to be the most comprehensive
checklist for assessing students’ program quality perceptions and experiences. For
example, students’ perception of having opportunities to be adequately challenged, to
learn new things, or to practice governance at different levels of programming might also
be important indicators of their perceptions of program quality.

Also, because the dosage computation was based on activity types that were
categorized by coders based on written information provided by program staff, there is
likely to be some error in the coding introduced by poor descriptions. For example, some
program staff only reported the activity name without giving the full description and
objective of the activity, and some staff were not skilled in providing useful descriptions

and objectives.

100

Another issue is the conceptualization of the leisure outcomes. While the two
leisure experience outcomes can legitimately be considered ways to think of
leisure--students’ active desire to attend the program reflects choice, self-determination,
and the consequence of uncoerced behavior, and program enjoyment represents students’
pleasure with the activities and engagement in the program as a whole--they do not
capture the multidimensional aspects of leisure. Given the constraints on data availability,
other leisure experiences such as flow (Csikszentrnihihalyi, 1990), awareness, boredom,
challenge, and distress/anxiety (Leisure Experience Battery for Adolescents; Caldwell,
Smith, & Weissinger, 1992) were not addressed in this study but could be important
aspects for understanding youth’s leisure experiences in ASP contexts.

As to the voluntary participation outcome, the study results indicated fewer
relationships between leisure participation and this outcome than the other leisure
experience outcome, program enjoyment. This might be due to the fact that the voluntary
participation outcome was a single-question dichotomous variable that provided less
variation than the enjoyment outcome, which was on a continuous scale. Note that the
original responses for the question “What is the MOST IMPORTANT reason that you
come to this program?” were: “I want to come,” “My parents want me to come,” and “A
teacher, principle, or counselor wants me to come.” Improvements can be made in the
future by designing this measure into a continuous scale or incorporating more questions
_ in the measure.

In addition, the two experience outcomes were measured globally; they addressed
students’ overall leisure experiences through their involvement in the whole program but

not immediate experiences resulting from specific activities. In other words, students’

101

participation in specific types of activities was presumed to relate more or less as a
fimction of dosage in contributing to their overall perceptions of the program. Although
this approach allows an exploration of the effect of leisure activity participation on ASP
leisure experiences over time, some other existing measures (e. g. Youth Pro gram Quality
Assessment; High/Scope Educational Research Foundation, 2007) and study designs (e. g.
Experience Sampling Method; see Larson, Hansen, & Moneta, 2006; Shemoff & Vandell,
2007; Vandell et al., 2005) would enable a detailed look at the specific processes that
promote positive leisure experiences through activity participation. Development and
validation of appropriate measures for examining leisure processes and experience is a
fruitfirl avenue for future research.

Finally, given the fact that most of the program-level characteristics were not
significant predictors of the outcomes, ASP characteristics that might influence
participation or experiences may have not been fiilly examined. The present study might
have overlooked some critical program-level factors given the data availability and/or the
study design, such as staff job satisfaction, turnover rates, staff and student ratios, staff
credentials, and how many years the program had been operating. Moreover, the
programmatic measures used here relied on administrator report, and might be more
effective predictors if site observations were implemented. Future studies need to further
investigate the critical programmatic factors that influence students’ program

participation and experiences.

102

Recommendations for Future Research

As the current political climate has placed high expectations on 21St CCLC

programs to produce favorable academic outcomes, the value of offering leisure and

recreation activities in 21St CCLC ASPs seemed degraded by many. With the intention to

highlight the role of leisure and recreation in encouraging ASP participation, especially

for programs like 21$t CCLC that have a strong academic emphasis but are not solely

devoted to academics, this study examined different types of the leisure participation and

then assocrations wrth students’ perceptions of lersure experiences in the 21S CCL ASP

contexts. In addition to the future directions described above designed to address the
limitations of the current study, other areas for future research are apparent based on the
literature review and the study results.

First, firture studies need to address additional outcomes beyond leisure
experiences. These might include academic performance or youth development outcomes
such as resilience, social-emotional competence and youth initiative (Darling, Caldwell,
& Smith, 2005; Dworkin, Larson, & Hansen, 2003; Larson, Hansen, & Walker, 2005).
Secondly, future studies might look in-depth at how both relaxed and transitional
activities play out in the interactional context. Measures of staff-student interaction at the
point of service—that is, with the context of a specific activity—are enabling studies
about the specific processes that encourage or discourage positive youth development.
For example, the Youth Program Quality Assessment (PQA) (High/Scope Educational
Research Foundation, 2007) assesses the ways that staff facilitate student engagement

and learning in the areas of support, engagement, and interaction to develop higher-level

103

processes such as problem—solving and critical thinking, and this measure is especially
in measuring interaction during transitional leisure activities. Well-designed studies that
look at the process of the program impacts on youth are highly needed.

