Content uploaded by Alyson J. Crozier
Author content
All content in this area was uploaded by Alyson J. Crozier on Jan 25, 2018
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
Author's personal copy
Group cohesion and adherence in unstructured exercise groups
Kevin S. Spink
a
,
*
, Jocelyn D. Ulvick
a
, Alyson J. Crozier
a
, Kathleen S. Wilson
b
a
University of Saskatchewan, College of Kinesiology, 87 Campus Drive, Saskatoon, SK, Canada, S7N 5B2
b
California State University, Fullerton, USA
article info
Article history:
Received 11 July 2013
Received in revised form
20 November 2013
Accepted 26 November 2013
Available online 17 December 2013
Keywords:
Cohesion
Adherence
Exercise
Context
abstract
Objectives: Previous research has reported a positive relationship between perceptions of cohesion and
adherence within structured exercise settings. Given that the social determinants of adherence can vary
across situations, this study aimed to examine the cohesioneadherence relationship in unstructured
exercise settings.
Design: This study employed a cross-sectional design.
Methods: Young adults (N¼125) recalled an unstructured exercise group where they had been par-
ticipants, and then rated their perceptions of cohesion with respect to that group as well as reported the
number of times/month they had been active in that group.
Results: Regression results revealed that cohesion was significantly related to adherence. Individuals
who reported higher levels of task and lower levels of social cohesion, with both dimensions of cohesion
reflecting the perceptions of the group as a totality, attended more sessions.
Conclusions: These findings extend research reporting that the cohesiveness perceived in a structured
exercise group is related to adherence. However, there were two findings that were not consistent with
previous research. The failure of the task dimension associated with satisfying personal needs and ob-
jectives to emerge as well as the emergence of a negative relationship with one of the social dimensions
of cohesion suggest that the relationship between cohesion and adherence may play out differently in an
unstructured versus structured setting with young adults.
Ó2013 Elsevier Ltd. All rights reserved.
A growing body of evidence indicates that individuals’exercise
participation is influenced by those around them. For instance,
numerous studies have shown that perceptions about the cohe-
siveness of an exercise group impacts whether or not individual
members adhere to the group (cf. Burke, Carron, & Shapcott, 2008).
Exercise researchers have spent the last two decades examining the
construct of cohesion across a wide range of populations, including
youth (Bruner & Spink, 2011), adults (Kwak, Kremers, Walsh, &
Brug, 2006), and older adults (Watson, Martin Ginis, & Spink,
2004). For the most part, these cohesion studies have focused on
individuals exercising within structured groups (e.g., fitness clas-
ses). However, what is absent from this body of literature is the
consideration that not all people who exercise with others do so in
structured settings (cf. Spink et al., 2006).
This void is notable given the popularity of the unstructured
exercise context. For example, among Canadian adults, more than
half (56%) are active in unstructured settings (Canadian Fitness &
Lifestyle Research Institute, 1997). Similarly, the United States
Department of Labor (2008) listed Americans’most “popular”
forms of exercise as activities that are presumably conducted in
unstructured settings, such as walking, weight lifting, and using
cardiovascular equipment. Given its popularity and prevalence,
examining physical activity in an unstructured setting appears
warranted.
Not only is the lack of attention paid to unstructured exercise
groups an apparent oversight, but also it is possible that the nature
of the relationship between group cohesion and individual adher-
ence to the group is situation-specific. As argued by Reis (2008) on a
broader social level, basic processes of social behavior are not
acontextual, and he suggests that it would not be wise to examine
them in this way. Not only has this sentiment underscored the
importance of distinguishing between different contexts (e.g.,
structured versus unstructured group settings), but also within
physical activity settings, this particular distinction appears to
result in different outcomes. Spink et al. (2006) found that the
psychosocial correlates associated with activity participation
differed across structured and unstructured activity settings. Inso-
far as the group is concerned, previous research also indicates that
group variables play out differently as a function of varying exercise
contexts. For instance, it was found that task dimensions of
*Corresponding author. Tel.: þ1 306 966 1074.
