You are on page 1of 13

This article was downloaded by: [Newcastle University]

On: 28 April 2014, At: 05:12


Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Journal of American College Health


Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/vach20

Gender Differences in College Leisure Time Physical


Activity: Application of the Theory of Planned Behavior
and Integrated Behavioral Model
a b c b
Jill M. Beville PhD , M. Renée Umstattd Meyer PhD, MCHES , Stuart L. Usdan PhD , Lori
b d e
W. Turner PhD , John C. Jackson PhD & Brad E. Lian PhD
a
Department of Campus Recreation , University of North Carolina at Greensboro ,
Greensboro , North Carolina
b
Department of Health Sciences , University of Alabama , Tuscaloosa , Alabama
c
Department of Health, Human Performance, and Recreation , Baylor University , Waco ,
Texas
d
Department of University Recreation , University of Alabama , Tuscaloosa , Alabama
e
Department of Community Medicine , Mercer University School of Medicine , Macon ,
Georgia
Accepted author version posted online: 12 Dec 2013.Published online: 05 Mar 2014.

To cite this article: Jill M. Beville PhD , M. Renée Umstattd Meyer PhD, MCHES , Stuart L. Usdan PhD , Lori W. Turner PhD ,
John C. Jackson PhD & Brad E. Lian PhD (2014) Gender Differences in College Leisure Time Physical Activity: Application of
the Theory of Planned Behavior and Integrated Behavioral Model, Journal of American College Health, 62:3, 173-184, DOI:
10.1080/07448481.2013.872648

To link to this article: http://dx.doi.org/10.1080/07448481.2013.872648

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
JOURNAL OF AMERICAN COLLEGE HEALTH, VOL. 62, NO. 3

Major Article

Gender Differences in College Leisure Time


Physical Activity: Application of the Theory
of Planned Behavior and Integrated
Behavioral Model
Downloaded by [Newcastle University] at 05:12 28 April 2014

Jill M. Beville, PhD; M. Renée Umstattd Meyer, PhD, MCHES;


Stuart L. Usdan, PhD; Lori W. Turner, PhD; John C. Jackson, PhD;
Brad E. Lian, PhD

P
Abstract. Objective: National data consistently report that males
participate in leisure time physical activity (LTPA) at higher rates
hysical activity is the cornerstone of a healthy lifestyle
than females. This study expanded previous research to examine and is cited as a key strategy for reducing the risk
gender differences in LTPA of college students using the theory of chronic conditions and diseases, including hyper-
of planned behavior (TPB) by including 2 additional constructs, tension, coronary heart disease, diabetes, cancer, and obe-
descriptive norm and self-efficacy, from the integrated behavioral sity.1 Benefits such as maintenance of ideal body weight,
model. Participants: Participants were college students (N = 621)
from a large public university in the southeastern United States.
prevention of premature death, and enhanced psychological
Methods: A self-report, classroom-based assessment with vali- well-being have been attributed to participation in regular
dated and reliable measures of LTPA, TPB constructs, descriptive physical activity.2 However, data from the 2011 Behavioral
norm, self-efficacy, and demographics was conducted in fall 2009. Risk Factor Surveillance System indicate that approximately
Results: Regression analyses revealed attitude (β = .119), inten- only 52% of US adults meet the recommendations for reg-
tion (β = .438), self-efficacy (β = .166), body mass index (BMI)
(β = −.084), and sports participation (β = .081) as significantly
ular leisure time physical activity (LTPA).3 The need to un-
associated with LTPA for females (R2 = .425, p < .001), whereas in- derstand physical activity behavior and implement effective
tention (β = .371) was significant for males (R2 = .202, p < .001). intervention strategies is paramount.4
Conclusions: Practitioners should consider tailoring promotional Despite the strong evidence for health promoting benefits
materials to address these gender differences in efforts to increase of engaging in regular physical activity, participation rates
LTPA participation among college students.
decrease over the life span, with the steepest decline in LTPA
Keywords: behavior change, exercise, males and females, univer- occurring in adolescence (approximately 15 to 18 years of
sity students, theory, young adults age) and young adulthood (approximately 20 to 25 years
of age).5–8 Epidemiological evidence indicates that levels
Dr Beville is with the Department of Campus Recreation at the of LTPA decline from high school to college and activ-
University of North Carolina at Greensboro in Greensboro, North ity patterns in college populations are generally insufficient
Carolina, and the Department of Health Sciences at the University to improve health and fitness.8 National data suggest that
of Alabama in Tuscaloosa, Alabama. Dr Umstattd Meyer is with only 47.4% of college students meet recommendations for
the Department of Health, Human Performance, and Recreation at moderate-intensity exercise, vigorous-intensity exercise, or
Baylor University in Waco, Texas. Dr Usdan and Dr Turner are with
the Department of Health Sciences at the University of Alabama a combination of the two.9 The college years provide unique
in Tuscaloosa, Alabama. Dr Jackson is with the Department of opportunities (and responsibilities) for campus communities
University Recreation at the University of Alabama in Tuscaloosa, to positively shape physical activity behaviors of their stu-
Alabama. Dr Lian is with the Department of Community Medicine dents.10 The college environment, with its potential to reach
at Mercer University School of Medicine in Macon, Georgia. a large audience in a relatively confined area, may be an
Copyright © 2014 Taylor & Francis Group, LLC

173
Beville et al

Experiential
attitude Knowledge and
skills
Attitude

Instrumental
attitude Salience of
behavior

Injunctive
norm
Perceived Norm INTENTION BEHAVIOR
Descriptive
norm

Environmental
constraints
Perceived
control

Personal Agency Habit

Self-efficacy
Downloaded by [Newcastle University] at 05:12 28 April 2014

FIGURE 1. Schematic representation of the integrated behavior model. Adapted from Montano DE, Kasprzyk
D. Figure 4.2: integrated behavioral model. In: Glanz K, Rimer BK, Viswanath K., eds. Health Behavior and
Health Education—Theory, Research and Practice . 4th ed. San Francisco, CA: Jossey-Bass; 2008:77 (color figure
available online).
© John Wiley & Sons Inc. Reproduced by Permission of John Wiley & Sons Inc. Permission to reuse must be
obtained from the rightsholder.