Furthermore, the utilization of participation “proportion” as the dosage was a new
way of capturing participation that has not been previously implemented. This strategy
was designed to present the emphasis of participation in different types of the activities
from the perspective of what youth experience. Our previous work has shown that
whereas the actual number of hours or days of participation in ASPs is associated with
better academic outcomes, the proportion of time that youth spend in different activities
is related to how students feel about the program (Van Egeren, Wu, & Reed, 2007).
When the majority of their time is spent in homework help, they are far less likely to feel
positively about the program than when the majority of their time is spent in the type of
activities that represent transitional leisure. Future studies might want to examine how
different ways of calculating leisure dosage (absolute participation, participation length,
breadth, and proportions) might best present the relationship between participation and
different leisure outcomes.

In addition, many activities are comprised of multiple activity types (say,
homework help plus mentoring or arts and academic enrichment); it would be interesting

to see whether these integrated activities result in different experiences or youth

outcomes than activities that are single-focused. As 21St CCLC funded ASPs are required

to employ a comprehensive curriculum and expected to produce improved academic
outcomes (N aftzger et al., 2007), the provision of the empirical studies on the relationship

between different types of the activity participation and the associated outcomes, as well

104

as how to provide a balanced curriculum, could help provide a foundation for 21St CCLC

program planners, Parks and Recreation professionals, as well as other ASP practitioners
to maximize the results of their work.
Recommendations for ASP Planners

The study results about different types of participation in leisure activities

and their relations wrth lelsure experiences may provrde some rnsrghts for 218 CCLC

program planners to consider. First, the results of this study suggest that students tend

to have greater enjoyment and a stronger desire to participate in 21St CCLC programs

when they spend more time in leisure (especially in sports) activity and/or when the
programs they participate in offer a greater variety of leisure activities. These results
support the use of leisure activities as incentives for greater motivation in

participation. As no program benefits can be conveyed to students unless they

actually participate (V andell & Shumow, 1999), 21St CCLC program planners need to

consider how to provide a variety of leisure activities and to best utilize them to
encourage academic activity participation and to sustain students’ overall program
involvement.

Secondly, because the study results indicated that students reported greater
enjoyment when they were given the opportunity for and actually participated more in

transitional leisure activities such as sports, arts, and youth development as opposed to

relaxed leisure (e.g., free play and social activities), 21St CCLC program planners need to

make sure that these activities are made available to them. The study results also showed

that male and younger students participated less than their counterparts in transitional

105

leisure. How to ensure that their program experiences are as enjoyable as the experiences
of other students can be a real challenge. In addition, given the fact that students’
participation in relaxed leisure was found to contribute to the overall enjoyment only
when students had good peer relationships in the program, program planners might need
to help structure the social and free-play time activity to make it a pleasant experience for
all participants, especially those who are new to the program or have fewer friends. In
fact, a series of analyses from the present study revealed a consistent pattern in that
students reported lower level of enjoyment and less desire for voluntary participation
when their programs offered a greater amount of time in relaxed leisure. Previous
literature cautioned that the provision of unstructured activities in youth centers might be
an indicator of low program quality and participation in unstructured activities might
increase, rather than decrease youth problem behaviors (Mahoney & Stattin, 2000;
Osgood, Anderson, & Shaffer, 2005). As some other researchers may argue that youth
need relaxing and safe ASP environments to de-stress life adversity, facilitate creativity,

explore possibilities, and to freely express themselves (Hirsch, 2005; Sutton-Smith, 1994),

findings from the present study may raise a red flag for 21St CCLC staff to pay closer

attention to prevent negative events occurred during the free-play and social time, and to
ensure the provision of high quality relaxed leisure time for all participants.

Third, the results showed that sports and arts activities were of interest to many

21St CCLC ASP participants, with females and younger students participating more in

arts and males more in sports. Although this might reflect the true preference of the

service population, program planners might want to consult with students who participate

106

less in each and ensure that their best interests were taken into considerations by the
curriculum design.

Moreover, participation in youth development activities was higher for the racial
minority group than the white students. This phenomenon was true even after taking into
account the extent to which youth development activities were offered by programs.
Although this might enhance minority youth’s resources and help them overcome life
challenges, program administrators might need to be attentive about how to include youth

of all racial backgrounds into the youth development classroom so that the benefits can

be conveyed to all. Also, 21St CCLC participants favored youth development activities

and seek for opportunities for governance in the program. Program planners might need

to employ strategies such as youth council/committee to meaningfully engage youth and

sustain their participation in 21St CCLC ASPs.

In a similar vein, older students reported lower levels of enjoyment than younger
students, although their participation was more likely to be based on their own choice. As
previous literature has documented the difficulty in retaining older students in the ASPs
(McNeal, 1999; Vandell & Shumow, 1999), program planners need to make extra efforts
to find out what activities are most relevant and appealing to them.

Nevertheless, students with higher days of attendance were more likely to report
that they did not choose to attend and had lower levels of program enjoyment than
students who attended fewer days. This seemed the reverse of what one would expect.
However, after taking into account the students’ perceptions of the program quality, this
pattern was no longer visible. The four program quality domains were shown to be very

relevant to ensure students’ leisure experiences. This suggests that program quality rather

107

than the length of attendance that predicts students’ leisure experiences in ASPs. Program
planners should ensure that students’ experience with staff, peers, and opportunities to
build skills and find meaning through governance activities are maximized to the fullest
extent to ensure students’ enjoyment and help maintain their active desire for

participation.

108

References

Baker, D. A., & Witt, P. A. (1996). Evaluation of two after-school recreational programs.
Journal of Park and Recreation Administration (submitted), Fall.

Baldwin, C. K., Caldwell, L. L., & Witt, P. A. (2005). Deliberate programming with logic
models: From theory to outcomes. In P. A. Witt & L. L. Caldwell (Eds),
Recreation and youth development (pp. 219-239). State College, Pennsylvania:
Venture Publishing, Inc.

Barber, B. L., Eccles, J. S., & Stone, M. R. (2001). Whatever happened to the Jock, the
Brain, and the Princes? Young adult pathways linked to adolescent activity
involvement and social identity. Journal of Adolescent Research, 16, 429-455.

Belle, D. (1999). The after-school lives of children: Alone and with others while parents
work. Mahwah, NJ: Lawrence Erlbaum Associate, Inc.

Blumer, H. (1969). Symbolic interactionism: Perspective and method. Englewood
Cliffs, NJ: Prentice Hall.

Borden, L. M., Perkins, D. F., Villarruel, F. A., Carleton-Hug, A., Stone, M. R., & Keith,
J. G. (2006). Challenges and opportunities to Latino youth Development:
Increasing meaningful participation in youth development programs. Hispanic
Journal of Behavioral Sciences, 28(2), 187 — 208.

Brown, B. B., & Theobald, W. (1998). Learning contexts beyond the classroom:
Extracurricular activities, community organizations, and peer groups. In K.
Borman & B. Schneider (Eds), The adolescent years .' Social influences and
educational challenges: Ninety-seventh yearbook of the National Society for the
Study of Education, Part I (pp. 109-141). Chicago, IL: The National Society for
the Study of Education.

Caldwell, L. L., Baldwin, C. K., & Walls, T. S., E. A. (2004). Preliminary effects of a
leisure education program to promote healthy use of free time among middle
school adolescents. Journal of Leisure Research, 36(3), 310-335.

Caldwell, L. L., & Darling, N. (1999). Leisure context, parental control, and resistance to
peer pressure as predictors of adolescent partying and substance use: An
ecological perspective. Journal of Leisure Research, 31, 57-77

Caldwell, L. L., & Smith, E. A. (2006). Leisure as a context for youth development and

delinquency prevention. The Australian and New Zealand Journal of Criminology,
39(3), 398-418.

Caldwell, L. L., Smith, E. A., & Weissinger, E. (1992). Development of a Leisure

Experience Battery for Adolescents: Parsimony, stability, and validity. Journal of
Leisure Research, 24(4), 361-376.

109

Calvert, M., Zeldin, S., & Weisenbach, A. (2002). Youth involvement for community,
organizational and youth development: Directions for research, evaluation and
practice. Madison: University of Wisconsin-Madison.

Carlin, J ., & Hocking, J. (1999). Design of cross-sectional surveys using cluster sampling:
an overview with Australian case studies. Aust N Z J Public Health, 23(5),
546-551.

Carnegie Council on Adolescent Development. (1992). A matter of time: Risk and

opportunity in the nonschool hours. New York: Carnegie Corporation of New
York.

Chang-Rios, K. (2007). The relationship between afterschool participation and academic
achievement: Does intensity of participation matter? Unpublished Dissertation,
University of Kansas, Lawrence, Kansas.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, NJ: Lawrence Erlbaum.

Colwell, M. J ., Pettit, G. S., Meece, D., Bates, J. E., & Dodge, K. A. (2001). Cumulative
risk and continuity in nonparental care from infancy to early adolescence.
Merrill-Palmer Quarterly, 47(2), 207-234.

Connell, J. P., & Wellbom, J. G. (1991). Competence, autonomy, and relatedness: A
motivational analysis of self-system processes. In M. R. Gunnar & L. A. Sroufe
(Eds), Self processes in development: The Minnesota symposium on child
psychology (Vol. 23, pp. 43—77). Hillsdale, NJ: Lawrence Erlbaum Associates.