ideal setting for disseminating information and delivering understand behavior and identify specific beliefs to target in
health promotion interventions.11 Data suggest that there is intervention strategies (see Figure 1).16 More specifically, the
a need to study physical activity determinants and behavior IBM builds on the TPB and includes self-efficacy, described
so that researchers can develop better programs and inter- by Albert Bandura as one’s degree of confidence in the abil-
ventions to improve the physical activity patterns of college ity to perform a behavior in the face of various obstacles or
students.8 challenges.17 In addition, the IBM describes subjective norm
as being made up of both injunctive and descriptive norms.
Theoretical Background Descriptive norms are defined as perceptions about what oth-
This study was conducted to gain a better understand- ers in one’s social or personal networks are doing in regards
ing of factors that impact LTPA behaviors of college stu- to a particular behavior.16
dents using the theory of planned behavior (TPB) and in- Although the IBM presents descriptive norm as an ad-
tegrated behavioral model (IBM). Physical activity inter- dition to the TPB, other researchers have incorporated this
ventions based on theory have been shown to improve ef- concept with the TPB in their investigation of LTPA.13,18–21
fectiveness.12 The TPB is one of several intrapersonal-level Additionally, previous research has been conducted using
value expectancy theories that have been regularly employed self-efficacy along with the TPB constructs.18,22 The results
to examine the disjuncture between awareness of the ben- of previous research has been mixed: one study found that
efits of physical activity and low levels of engagement in neither self-efficacy or perceived behavioral control were as-
physical activity.13 Value expectancy theories are based on sociated with physical activity,18 whereas another found that
the foundational belief that behavior is achievable when in- both perceived behavioral control and self-efficacy each had
dividuals (1) “value” an outcome related with a behavior unique and independent relationships with physical activ-
and (2) “expect” that the specific behavior will lead to this ity.22 To date, the IBM has not been used in previous re-
outcome.14 search targeting LTPA.16 Therefore, conducting a study that
The central construct of the TPB is the person’s intention utilizes the constructs contained within the TPB and descrip-
to engage in a particular behavior.15 Furthermore, the TPB tive norm and self-efficacy from the IBM is unique to this
posits that intention is a function of 3 determinants: atti- study and adds to the current LTPA literature. In addition, in-
tude, subjective norm, and perceived behavioral control. The vestigating college students’ engagement in LTPA provides
IBM includes constructs from the TPB and other influen- a distinct context within the current literature. The potential
tial frameworks to provide a theoretical basis from which to public health benefits of using campus settings to positively

174 JOURNAL OF AMERICAN COLLEGE HEALTH


Gender Differences in College Leisure Time Physical Activity

influence the physical activity habits of young adults should ten form that their participation was voluntary and that the
not be underestimated.10 research was completely anonymous. In addition, the stu-
Previous research suggests that LTPA among college stu- dents were informed of and provided with a written defini-
dents differs by gender.23–25 In a study exploring knowledge tion of regular LTPA (either 30 minutes of moderate-intensity
of physical activity among a group of college students, gender LTPA—ie, brisk walking, tennis, easy bicycling, on at least
differences were reported for awareness of the definition and 5 days per week—or 20 minutes of vigorous LTPA—ie, run-
health benefits of physical activity and the components of the ning, aerobics, fast bicycling, on at least 3 days per week).
public health recommendations for this behavior.23 A study When the students completed their survey, it was returned
by Blanchard and colleagues revealed gender differences in to the course instructor who then returned all surveys to the
behavioral, normative, and control beliefs among college stu- researchers. This process was implemented in an attempt to
dents.24 For Caucasians, the association between affective greater protect anonymity. There were no incentives provided
attitudes and intention was stronger for females. However, for participation.
in African Americans, this association was stronger among
males. National data consistently acknowledge that males Measures
participate in LTPA at higher rates than females.9,26 Demographics were assessed by self-report using ques-
tions from the American College Health Association–
Potential Confounding Factors National College Health Assessment II (ACHA-NCHA II).
Several potential confounding factors for LTPA have been Questions included age, gender, self-reported height and
Downloaded by [Newcastle University] at 05:12 28 April 2014

cited in the literature. More specifically, for college students weight, year in school, ethnicity, marital status, on-/off-
several factors, including body mass index (BMI),11,25,27 year campus residence, affiliation with a social fraternity/sorority,
in school,28 race,24,25,28,29 Greek affiliation,30 and place of grade point average (GPA), and participation in varsity, club,
residence,30,31 have all been related with physical activity in and/or intramural sports.
previous research. Therefore, these factors were examined
as potential confounders when testing the relationships of Attitude
interest. Attitude can be defined as the degree to which a person has
This study sought to apply the TPB and the 2 specific a favorable or unfavorable evaluation or appraisal of the be-
IBM constructs, descriptive norm and self-efficacy, to in- havior in question.15 A 6-item scale developed by Blanchard
vestigate LTPA behaviors of college students. These find- et al was used to assess attitude.24 Blanchard and colleagues29
ings should aid in identifying specific factors associated with previously reported the affective attitude scale to have ac-
students’ participation in LTPA and provide insight regard- ceptable internal consistency for both African American
ing how to increase students’ engagement in regular LTPA. (α = .74) and Caucasian (α = .79) students. The instru-
This research investigated potential differences in LTPA mental attitude scale was also found to have acceptable in-
participation between male and female students using the ternal consistency for both African American (α = .70) and
TPB and IBM. Identifying gender differences has practical Caucasian (α = .81) students.29 Based on Blanchard et al’s
applications for practitioners to allow for better tailoring of approach, the statement “For me to participate in moderate-
LTPA social marketing messages and interventions among intensity leisure time physical activity over the next week will
the students. We hypothesized that TBP constructs, descrip- be. . .” preceded a list of adjectives. Participants were asked
tive norm, and self-efficacy would be independently associ- to rate 6 items related to attitude on a 7-point semantic differ-
ated with LTPA. In addition, based on previous research, we ential adjective scale. This scale taps both the instrumental
hypothesized that there would be gender differences in the (harmful–beneficial, bad–good, useless–useful) and affective
TPB and IBM descriptive norm and self-efficacy constructs (unpleasant–pleasant, boring–fun, unenjoyable–enjoyable)
in relation to participation in LTPA.23–25 aspects of attitude. Scale scores were formed by averaging
the responses to the affective and instrumental norm ques-
tions (higher scores indicate a more favorable attitude). Inter-
METHODS nal consistency was also acceptable for attitude in the present
Participants study (α = .85).
After obtaining Institutional Review Board approval,
paper-and-pencil questionnaires were administered to stu- Subjective Norm (Injunctive Norm)
dents enrolled in 15 large lecture undergraduate courses from Subjective norm is the perceived social pressure to perform
4 different colleges within a large public university in the or not to perform a particular behavior15 and was measured
Southeast. Surveys were completed between November and using 3 items with a Likert scale rated from 1 (strongly dis-
December of the 2009 Fall semester. The following recruit- agree) to 7 (strongly agree). The 3 items for the subjective
ment approach was taken to obtain a sample of students norm construct measure injunctive norms, which are the nor-
that represented the enrollment of the institution. Male and mative beliefs about what others think one should do.16 These
female college students who were enrolled in the selected items included “Most people important to me (1) definitely
undergraduate courses were informed verbally and in writ- think I should, (2) definitely approve of me, and (3) definitely