Csikszentrnihalyi, M., & Larson, R. (1984). Being adolescent: Conflict and growth in the
teenage years. New York: Basic Books.

Csikszentmihalyi, M., Larson, R., & Prescott, S. (1977). The ecology of adolescent
activity and experience. Journal of Youth and Adolescence, 6, 281-294.

Csikszentrnihihalyi, M. (1990). Enjoyment and the Quality of Life, The Conditions of
Flow. In Flow: The Psychology of Optimal Performance (pp. 43-??). New York:
Harper 7 Row Publishers.

Darling, N., Caldwell, L. L., & Smith, R. (2005). Participation in school-based
extracurricular activities and adolescent adjustment. Journal of Leisure Research,
37(1), 51-76.

Davalos, D. B., Chavez, E. L., & Guardiola, R. J. (1999). The effects of extracurricular
activity, ethnic identification, and perception of school on student dropout rates.
Hispanic Journal of Behavioral Sciences,, 21(1), 61-77.

deGrazia, S. (1962). Of time, work, and leisure. New York: Twentieth Century Fund.

110

Devereux, E. (1976). Backyard versus Little League: The impoverishment of children's
games. In D. Landers (Ed.), Social problems in sport (pp. 37-56). Urbana:
University of Illinois Press.

Dumazedier. The Chips Are Not Down (S. E. McClure, Trans). In Toward A Society of
Leisure (pp. 3-32). New York: The Free Press.

Durlak, J. A., & Weissberg, R. P. (2007). The impacts of after-school programs that
promote personal and social skills. Chicago, IL: Collaborative for Academic,
Social, and Emotional Learning.

Dworkin, J. B., Larson, R., & Hansen, D. (2003). Adolescents' accounts of growth
experiences in youth activities. Journal of Youth and Adolescence, 52(1), 17-26

Eccles, J ., & Templeton, J. (2002). Extracurricular and other after-school activities for
youth. Review of Research in Education, 26, 113-180.

Eccles, J. S. (2005). The present and future of research on activity settings as
developmental contexts. In J. L. Mahoney, R. W. Larson & J. S. Eccles (Eds),
Organized activities as contexts of development (pp. 353-371). New Jersey:
Lawrence Erlbaum Associates.

Eccles, J. S., & Barber, B. L. (1999). Student council, volunteering, basketball, or
marching band: What kind of extracurricular involvement matters? Journal of
Adolescent Research, 14, 10-43.

Eccles, J. S., Early, D., Fraser, K., Belansky, E., & McCartby, K. (1997). The relation of
connection, regulation, and support for autonomy to adolescents' functioning.
Journal of Adolescent Research, 12, 263-286.

Eccles, J. S., & Gootrnan, J. A. (2002). Community programs to promote youth
development. Washington, DC: National Academy Press.

Edginton, C. R., Jordan, D. J ., DeGraaf, D. G., & Edginton, S. R. (2002). Leisure and life
satisfaction: Foundational perspectives (Third ed.). New York, NY:
McGraw-Hill.

Elley, C. R., Kerse, N., Chondros, P., & Robinson, E. (2005). Intraclass correlation
coefficients from three cluster randomised controlled trials in primary and

residential health care. Australian and New Zealand Journal of Public Health,
29(5), 461-467.

Fiester, L. M., Simpkins, S. D., & Bouffard, S. M. (2005). Present and accounted for:
Measuring attendance in out-of-school-time programs. New Directions for
Youth Development, 105, 91-107.

Fox, K. M. (2000). Echoes of leisure: Questions, challenges, and potentials Journal of
Leisure Research, 32(1), 32-36.

111

Fredricks, J. A., Alfred-Liro, C. J ., Hruda, L. 2., Eccles, J. S., Patrick, H., & Ruan, A. M.
(2002). A qualitative exploration of adolescents' commitment to athletics and the
arts. Journal of Adolescent Research, I 7, 68-97.

Gayl, C. L. (2004). After—school programs: Expanding access and ensuring quality:
Progressive Policy Institute.

George, R., Cusick, G. R., Wasserman, M., & Gladden, R. M. (2007). After-school
programs and academic Impact: A study of Chicago 's After School Matters.
Chicago: Chapin Hall Center for Children.

Gerber, S. B. (1996). Extracurricular activities and academic achievement. Journal of
Research and Development in Education, 30, 42—50.

Gilman, R. (2001). The relationship between life satisfaction, social interest, and
frequency of extracurricular activities among adolescent students. Journal of
Youth and Adolescence, 30, 749—767.

Goerge, R., Cusick, G. R., Wasserman, M., & Gladden, R. M. (2007). After-school
programs and academic Impact: A study of Chicago ’s After School Matters.
Chicago: Chapin Hall Center for Children.