VOL 62, APRIL 2014 175


Beville et al

support me in engaging in moderate LTPA.”24,29 Scale scores with college students to determine if descriptive norm was a
were calculated by averaging the responses to the 3 questions significant and positive predictor of intention to participate in
(higher scores indicate stronger subjective norm for LTPA). LTPA and participation in LTPA. The first item asks whether
Internal consistency (Cronbach’s α) for the 3-item scale was most of the subjects’ family members will engage in mod-
previously reported as acceptable (α = .81).24 Internal con- erate LTPA during the next week. The second asks whether
sistency for this scale was also similar in the present study or not the subjects’ friends will engage in moderate LTPA
(α = .82). during the next week.19 Okun et al19 originally measured
these items on a Likert scale from 0 (strongly disagree) to 6
Perceived Behavioral Control (strongly agree). However, for consistency of this research
Perceived behavioral control is defined as the perceived project with the other scales, the ability to compare means
ease or difficulty of performing a behavior and it reflects and improve understandability for the students, both ques-
past experience as well as anticipated impediments and ob- tions were measured on a scale from 1 (strongly disagree)
stacles.15 This construct was measured using a 3-item scale to 7 (strongly agree). A scale score was formed by averag-
from Blanchard and colleagues.29 These items included the ing the responses of the 2 questions (higher scores indicate
following: (1) “During the next week, if I wanted to I could stronger descriptive norm for LTPA). Internal consistency for
engage in moderate leisure time physical activity” rated on this scale has not been previously reported. Internal consis-
a scale of 1 (strongly disagree) to 7 (strongly agree); (2) tency in the present study was α = .53. Nunnally32 has indi-
“During the next week, it would be for me to cated .70 to be an acceptable reliability coefficient, but lower
Downloaded by [Newcastle University] at 05:12 28 April 2014

engage in moderate leisure time physical activity” rated on a thresholds are sometimes used in the literature. Although in-
scale of 1 (extremely difficult) to 7 (extremely easy); and (3) ternal consistency was lower than usually acceptable, given
“During the next week, how much control do you believe you that these items were specifically created for use with col-
have to engage in moderate leisure time physical activity” lege students, they were retained in bivariate analyses in the
rated on a scale of 1 (extreme lack of control) to 7 (extreme present study to allow for rudimentary examination of all
control). Scale scores were calculated by averaging the re- constructs of interest. However, given low internal consis-
sponses to the 3 questions (higher scores indicate greater tency, it was not included in multivariate analyses.
perceived control for LTPA). Blanchard and colleagues29
previously reported this perceived behavioral control scale to Self-efficacy
have an acceptable internal consistency for African American Self-efficacy is one’s degree of confidence in the ability to
(α = .81) and Caucasian (α = .76) students. This perceived perform a given behavior in a specific setting or context.14
behavioral control scale also had acceptable internal consis- One form of self-efficacy for LTPA, barrier self-efficacy, was
tency in the present study (α = .75). measured using the 5-item scale developed by Marcus and
colleagues.33 This scale measures how confident participants
Intention are that they can participate in LTPA when faced with com-
Intention to perform a given behavior indicates how hard mon barriers: negative affect, resisting relapse, and making
people are willing to try and how much effort they plan to time for exercise. Specifically, the 5 questions included “Dur-
exert in order to perform a behavior that is under volitional ing the next week, I am confident I can engage in leisure time
control.15 Intention toward the behavior was measured using physical activity even when (1) I am tired, (2) I am in a bad
a previously developed 3-item scale.24,29 The first 2 items, mood, (3) I feel I don’t have time, (4) I am on vacation, and
(1) “During the next week, I intend to engage in moderate (5) the weather is bad.” Participants rated their confidence on
leisure time physical activity” and (2) “During the next week, a Likert scale from 1 (not at all confident) to 7 (very confi-
I will engage in moderate leisure time physical activity,” dent). Self-efficacy scores were calculated by summing the
were measured on a Likert scale from 1 (strongly disagree) scores on the individual items (higher scores indicate greater
to 7 (strongly agree). The third item states, “During the next self-efficacy). This scale has previously reported acceptable
week, my goal is to engage in moderate-intensity leisure time 2-week test-retest reliability (r = .90) and internal consis-
physical activity on at least days.” As previously reported, tency (α = .82).33 In the present study, the self-efficacy scale
a composite intention score was formed by averaging the also had acceptable internal consistency (α = .81).
responses to the 3 items (higher scores indicate a greater
intention to participate in LTPA).24,29 Acceptable internal Leisure Time Physical Activity
consistency (α = .87) for this scale was previously reported.24 LTPA behavior was measured using the Godin Leisure
Internal consistency in the present study was also acceptable Time Exercise Questionnaire.34 The questionnaire is reliable,
(α = .89). valid, and easy to complete quickly without a need for de-
tailed review.34 The questionnaire is relatively brief and asks
Descriptive Norm participants to recall their participation in LTPA (for more
Descriptive norm is defined as perceptions about what oth- than 15 minutes) during the past 7 days. The instrument
ers in one’s social or personal networks are doing.16 Okun and contains 3 open-ended questions covering the frequency of
colleagues19 developed this 2-item scale specifically for a use mild (eg, easy walking), moderate (eg, fast walking), and

176 JOURNAL OF AMERICAN COLLEGE HEALTH


Gender Differences in College Leisure Time Physical Activity

strenuous (eg, jogging) exercise completed during free time. allows researchers to use standard complete-data meth-
As previously done by Rhodes and colleagues, the phrase ods of analysis on the filled in data.42 Based on this ap-
“leisure time physical activity” was substituted for “exer- proach, 38 participant TPB subscales (intention, n = 34;
cise” on the instrument.35 This survey instrument has been attitude, n = 4) had means imputed for one missing item
consistently used to identify engagement in LTPA and is response.
widely used with the college student population.19,29,35–39 The final sample consisted of 621 (80.1% of the returned
LTPA scale scores were created according to the Godin survey sample) undergraduate students enrolled in classes
Leisure Time Physical Activity Questionnaire. Weekly fre- during the fall 2009 at a large public university in the south-
quencies of strenuous, moderate, and light activities were eastern United States (see Table 1). The majority of the
multiplied by 9, 5, and 3, respectively. Overall LTPA scale sample were female (n = 421; 67.8%) with a mean age
scores were then calculated by summing the products of the of 20.18 (SD = 1.572). With regard to ethnicity, the sam-
3 separate components.34 For the purposes of this research, ple was predominantly white (n = 504; 81.2%) or black
LTPA was analyzed as a continuous variable. (n = 91; 14.7%). Just over a fifth of the sample reported
being in their first year of college (n = 130; 20.9%), 25.9%
Analyses in their second year (n = 161), 19.6% in their third year
All data were entered, cleaned, and analyzed using SPSS (n = 122), and 33.5% reported being in their fourth or
version 17 statistical software (SPSS, Chicago, Illinois). Sig- fifth year of college. The majority (n = 407; 65.5%) re-
nificance was examined using p ≤ .05, although exact p val- ported that they were not affiliated with a Greek (frater-
nity or sorority) organization and 60.5% (n = 376) in-
Downloaded by [Newcastle University] at 05:12 28 April 2014