Gold, M., & Osgood, D. W. (1992). Personality and peer influence in juvenile
corrections. Westport, CT: Greenwood Press.

Gordon, W. R., & Caltabiano, M. L. (1996). Urban-rural differences in adolescent
self-esteem, leisure boredom, and sensation-seeking as predictors of leisure-time
usage and satisfaction. Adolescence, 3 I (124), 883-902.

Guinn, B., Semper, T., & Jorgensen, L. (1996). Perceived leisure freedom and activity
involvement among Mexican American adolescents. Hispanic Journal of
Behavioral Sciences, 18(356-366).

Halpem, R. (1999). After-school programs for low-income children: Promise and
challenges. The Future of Children, 9(2), 81-95.

Halpem, R. (2002). A different kind of child development institution: The history of

after-school programs for low-income children. Teachers College Record, 101(2),
178-211.

Halpem, R. (2003). Making play work: The promise of after-school programs for
low-income children. New York: Teachers College Press.

Hamilton-Smith, E. (1991). The construction of leisure. In B. L. Driver, P. J. Brown & G.
L. Peterson (Eds), Benefits of leisure. State College, PA: Venture.

112

Hansen, D., Larson, R., & Dworking, J. (2003). What adolescents learn in organized
youth activities: A survey of self-reported developmental experiences. Journal of
Research on Adolescence, 13, 25-55.

Hazel, K. L., Langenau, E. E., & Levine, R. L. (1990). Dimensions of hunting
satisfaction: Multiple-satisfactions of wild turkey hunting. Leisure Sciences, 12(4),
383-393.

Henggeler, S., Schenwald, S., Borduin, C., Rowland, M., & Cunningham, P. (1998).
Multisystemic treatment of antisocial behavior in children and adolescents. New
York: Guilford.

High/ Scope Educational Research Foundation. (2007). The youth program quality
assessment: A research-validated instrument and comprehensive system for
accountability, evaluation and program improvement.

Hirsch, B. J. (2005). A place to call home: After-school programs for urban youth.
Washington, DC: American Psychological Association.

Hultsman, J. T., & Kaufinan, J. E. (1990). The experience of leisure by youth in a
therapeutic milieu. Youth & Society, 21, 496-510

Innovation by Design and Center for Teen Empowerment. (2000). Afterschool programs
in Boston: What young people think and want. Boston, MA: Center for Teen
Empowerment.

Iso-Ahola, S. (1980). The social psychology of leisure and recreation. Dubuque, IA:
William C. Brown.

Iso-Ahola, S. (1999). Motivational foundations of leisure. In E. L. Jackson & T. L.
Burton (Eds), Leisure studies: Prospects for the twenty-first century (pp. 35—51).
State College, PA: Venture.

J ames-Burdumy, S., Dynarski, M., Moore, M., Deke, J ., Mansfield, W., & Pistorino, C.
(2005). When schools stay open late: The national evaluation of the 21st Century
Community Learning Centers Program: Final Report. Retrieved 09/01/2007,
from http://www.ed.gov/ics/ncee.

J uniu, S. (2000). Downshifting: Regaining the essence of leisure. Journal of Leisure
Research, 32(1), 69-73.

J uniu, S., & Henderson, K. (2001). Problems in researching leisure and women: Global
considerations. World Leisure Journal, 43(4), 3-10.

Kirshner, B., O'Doonoghue, J. L., & McLaughlin, M. W. (2002). Youth participation:
Improving institutions and communities. In B. Kirshner, J. L. O'Doonoghue & M.
W. McLaughlin (Eds), New Direction for Youth Development (V 01. 96, pp. 5-8).
San Francisco, CA: Jossey-Bass.

113

Kleiber, D., Larson, R., & Csikszentmihalyi, M. (1986). The experience of leisure in
adolescence. Journal of Leisure Research, 18, 169-176.

Kleiber, D. A. (1999). Leisure experience and human development. New York: Basic
Books.

Kleiner, B., Nolin, M. J ., & Chapman, C. (2004). Before- and after-school care,
programs, and activities of children in kindergarten through eighth grade: 2001.
National Center for Educational Statistics analysis report. Washington, DC:
National Center for Educational Statistics.

Kraus, R. (1990). Recreation and Leisure in Modern Society (Forth ed.). New York:
HarperCollins.

Larson, R. (2000). Toward a psychology of positive youth development. American
Psychologist, 55, 170—183.

Larson, R., Hansen, D., & Walker, C. (2005). Everybody's gotta give: Development of
initiative and teamwork within a youth program. In J. Mahoney, R. Larson & J.
Eccles (Eds), Organized activities as contexts of development: Extracurricular
activities, after-school and community programs (pp. 159-184). Mahwah, NJ:
Lawrence Erlbaum.