ues are reported. Due to positively skewed data, LTPA scores


were corrected for gross errors by Windsorizing.40 The pro- dicated not participating in intramural and/or club sports.
cess of Windsorizing eliminates the highest or lowest values Mean BMI calculated from reported height and weight, was
and replaces them with the next highest or lowest values to 23.51 (SD = 4.189), and the mean self-reported GPA for the
correct large errors or large abnormalities in distributions. In 502 students who responded to this item was 3.139 (SD =
this sample, several LTPA scale scores were high (n = 13), 0.481). Scores for each theoretical construct are presented in
and were therefore replaced with values based on the 98th Table 2.
percentile value.40
To determine if there were gender-related differences Bivariate Analyses
in the relationships of specific TPB/IBM constructs with One-way ANOVAs indicated that gender was significantly
LTPA, several analyses were conducted. Analysis of vari- related with BMI, attitude, subjective norm, intention, self-
ance (ANOVA) and chi-square analyses were used to test for efficacy, and LTPA. However, gender was not related with
demographic differences among the gender groups. Pearson age (F = 2.622, p = .106), descriptive norm (F = 0.902,
correlations were also examined to determine whether or not p = .343), or perceived behavioral control (F = 3.705,
the variables related as hypothesized by the TPB/IBM among p = .055). In this college sample, male students had signifi-
each gender. Similar methods were employed by Blanchard cantly higher (F = 34.52, p < .001) BMI scores (M = 24.92,
et al.29 Specifically, the TPB and IBM state that attitudes to- SD = 4.03) than female students (M = 22.84, SD = 4.10).
ward the behavior, subjective norm/perceived norm, and per- In addition, males had significantly higher (F = 11.183,
ceived behavioral control/personal agency should be related p = .001) attitude scores (M = 5.99, SD = 0.86) than fe-
with intention. Finally, linear regression analyses were con- males (M = 5.69, SD = 0.95). Males had significantly lower
ducted to determine if associations of TPB/IBM constructs (F = 10.679, p = .001) subjective norm scores (M = 5.58,
with LTPA behavior differed by genders. The TPB/IBM con- SD = 1.00) than females (M = 5.89, SD = 1.13). Intention
structs, intention, and LTPA were all continuous variables in scores were also significantly different (F = 11.544, p =
this analysis. .001) between males (M = 5.01, SD = 1.19) and females
(M = 4.64, SD = 1.34). Male students had significantly
RESULTS higher (F = 23.649, p < .001) self-efficacy scores (M =
21.40, SD = 6.46) than female students (M = 18.66,
Characteristics of the Sample SD = 6.63). In addition, there was a significant difference in
Of the 1,004 students enrolled in the participating courses, reported LTPA (F = 14.627, p < .001) between males (M =
a total of 775 surveys were returned (77.2% response rate). 57.26, SD = 25.92) and females (M = 49.05, SD = 24.54).
Subsequently, respondents were not included in the analyses Cross-tabulation and chi-square analyses indicated sig-
if they reported being involved in varsity athletics (n = 75), nificant differences in year in school (χ 2 = 11.910, p =
failed to complete demographic items (n = 22), were not .008), Greek affiliation (χ 2 = 6.329, p = .012), and sports
of traditional college age (18–24; n = 37), were graduate participation (χ 2 = 93.720, p < .001) based on gender.
students (n = 4), or if they were missing 2 or more responses Forty percent of first-year students were male (60% female),
for a scale of interest (n = 16).41 22.3% of second-year students were male (77.6% female),
Among the wide variety of procedures used to handle 31.1% of third-year students were male (68.9% female), and
missing data, imputing means for missing values is a com- 35.6% of fourth- and fifth-year students were male (64.4%
mon strategy used to address missing item responses, which female). Only 27.5% of males were affiliated with Greek

VOL 62, APRIL 2014 177


Beville et al

TABLE 1. Demographics and Leisure Time Physical Activity Behavior of a Sample of College Studentsa

Characteristic n % M SD

Gender
Male 200 32.2
Female 421 67.8
Age 20.18 1.572
Body mass index 23.51 4.189
Grade point averageb 3.139 0.481
Year in school
First year 130 20.9
Second year 161 25.9
Third year 122 19.6
Fourth year 135 21.7
Fifth year 73 11.8
Racial/Ethnic group
White 504 81.2
Black 91 14.7
Hispanic/Latino 5 0.8
Asian/Pacific Islander 4 0.6
Downloaded by [Newcastle University] at 05:12 28 April 2014

American Indian/Alaskan Native/Native Hawaiian 1 0.2


Biracial or multiracial 15 2.4
Other 1 0.2
Greek affiliation
Yes 214 34.5
No 407 65.5
Sports participation
Club sports 32 5.2
Intramural sports 192 30.9
Club sports and intramural sports 21 3.4
None 376 60.5
Residence
Campus residence hall 168 27.1
Fraternity or sorority house 21 3.4
Other college/university housing 31 5.0
Parent/Guardian’s home 23 3.7
Other off-campus housing 371 59.7
Other 7 1.1
Marital status
Single 594 95.7
Married/partnered 18 2.9
Divorced 2 0.3
Other 7 1.1
aN = 621.
bN = 502 (excludes those who answered “Not applicable”).

TABLE 2. Theory of Planned Behavior and Integrated Behavioral Model Subscale Scores

Scale/Subscale M SD Range

Intention 4.76 1.30 0.67–7.00


Attitude 5.78 0.93 1.00–7.00
SJNorm 5.79 1.10 1.00–7.00
PBC 5.79 1.04 1.67–7.00
DSNorm 4.34 1.14 1.00–7.00
SEfficacy 19.54 6.69 5.00–35.00
LTPA 51.69 25.27 0.00–116.00

Note. SJNorm = subjective norm (injunctive norm); PBC = perceived behavioral control; DSNorm = descriptive norm; SEfficacy = self-efficacy;
LTPA = leisure time physical activity.

178 JOURNAL OF AMERICAN COLLEGE HEALTH


Gender Differences in College Leisure Time Physical Activity

TABLE 3. Pearson Correlation Coefficients of Theory of Planned Behavior Constructs, Descriptive Norm,
Self-efficacy, and Leisure Time Physical Activity of a Sample College Students

Intention Attitude SJNorm PBC DSNorm SEfficacy LTPA

Intention — .431∗∗ .307∗∗ .562∗∗ .427∗∗ .664∗∗ .622∗∗


Attitude .523∗∗ — .228∗∗ .368∗∗ .226∗∗ .408∗∗ .389∗∗
SJNorm .198∗∗ .218∗∗ — .241∗∗ .310∗∗ .232∗∗ .152∗∗
PBC .429∗∗ .291∗∗ .207∗∗ — .327∗∗ .430∗∗ .373∗∗
DS Norm .228∗∗ .093 .154∗ .087 — .370∗∗ .340∗∗
SEfficacy .699∗∗ .447∗∗ .176∗ .401∗∗ .228∗∗ — .535∗∗
LTPA .459∗∗ .190∗∗ −.048 .185∗∗ .190∗∗ .400∗∗ —

Note. Males (below diagonal): n = 200; females (above diagonal): n = 421. SJNorm = subjective norm (injunctive norm); PBC = perceived
behavioral control; DSNorm = descriptive norm; SEfficacy = self-efficacy; LTPA = leisure time physical activity.
∗∗ p < .01; ∗ p < .05.

organizations, and 37.8% of females were affiliated with only significant correlate of LTPA for male students was in-
Downloaded by [Newcastle University] at 05:12 28 April 2014

Greek organizations. For sports participation, 67% of males tention (p < .001). Specifically, for male students, the higher
participated in sports, whereas only 26.4% of female students one’s intention score, the more likely they were to participate
participated in sports. Chi-square results for ethnicity (χ 2 = in LTPA.
3.248, p = .197), marital status (χ 2 = 0.016, p = .898), and For female students, the model was also significant (F =
residence (χ 2 = 0.149, p = .700) did not differ by gender. 33.726, p < .001) and explained 42.5% of the variance
Pearson correlation coefficients were calculated by gender in LTPA. For female students, attitude (p = .006), in-
with regard to the TPB and IBM constructs and LTPA (see Ta- tention (p < .001), self-efficacy (p = .001), BMI (p =
ble 3). For males (n = 200), the constructs attitude, perceived .033), and sports participation (p = .034) were signifi-
behavioral control, intention, descriptive norm, and self- cantly related with greater participation in LTPA. Specifi-
efficacy were correlated with LTPA (p < .001). Self-efficacy, cally, the higher a female student’s attitude, intention, and
attitude, and perceived behavioral control were related with self-efficacy scores, the more likely she was to participate in
intention for males (p < .001). For females (n = 421), LTPA. Additionally, a lower BMI and participation in sports
all of the constructs—attitude, subjective norm, perceived were also related with greater participation in LTPA for fe-
behavioral control, intention, descriptive norm, and self- male students. No other variables were significant in this
efficacy—were correlated with LTPA, and self-efficacy, per- model.
ceived behavioral control, and attitude were all related with
intention (p < .001) (see Table 3). COMMENT
Conclusions
Multivariate Analyses This study was unique in that it examined the relationship
A multiple linear regression analysis was conducted to of the IBM constructs descriptive norm and self-efficacy
examine associations between LTPA and attitude, subjective regarding LTPA behavior in addition to TPB constructs.
norm, perceived behavioral control, intention, self-efficacy, The IBM, and more specifically descriptive norm and self-
and intention by gender. Given the low internal consistency efficacy, have not been examined in other LTPA studies using
coefficient for descriptive norm (α = .53), it was not in- the TPB.24,29,43–45 This study supported the IBM’s inclusion
cluded in multivariate analyses. Results from the chi-square of self-efficacy in examinations of LTPA behavior among this
analyses and ANOVAs indicated that year in school, Greek population, as it was significantly associated with and further
affiliation, sports participation, and BMI were significantly explained participation in LTPA. Although descriptive norm
associated with LTPA and therefore were included in the was not included in multivariate analyses, bivariate findings
multiple linear regression analysis as potential confounding suggest that it is potentially an important concept to consider
variables. after measures are further developed and validated.
Based on bivariate results, the following independent vari- As hypothesized, the findings of this research indicate
ables were included in the multiple linear regression models that there are gender differences with respect to the the-
examining correlates of LTPA for males and females: sub- oretical constructs related with LTPA. For male students,
jective norm (females only), attitude, perceived behavioral attitude, perceived behavioral control, intention, descriptive
control, self-efficacy, intention, BMI, year in school, Greek norm, and self-efficacy were all correlated with LTPA be-
affiliation, and sports participation (see Table 4). For male havior in bivariate analyses. However, subjective norm was
students, the model was significant (F = 6.038, p < .001) not significant. Among female students, all of the constructs
and explained 20.02% of the variance in LTPA. However, the were positively correlated with LTPA in bivariate analyses.

VOL 62, APRIL 2014 179


Beville et al

TABLE 4. Multiple Regression Model Examining Correlates of Leisure Time Physical Activity Among a Sample
of College Students by Gender

Item β p value t value Range M/% SD

Attitude M: −.098 .211 −1.25 2.00–7.00 5.96 0.86


F: .119 .006 2.76 1.00–7.00 5.69 0.95
SJNorm M: a a a
1.00–7.00 5.58 1.00
F: −.046 .253 −1.15 1.00–7.00 5.89 1.13
PBC M: −.065 .376 −0.89 2.00–7.00 5.91 1.05
F: .008 .870 0.16 1.67–7.00 5.74 1.04
Intention M: .371 <.001 3.79 1.33–7.00 5.01 1.19
F: .438 <.001 7.65 0.67–7.00 4.64 1.34
SEfficacy M: .157 .094 1.69 5.00–35.00 21.40 6.46
F: .166 .001 3.24 5.00–35.00 18.66 6.63
BMI M: .040 .545 0.61 16.04–44.91 24.92 4.03
F: −.084 .033 −2.14 16.13–45.72 22.84 4.10
Participates in sports M: .027 .692 0.40 67.0%
F: .081 .034 2.12 26.4%
School M: −.072 .280 −1.08
F: −.017 .660 −0.44
Downloaded by [Newcastle University] at 05:12 28 April 2014

Greek affiliated M: −.041 .546 −0.61 27.5%


F: −.021 .585 −0.55 37.8%

Note. M = males: n = 200; F = females: n = 421. Males: R2 = .202, p < .001; females: R2 = .425, p < .001. SJNorm = subjective norm; PBC =
perceived behavioral control; SEfficacy = self-efficacy; BMI = body mass index.
aSubjective norm was not included in the model for males, based on the bivariate analyses.