Larson, R., & Richards, M. H. (1994). Divergent realities: The emotional lives of mothers,
fathers, and adolescents. New York: Basic Books.

Larson, R. W., & Verma, S. (1999). How children and adolescents spend time across the
world: Work, play and developmental opportunities. Psychological Bullentin,
125(6), 701-736.

Lauer, P. A., Akiba, M., Wilkerson, S. B., Apthorp, H. S., Snow, D., & Martin-Glenn, M.
L. (2006). Out-of-school-time programs: A meta-analysis of effects for at-risk
students. review of Educational Research, 76(2), 275-313.

Lauver, S., & Little, P. (2005). Finding the right hook. The School Administrator, 5(62),
27-30.

Lerner, R. M. (2005). Promoting positive youth development: Theoretical and empirical
bases. Washington, DC: Institute for Applied Research in Youth Development.

Luke, D. A. (2004). Multilevel modeling. Thousand Oaks, CA: Sage Publications.

Mahoney, J ., Larson, R., & Eccles, J. (2005). Organized activities as contexts of
development: Extracurricular activities, after-school and community programs.
Hillsdale, NJ: Lawrence Erlbaum Associates.

Mahoney, J ., Larson, R., Eccles, J. S., & Lord, H. (2005). Organized activities as
developmental contexts for children and adolescents. In J. Mahoney, R. Larson &

114

J. S. Eccles (Eds), Organized activities as contexts of development (pp. 3-22).
Mahwah, New Jersey: Lawrence Erlbaum Associates.

Mahoney, J. L., Lord, H., & Carryl, E. (2005). An ecological analysis of after-school
program participation and the development of academic performance and

motivational attributes for disadvantaged children. Child Development, 76,
811-825.

Mahoney, J. L., & Stattin, H. (2000). Leisure activities and adolescent antisocial behavior:
The role of structure and social context. Journal of Adolescence, 23(113-127).

Marsh, H. W. (1992). Extracurricular activities: Beneficial extension of the traditional

curriculum or subversion of academic goals? Journal of Educational Psychology,
84, 553—562. ’

Marsh, H. W., & Kleitman, S. (2002). Extracurricular school activities: The good, the‘bad,
and the nonlinear. Educational Review, 72, 464~514.

McCallion, G. (2003). 21 st Century Community Learning Centers in PL. 107-110:
Background and funding. Washington, DC: Congressional Research Service.

McCartney, K., & Rosenthal, R. (2000). Effect size, practical importance, and social
policy for children. Child Development, 71(1), 173-180.

McGillis, D. (1996). Beacons of hope: New York City's school-based community centers.
NIJ Program Focus.

McKenzie, J. F ., & Smeltzer, J. L. (2001). Planning, implementing, and evaluating health
promotion programs (3rd ed.). Boston: Allyn & Bacon.

McLaughlin, M., Irby, M., & Langrnan, J. (1994). Urban sanctuaries: Neighborhood
organizations in the lives and future of inner city youth. San Francisco:
J ossey—Bass.

McNeal, R. B. (1999). Participation in high school extracurricular activities: Investigating
school effects. Social Science Quarterly, 80(2), 291-309.

Meyer, H. D., & Brightbill, CK. (1964). Community recreation: A guide to its
organization (3rd Ed. ). Englewood Cliffs: Prentice-Hall, Inc.

Michigan let Century Community Learning Centers. (2008). 21st Century Community
Learning Centers grant program criteria and priority. Lansing, MI: Michigan
Department of Education.

Mobily, K. E. (1989). Meaning of recreation and leisure among adolescents. Leisure
Studies, 8(1), 11-23.

115

Murphy, E. (2000). Risk, responsibility, and rhetoric in infant feeding. Journal of
Contemporary Ethnography, 29(3), 291-325.

Muthen, L. K., & Muthen, B. (1999). Intraclass correlations. Retrieved April, 2008,
from http://wwwstatmodel.com/discussion/messages/12/1 8.html? 1206899092

Naftzger, N., Bonney, C., Donahue, T., Hutchinson, C., Margolin, J ., & Vinson, M.
(2007). let Century Community Learning Centers (21st CCLC) analytic support
for evaluation and program monitoring: An overview of the 21 st CCLC
performance data: 2005-06. Naperville, IL: Learning Point Associates.

National Institute on Out-of-School Time. (2007). Making the case: A fact sheet on
children and youth in out-o -school time. Retrieved 09/01/2007, from
http://www.niost.org/publications/FinaIZOOTFactSheetpdf.

Neulinger, J ., & Breit, M. (1969). Attitude Dimensions of Leisure. Journal of Leisure
Research.

Noam, G. (2004). After-school worlds: Creating a new social space for development and
learning. San Francisco: Jossey—Bass.