For both genders, intention was the strongest correlate with may be different among college students.47 There may be
LTPA, followed by self-efficacy, which supports the frame- some cultural influences that perpetuate an idea that males
work of the theories, where the most direct determinant of are supposed to play sports and be active and that it is not
behavior is intention.15,16 as accepted for females to engage in this behavior.48 Past
The findings of this research are consistent with previous behavior may also play an important role in explaining these
research in that others suggest that there are gender differ- gender differences.18 Participation in organized sport during
ences when it comes to LTPA among college students.23–25 adolescence may affect whether or not students continue to be
Behrens et al23 explored the understanding of moderate phys- active in college. The opportunities available to adolescents
ical activity among a group of college students. Their findings to play organized youth sports may have gender differences,
suggest that there are gender differences in the awareness of where males have more opportunities than females.49 How-
the definition and health benefits of physical activity and the ever, on a college campus, opportunities exist for organized
components of the public health recommendations for this sport through intramural and club sport participation for both
behavior.23 A study by Blanchard and colleagues24 found males and females. In this study, females who participated in
gender differences in the behavioral, normative, and control sports activities were more likely to engage in LTPA.
beliefs among black students, but not white students. Sum- One purpose of this research was to investigate the TPB
iniski et al25 used stages of change from the Transtheoretical while including descriptive norm and self-efficacy from the
Model to investigate gender differences in LTPA among col- IBM. The measure of descriptive norm was taken from pre-
lege students. They found that more males were in the main- viously published research; however, no previous tests of
tenance stage and more females were in the Contemplation validity or reliability were reported. In the present study
stage.25 descriptive norm was retained in bivariate analyses, which
Although gender is consistently related with physical ac- resulted in a positive correlation with LTPA. However, due
tivity in the literature, theoretical explanation for these as- to the low internal consistency coefficient for descriptive
sociations is scant.46 Even though theoretical frameworks norm (α = .53), it was not included in multivariate anal-
may not be able to explain the gender differences that exist yses. Future studies should develop a new measure of de-
with regard to physical activity, they are still present. Na- scriptive norm based on the process described in the TPB
tional data consistently demonstrate that males participate and subsequently establish reliability and validity within the
in higher rates of LTPA than females.9,26 There are several population of college students. Additionally, the scale used
possibilities that account for the gender differences found to measure descriptive norm in the present study only con-
among this population of college students. The social pres- sisted of 2 items, 1 item for friends and 1 for family. Thus,
sures on males and females to participate in physical activity conceptually each item could be described as its own 1-item

180 JOURNAL OF AMERICAN COLLEGE HEALTH


Gender Differences in College Leisure Time Physical Activity

subscale. Development of a valid and reliable scale is war- different opportunities on campus to engage in LTPA could
ranted, considering that rudimentary bivariate results seen in make students consider their intentions on attending these
the current study potentially suggest that descriptive norm programs and opportunities. In addition, social networking
is important to consider in addition to subjective norms. sites, such as FaceBook or Twitter, could be used to form
However, these results should be considered with caution groups, send event invitations, or make announcements about
given the low internal consistency seen in this present study. LTPA opportunities on campus.
Given these limitations regarding the current measurement In addition, practitioners and other college professionals
of descriptive norm, further examination of a more complete could specifically target female students regarding the di-
multivariate IBM model, including descriptive norm, is rec- rect relationships between LTPA and both attitude and self-
ommended after a valid and reliable descriptive norm scale is efficacy. Reaching out to female students in residence halls
developed. or in sorority houses to promote LTPA with messages geared
These findings provide a better understanding of LTPA towards improving LTPA attitudes may be an effective strat-
behaviors of college students by using the TPB and self- egy to increase LTPA participation. Research by Chatzis-
efficacy. This examination of LTPA among college students arantis and Hagger support the proposition that interventions
should be of interest to those fostering programs, services, can produce change in attitudes and intentions by address-
and facilities to support this behavior, and particularly cam- ing modal salient beliefs.51 In their study, the most popular
pus recreation professionals. In addition, the results parallel salient beliefs related with physical activity were identified:
previous research that has consistently supported the promi- having fun, staying fit, improving skill, getting an injury,
Downloaded by [Newcastle University] at 05:12 28 April 2014

nent roles of attitude and perceived behavioral control in and feeling hot and sweaty. They addressed these beliefs
the prediction of intention, with a lesser role for subjective through communication strategies and found that individuals
norms, and potential mediation of the effects of these con- who were given messages containing salient beliefs reported
structs on behavior by intentions.50 Future research should more positive attitudes and stronger intention than those
aim to examine mediating relationships using a prospective who were given messages with nonsalient beliefs. There-
study design. fore, identifying salient beliefs among college women for
Findings from this research suggest that it may be ad- use in marketing and promotional efforts could be an effec-
vantageous to tailor educational and promotional efforts for tive strategy to for improving attitudes and intentions towards
LTPA to college students and more specifically by gender. LTPA.
Distinct differences exist in the correlates of LTPA between Providing opportunities to enhance self-efficacy for fe-
male and female college students. The practical applications males would also be beneficial. In 2011, Williams and French
of the current findings suggest that practitioners should tailor reported an increase in self-efficacy and physical activity be-
educational and promotional materials and that one format havior when interventions used strategies that included action
may not be appropriate for all students. More specifically, planning, providing instruction, prompting practice, reinforc-
promotional material targeting male students should focus ing effort toward the behavior, prompting self-monitoring,
on intention, the significant predictor of LTPA among this setting graded tasks, planning social support, and/or relapse
college sample. Practitioners could take it one step further prevention.52 The researchers pointed out that identifying
and use these results showing that attitude, perceived behav- specific plans for when, where, and how the behavior will be
ioral control, and self-efficacy were positively correlated to performed improved both self-efficacy and physical activity.
intention for the male students. Therefore, these constructs Examples of how to tap into these recommendations include
could also be targeted in promotional materials, as this could offering group exercise classes or walking clubs at specific
indicate mediating relationships as supported by both the times and locations through the residence halls and/or soror-
TPB and IBM theoretical models. On the other hand, for ities. Engaging female students in smaller groups and/or in
females, promotional efforts should address attitude, inten- familiar settings may assist in increasing their confidence to
tion, and self-efficacy, which were the significant predictors perform LTPA. Some of the benefits of group physical ac-
of LTPA in this sample, suggesting potential partial and full tivity include exposure to a social and fun environment, a
mediating relationships between various constructs for fe- safe and effectively designed workout, a consistent exercise
males. Given the results of this study and the potential im- schedule, an accountability factor for participating in exer-
plications for practitioners, additional research is needed to cise, and a workout that requires no prior exercise knowledge
further confirm and understand these relationships. or experience.53 Campus recreation professionals could also
Practitioners or college professionals could use the present offer weight room orientations or other small-group instruc-
finding supporting the relationship between intention and tional opportunities that could enhance the students’ com-
LTPA to create intervention strategies for all students. Prac- fort and confidence in using equipment. In addition, finding
titioners could develop a promotional campaign for LTPA strategies to reinforce regular LTPA may have favorable re-
that stresses the importance of regular engagement in the sults.52 This could be accomplished through providing in-
behavior to target the intention construct. For example, a centives that reward or acknowledge small successes while
campaign entitled “Make It a Part of Your Day” could be working towards a larger goal. Using the findings from this
created to remind and encourage students to participate in research, grounded in theory, to create programs and promo-
LTPA. Sending e-mail reminders or announcements about tional strategies should prove beneficial.

VOL 62, APRIL 2014 181


Beville et al

Limitations hand, intrinsically motivated students expressed preference


There were limitations of this research that warrant further in being active in noncompetitive types and locations for ac-
discussion. First, this study relied on respondents to honestly tivity with lower visibility. However, intrinsically motivated
self-report their LTPA behavior. Although self-reporting is active students still reported using university recreation fa-
not unique to this particular study,24,29 there was potential cilities. More research to better understand the utilization
for students to inflate or underreport their LTPA participa- of campus recreation facilities of active and sedentary stu-
tion, whereby results support inflation. Therefore, to adjust dents is needed to help understand how sedentary college
for extreme values reported, data were Windsorized to help students can best be served and encouraged to engage in
eliminate large errors in self-reported LTPA.40 This research LTPA.
was also limited in that it was an exploratory cross-sectional Despite these limitations, the results of this study support
study. The cross-sectional study design limited our ability to the development and implementation of educational and pro-
examine causality. A more accurate picture of LTPA and the motional materials that target increasing college students’
examined relationships could be attained through a prospec- participation in LTPA. This research provides a better un-
tive study design. A prospective study would also further derstanding of the factors associated with participation in
validate the utility of the TPB and IBM in examining LTPA LTPA. The findings of this research also indicate that tai-
behavior and allow for potential mediating relationships to be loring materials based on gender would be beneficial, as the
explored. Future research should consider using a random- results demonstrated that different constructs were signifi-
ized prospective study to allow for examination of causality cantly related with LTPA dependent on gender. In addition,
Downloaded by [Newcastle University] at 05:12 28 April 2014

and these potential indirect effects. development of a valid and reliable scale for the descriptive
Another limitation was the lack of generalizability and norm construct is recommended. Findings of this research
potential selection bias associated with the use of a conve- provide preliminary support for the inclusion of the IBM
nience sample. This research used a convenience sample of construct self-efficacy, and potentially descriptive norm, in
college students from classes at 1 university in the southeast- conjunction with TPB in examining LTPA behavior. How-
ern United States. Thus, the results may not be generalizable ever, further research is necessary to fully examine the rela-
to other-aged populations or even other samples of college tionships among IBM and TPB constructs in relation to LTPA
students. However, this research attempted to minimize se- behavior.
lection bias by using general education classes from multi-
ple colleges that included students from several departments FUNDING
and various majors; thus, the sample should be relatively No funding was used to support this research and/or the
generalizable to other students at the university in which preparation of the manuscript.
the research was conducted. Overall, based on the referent
university’s published statistics, the demographic rates of CONFLICT OF INTEREST DISCLOSURE
this sample were fairly consistent with the entire under- The authors have no conflicts of interest to report. The
graduate population of the university, although this sample authors confirm that the research presented in this article
had a disproportionately higher number of females (67.8%) met the ethical guidelines, including adherence to the legal
and Greek-affiliated students (34.5%). In fall 2009, univer- requirements, of the United States and received approval
sity published statistics documented that there were 12,426 from the Institutional Review Board of Baylor University.
(52.4%) female students and 6,399 (27%) students affili-
ated with a Greek organization. In addition, this sample had NOTE
a slightly lower percentage of white students (81.2%) than
For comments and further information, address corre-
enrolled at the institution (82.9%), and a slightly higher per-
spondence to M. Renée Umstattd Meyer, Department of
centage of black students (14.7%) compared with the total
Health, Human Performance, and Recreation, Baylor Univer-
percent (11.8%) enrolled.
sity, One Bear Place 97313, Waco, TX 76798, USA (e-mail:
The influence of campus exercise or fitness facilities and
Renee Umstattd@Baylor.edu).
their hours of operation on college students’ physical activity
behaviors is still unclear.28 There is a lack of research re-
garding how many sedentary students use campus recreation REFERENCES
facilities and this should be considered in future planning of 1. Staten RR, Miller K, Noland MP, Rayens MK. College stu-
dents’ physical activity: application of an ecological model. Am J
interventions to increase LTPA. One qualitative study54 ex- Health Stud. 2005;20:58–65.
amining differences in physical activity location preference 2. Fischer DV, Bryant J. Effect of certified personal trainer
among active and inactive college students found location services on stage of exercise behavior and exercise mediators in
preferences to be related with intrinsic/extrinsic orientation, female college students. J Am Coll Health. 2008;56:369–376.
although inactive students engaged in less physical activity 3. Centers for Disease Control and Prevention. U.S. physical
activity statistics. Available at: apps.nccd.cdc.gov/brfss. Published
regardless of orientation. In this study, extrinsically moti- 2011. Accessed April 20, 2010.
vated students (active and inactive) preferred being active 4. Rhodes RE, Plotnikoff RC. Can current physical activ-
in places where they could be seen or see others and most ity act as a reasonable proxy measure of future physical activ-
often reported university recreation facilities. On the other ity? Evaluating cross-sectional and passive prospective designs

182 JOURNAL OF AMERICAN COLLEGE HEALTH


Gender Differences in College Leisure Time Physical Activity

with the use of social cognition models. Prev Med. 2005;40:547– 25. Suminski RR, Petosa RL, Utter AC, Zhang JJ. Physical ac-
555. tivity among ethnically diverse college students. J Am Coll Health.
5. Caspersen CJ, Powell KE, Christenson GM. Physical ac- 2002;51:75–80.
tivity, exercise, and physical fitness: definition and distinctions for 26. American College Health Association. American College
health-related research. Public Health Rep. 1985;100:126–131. Health Association- National College Health Assessment II: Refer-
6. Dishman RK, Buckworth J. Increasing physical activity: a ence Group Executive Summary Fall 2009. Baltimore, MD: Amer-
quantitative study. Med Sci Sports Exerc. 1996;28:706–719. ican College Health Association, 2009.
7. Stephens T, Jacobs DR, White CC. A descriptive epi- 27. Cilliers J, Senekal M, Kunneke E. The association between
demiology of leisure-time physical activity. Public Health Rep. body mass index of first-year female university students and their
1985;100:147–157. weight related perceptions and practices, psychological health,
8. Kilpatrick M, Hebert E, Bartholomew J. College students’ physical activity and other physical health indicators. Public Health
motivation for physical activity: differentiating men’s and women’s Nutr. 2006;9:234–243.
motives for sport participation and exercise. J Am Coll Health. 28. Keating XD, Guan J, Pinero JC, Bridges DM. A meta-
2005;54:87–94. analysis of college students’ physical activity behaviors. J Am Coll
9. American College Health Association. American College Health. 2005;54:116–125.
Health Association–National College Health Assessment II: Refer- 29. Blanchard CM, Fisher J, Sparling PB, et al. Understanding
ence Group Executive Summary Fall 2011. Baltimore, MD: Amer- physical activity behavior in african american and caucasian college
ican College Health Association; 2011. students: an application of the theory of planned behavior. J Am Coll
10. Leslie E, Sparling PB, Owen N. University campus set- Health. 2008;56:341–346.
tings and the promotion of physical activity in young adults: 30. Dinger M. Physical activity and dietary intake among college
lessons from research in Australia and the USA. Health Educ. students. Am J Health Stud. 1999;15:139–148.
2001;101:116–125. 31. Irwin JD. Prevalence of university students’ sufficient
Downloaded by [Newcastle University] at 05:12 28 April 2014