Noam, G. (2005). The significance of enrollment, attendance, and engagement. In H. B.
Weiss, P. Little & S. Bouffard (Eds), New Directions for Youth Development
(Vol. 105, pp. 1-4). San Franciso, California: Jossey—Bass.

Osgood, D. W., Anderson, A. L., & Shaffer, J. N. (2005). Unstructured leisure in the
alter-school hours. In J. L. Mahoney, R. W. Larson & J. S. Eccles (Eds),
Organized activities as contexts of development (pp. 45-64). Mahwah, New Jersey:
Lawrence Erlbaum Associates.

Oulette, M., Hutchinson, A. M., & Frant, N. (2005). The afterschool hours: A new focus
for America ’3 cities. Washington, DC: National League of Cities.

Parrish, T. B., Hikido, C. S., & Fowler, W. J. (1998). Inequalities in public school district
revenues. Washington, DC: US. Department of Education.

Pellegrini, A. D., & Smith, P. K. (1993). School recess: Implications for education and
development. Review of Educational Research, 63(1), 51-67.

Perkins, D. F., Borden, L. M., Villan'uel, F. A., Carlton-Hug, A., Stone, M. R., & Keith, J.
G. (2007). Participation in structured youth programs. Youth & Society, 38(4),
420-442.

Peugh, J. L., & Enders, C. K. (2004). Missing data in educational research: A review of

reporting practices and suggestions for improvement. Review of Educational
Research, 74(4), 525-556.

116

Posner, J. K., & Vandell, D. L. (1999). After-school activities and the development of
low-income urban children: A longitudinal study. Developmental Psychology,
35(3), 868—879.

Quinn, J. (1999). Where need meets opportunity: youth development programs for early
teens. The Future of Children, 9(2), 96-116.

R Development Core Team. (2008). R: A language and environment for statistical
computing (Version 2. 7.0). Vienna, Austria: R Foundation for Statistical
Computing.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models (2nd ed.). London:
Sage.

Raymore, L. A., Godbey, G., & Crawford, D. W. (1994). Self-Esteem, Gender, and
Socioeconomic Status: Their Relation to Perceptions of Constraint on Leisure
Among Adolescents. Journal of Leisure Research, 26(2), 99-1 18.

Reisner, E. R., White, R. N., Russell, C. A., & Birmingham, J. (2004). Building quality,
scale, and effectiveness in after-school programs: Summary report of the T ASC
evaluation. Retrieved 02/12/08, from
http://www.policystudies.com/studies/youth/TASC%20$ummary%20chort“ 020

Finalpdf.

Rivkin, M. S. (1995). The great outdoors: Restoring children 's right to play outside.
Washington, DC: National Association for the Education of Young Children.

Roadburg, A. (1983). Freedom and enjoyment: Disentangling perceived leisure journal of
Leisure Research, 15, 15-26.

Roj ek, C. Leisure and Recreation Theory. In Understanding Leisure and Recreation.

Ruskin, H., Cortez, L., Dempsey, S., Estrada, A., Foley, J ., Hope, D., et a1. (2001). World
Leisure and Recreation Association International Position Statement on Leisure
Education and Youth at Risk. Leisure Sciences, 23, 201-207.

Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of
intrinsic motivation, social development, and well-being. American Psychologist,
55, 68-78.

Schafer, J. L. (2007). mix: Estimation/multiple imputation for mixed categorical and
continuous data (R Package Version 1.0-6).

Shaw, S. M. (1985). The meaning of leisure in everyday life. Leisure Sciences, 7, 1-24.

Shaw, S. M. (1986). Leisure, recreation or freedom? Measuring time usage. Journal of
Leisure Research, I 8, 177-189.

117

Sherman, L., Gottfredson, D., macKenzie, D., Eck, J ., Reuter, P., & Bushway, S. (1998).
Preventing crime: l/Wzat works, what doesn 't, what's promising: National
Institute of Justice.

Shemoff, D. J ., & Vandell, D. L. (2007). Engagement in after-school program activities:
quality of experience fi'om the perspective of participants. Journal of Youth and
Adolescence, 2007(36), 891-903.

Shemoff, D. J ., & Vandell, D. L. (2008). Youth engagement and quality of experience in
afterschool programs. Afierschool Matters, Spring(9), 1-14.

Silbereisen, R. K., & Eyferth, K. (1986). Development as action in context. In R. K.
Silbereisen, K. Eyferth & G. Rudinger (Eds), Development as action in context:
Problem behavior and normal youth development (pp. 3-16). Berlin:
Springer-Verlag.

Simpkins, S. D., Ripke, M., Huston, A. C., & Eccles, J. S. (2005). Predicting
participation and outcomes in out-of-school activities: Similarities and differences
across social ecologies. In H. B. Weiss, P. M. D. Little & S. M. Buffard (Eds),
New direction for youth development (Vol. 105, pp. 51-69). San Francisco, CA:

J ossey—Bass.

Smith, P. (2000). What good schools can do about bullying Childhood, 7(2), 193-212.

Smith, P. K., & Sharp, S. (Eds). (1994). School bullying: Insights and perspectives
London: Routledge.

Sutton-Smith, B. (1990). School playground as festival. Children 's Environment
Quarterly, 7, 3-7.

Sutton-Smith, B. (1994). Does play prepare the firture? In J. H. Goldstein (Ed.), Toys,
play, and development (pp. 130—146). Cambridge, England: Cambridge
University Press.

Sutton-Smith, B. (1997). The ambiguity of play. Cambridge, MA: Harvard University
Press.

Theokas, C., Lerner, J. V., Phelps, E., & Lerner, R. M. (2006). Cacophony and change in
youth after school activities: Findings from the 4-H study of positive youth
development. Journal of Youth Development, 1(1).

United Nations. (2007). World youth report 2007 : Young people ’s transition to adulthood:
progress and challenges. New York: Department of Economic and Social Affairs,
United Nations.

US Department of Education. (2007). 21st-Century Community Learning Centers.
Retrieved. from http:r’/www.edgm’programs/Z 1 stcclc/index.html.

118

Van Egeren, L., & Reed, C. S. (2007). Michigan 2lst Century Community Learning
Centers: Five Years Old and Growing [Electronic Version]. 21 st CCLC Fall
Kick-Ojf2007-08. Retrieved 09/01/2007 from
http://www.outreach.msu.edu/cerc/21cclc.asp.

Van Egeren, L., Wu, H.-C., & Reed, C. (2007). What makes a successful after-school
program: The interaction between direct contact with student and good

management. Paper presented at the Society for Research in Child Development,
Boston, MA.

Vandell, D., & Ramanan, J. (1991). Children of the national longitudinal survey of youth:

Choices in after-school care and child development. Developmental Psychology,
27(4), 637-643.

Vandell, D., Shumow, L., & Posner, J. (2005). After-school programs for low-income
children: Differences in program quality. In J. L. Mahoney, R. W. Larson & J. S.
Eccles (Eds), Organized activities as contexts of development (pp. 437-456).
Mahwah, New Jersey: Lawrence Erlbaum Associates.

Vandell, D. L., Shemoff, D. F., Pierce, K. M., Bolt, D. M., Dadisman, K., & Brown, B. B.
(2005). Activities, engagement, and emotion in after-school programs (and
elsewhere). New Directions for Youth Development, 105, 121-129.

Vandell, D. L., & Shumow, L. (1999). After-school childcare programs. Future of
Children, 9(2), 64-80.

Villarruel, F. A., & Lerner, R. M. (1994). Development and context and the contexts of
learning. In F. A. Villarruel & R. M. Lerner (Eds), Promoting Community-Based
Programs for Socialization and Learning (Vol. 63, pp. 3-10). San Francisco:
Jossey—Bass Publishers.

Villarruel, F. A., Montero-Sieburth, M., Dunbar, C., & Outley, C. W. (2005). Dorothy,
There is no yellow brick road: The paradox of community youth development
approaches for Latino and African American Urban Youth. In J. Mahoney, R.
Larson & J. S. Eccles (Eds), Organized activities as contexts of development (pp.
111-129). Mahwah, New Jersey: Lawrence Erlbaum Associates, Inc.

Villarruel, F. A., Perkins, D. F., Borden, L. M., & Keith, J. G. (2003). Community Youth
Development: Programs, Policies and Practices.

Walker, K. E., & Arbreton, A. J. A. (2001). Working together to build Beacons Centers
in San Francisco: Evaluation findings from 1998—2000. Philadelphia:
Public/Private Ventures.

Weiss, H. R., Little, R, & Bouffard, S. (2005). More than just being there: Balancing the
participation equation. New Directions for Youth Development, 105, 15-31.

119

West, S., & Crompton, J. (2001). A review of the impact of adventure programs on
at-risk youth. Journal of Park and Recreation Administration, 19(2), 113-140.

Wimer, C., Bouffard, S. M., Caronongan, P., Dearing, E., Simpkins, S., Little, P. M. D.,
et al. (2006). What are kids getting into these days? Demographic differences in
youth out-of school time participation, Harvard Family Research Project:
Harvard Family Research Project.

Witt, P. A., & Caldwell, L. L. (2005). Recreation and youth development. State College,
Pennsylvania: Venture Publishing, Inc.

Witt, P. A., & Crompton, J. L. (1996). The at-risk youth recreation project. Journal of
Park and Recreation Administration, 14(3), 1-9.

120