11. Doerksen SE, Umstattd MR, McAuley E. Social cognitive physical activity: a systematic review. Percept Mot Skills.
determinants of moderate and vigorous physcial activity in college 2004;98:927–943.
freshmen. J Appl Soc Psychol. 2009;39:1201–1213. 32. Nunnally JC. Psychometric Theory. 2nd ed. New York, NY:
12. Suminski RR, Petosa R. Web-assisted instruction for chan- McGraw-Hill; 1978.
ing social cognitive variables related to physical activity. J Am Coll 33. Marcus BH, Selby VC, Niaura RS, Rossi JS. Self-efficacy
Health. 2006;54:219–225. and the stages of exercise behavior change. Res Q Exerc Sport.
13. Okun MA, Ruehlman L, Karoly P, Lutz R, Fairholme C, 1992;63:60–66.
Schaub R. Social support and social norms: do they both con- 34. Godin G. A simple method to assess exercise behavior in the
tribute to predicting leisure-time exercise? Am J Health Behav. community. Can J Appl Sport Sci. 1985;10:141.
2003;27:493–507. 35. Rhodes RE, Blanchard CM, Matheson DH, Coble J. Disen-
14. Glanz K, Rimer BK, Viswanath K, eds. Health Behavior tangling motivation, intention, and planning in the physical activity
and Health Education: Theory, Research, and Practice. 4th ed. San domain. Psychol Sport Exerc. 2006;7:15–27.
Francisco, CA: Jossey-Bass; 2008. 36. Rhodes RE, Courneya KS. Relationships between person-
15. Ajzen I. The theory of planned behavior. Organ Behav Hum ality, an extended theory of planned behaviour model and exercise
Decis Process. 1991;50:179–211. behaviour. Br J Health Psychol. 2003;8:19–36.
16. Montano DE, Kasprzyk D. Theory of reasoned action, the- 37. Winters ER, Petosa RL, Charlton TE. Using social cogni-
ory of planned behavior, and the integrated behavioral model. In: tive theory to explain discretionary, “leisure-time” physical exer-
Glanz K, Rimer BK, Viswanath K, eds. Health Behavior and Health cise among high school students. J Adolesc Health. 2003;32:436–
Education—Theory, Research and Practice. 4th ed. San Francisco, 442.
CA: Jossey-Bass; 2008:67–96. 38. Gyurcsik NC, Bray SR, Brittain DR. Coping with barriers
17. Bandura A. Social Foundations of Thought and Action: A to vigorous physical activity during transition to university. Fam
Social Cognitive Theory. Upper Saddle River, NJ: Prentice-Hall; Community Health. 2004;27:130–142.
1986. 39. Rhodes RE, Jones LW, Courneya KS. Extending the theory
18. Jackson C, Smith RA, Conner M. Applying extended version of planned behavior in the exercise domain: a comparison of social
of the theory of planned behavior to physical activity. J Sports Sci. support and subjective norm. Res Q Exerc Sport. 2002;73:193–
2003;21:119–133. 199.
19. Okun MA, Karoly P, Lutz R. Clarifying the contribution of 40. Kirk RE. Experimental Design: Procedures for the Behav-
subjective norm to predicting leisure-time exercise. Am J Health ioral Sciences. 3rd ed. Pacific Grove, CA: Brooks/Cole Publishing
Behav. 2002;26:296–305. Company; 1995.
20. Rhodes RE, Courneya KS. Investigating multiple conpo- 41. Little RJA, Rubin DB. Statistical Analysis With Missing
nents of attitude, subjective norm, and perceived control: an exam- Data. 2nd ed. Hoboken, NJ: John Wiley & Sons; 2002.
ination of the theory of planned behaviour in the exercise domain. 42. Huisman M. Imputation of missing item responses: some
Br J Soc Psychol. 2003;42:129–146. simple techniques. Qual Quant. 2000;34:331–351.
21. Rhodes RE, Blanchard C, Matheson DH. A multicomponent 43. Ajzen I, Driver BL. Application of the theory of planned
model of the theory of planned behavior. Br J Health Psychol. behavior to leisure choice. J Leis Res. 1992;24:207–224.
2006;11:119–137. 44. Bozionelos G, Bennett P. The theory of planned behavior as
22. Motl RW, Dishman RK, Ward DS, et al. Comparison of predictor of exercise. J Health Psychol. 1999;44:517–529.
barriers self-efficacy and perceived behavioral control for explain- 45. Bryan AD, Rocheleau CA. Predicting aerobic versus resis-
ing physical activity across 1 year among adolescent girls. Health tance exercise using the theory of planned behavior. Am J Health
Psychol. 2005;24:106–111. Behav. 2002;26:83–94.
23. Behrens TK, Dinger MK, Heesch KC, Sisson SB. College 46. Rhodes RE, Blanchard CM, Blacklock RE. Do physical ac-
students’ understanding of moderate physical activity: a qualitative tivity beliefs differ by age and gender? J Sport Exerc Psychol.
study. Am J Health Stud. 2005;20:129–134. 2008;30:412–423.
24. Blanchard CM, Rhodes RE, Nehl E, Fisher J, Sparling PB, 47. Dishman RK, Sallis JF, Orenstein MR. The determi-
Courneya KS. Ethnicity and the theory of planned behavior in the nants of physical activity and exercise. Public Health Rep.
exercise domain. Am J Health Behav. 2003;27:579–591. 1985;100:158–171.

VOL 62, APRIL 2014 183


Beville et al

48. Shropshire J, Carroll B, Yim S. Primary school children’s physical activity behaviour—and are they the same? Health Educ
attitudes to physical education: gender differences. Phys Educ Sport Res. 2011;26:308–322.
Pedagog. 1997;2:23–38. 53. Dolan S. Benefits of group exercise. Available at: http://
49. Vilhjalmsson R, Kristjansdottir G. Gender differences in www.acsm.org/access-public-information/articles/2012/01/20/ben
physical activity in older children and adolescents: the central role efits-of-group-exercise. Published 2012. Accessed July 29,
of organized sport. Soc Sci Med. 2003;56:363–374. 2013.
50. Armitage CJ, Conner M. Efficacy of the theory of 54. Baller SL. An Investigation of Materialistic Values
planned behaviour: a meta-analytic review. Br J Soc Psychol. and Physical Activity Participation, Location, and Experience.
2001;40:471–499. Tuscaloosa: Health Sciences, University of Alabama, Tuscaloosa;
51. Chatzisarantis NLD, Hagger MS. Effects of a brief interven- 2010.
tion based on the theory of planned behavior on leisure-time physi-
cal activity participation. J Sport Exerc Psychol. 2005;27:470–487. Received: 19 January 2013
52. Williams SL, French DP. What are the most effective inter- Revised: 27 October 2013
vention techniques for changing physical activity self-efficacy and Accepted: 3 December 2013
Downloaded by [Newcastle University] at 05:12 28 April 2014

184 JOURNAL OF AMERICAN COLLEGE HEALTH

You might also like