E-mail address: kevin.spink@usask.ca (K.S. Spink).
Contents lists available at ScienceDirect
Psychology of Sport and Exercise
journal homepage: www.elsevier.com/locate/psychsport
1469-0292/$ esee front matter Ó2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.psychsport.2013.11.008
Psychology of Sport and Exercise 15 (2014) 293e298
Author's personal copy
cohesion were related to exercise participation in university set-
tings whereas it was the social measures of cohesion that were
related to adherence in private fitness clubs (Spink & Carron, 1994).
In terms of the structured/unstructured comparison, there also
would appear to be some key distinctions that might suggest that
cohesion could differ by context in the activity setting (Spink et al.,
2006). One distinction concerns how the two groups typically form
within each setting. Specifically, structured exercise groups (e.g.,
fitness class at a gym, running club) are ones in which the type,
time, place, and frequency of exercise participation are pre-
determined by someone else outside of the group. In contrast,
unstructured exercise groups (e.g., lifting weights with coworkers,
walking with family or friends) are characterized by the fact that
the type, time, place, and frequency of exercise participation are set
by individual group members. Whereas members of a structured
exercise group simply have to show up to a pre-scheduled exercise
class, members of an unstructured exercise group have to cohere
around factors related to group involvement in order to ensure that
the group will meet and what will be done (e.g., choosing a time
and place, agreeing on the specific activity). This implies a much
greater degree of interdependence in that individuals will likely be
influencing each other’s activities and outcomes right from the
outset of the proposed activity interaction (Reis, 2008). With this
greater interdependence in unstructured settings, it is possible that
the perceptions that members hold toward the cohesiveness of the
group will align more with the group as a totality. This would
contrast with exercising in a structured setting, where adhering to
the group may be more associated with satisfying individual needs
and objectives. This separation of the group as a totality versus
individual needs has been a longstanding distinction made in the
group literature (Zander, 1971).
A second distinction relates to differences that may exist in
terms of the member interaction that occurs within the two exer-
cise contexts. It is possible that members in the unstructured group
may have more interactionwith each other simply based on the fact
that there is more necessity to interact. In unstructured settings,
typically multiple interactions are required to allow the activity
session to run efficiently and for the goals of the activity to be
realized. Whether it is the reciprocity required to use the equip-
ment in the gym, the need to decide which trail to take, or even the
decision to end the session, all of these activities require interaction
between members. While some structured exercise groups might
support interaction and communication among individual group
members (e.g., circuit classes, boot camps, run clubs), the norm is
that many offer little to no opportunity for group member in-
teractions (e.g., step aerobics, indoor cycling, yoga, Pilates). Once
again, this assumed increased communication and interaction
found in unstructured activity settings might heighten the inter-
dependence among members in this group environment, and
possibly change how members perceive the cohesiveness with the
group (i.e., focusing on group versus individual perceptions of
cohesion).
In recognizing that there may be contextual differences between
structured and unstructured groups insofar as the salience/emer-
gence of different cohesion perceptions, the current study sought to
extend previous research by examining the relationship between
cohesion and exercise adherence in unstructured exercise groups.
Carron, Widmeyer, and Brawley’s (1985) conceptual model of
cohesion has been used as a framework for understanding cohesion
in the exercise setting, and one of the model’s assumptions cap-
tures the interdependence distinction highlighted above. Specif-
ically, the cohesion model suggests that the cognitions that
members hold about the cohesiveness of their group relate to
perceiving the group in terms of satisfying personal needs and
objectives as well as perceiving the group as a totality. The former
construct (related to personal needs) is captured by the group
member’s attraction to the group (ATG); the latter construct
(related to group totality) is captured by member’s perceptions of
group integration (GI). Studies with young adults examining group
cohesion from this perspective in a structured exercise setting
typically find that individuals who report greater adherence to the
group are the ones who perceive cohesiveness more in terms of
satisfying personal needs and objectives (ATG factors) than
perceiving the group as a totality. This finding has emerged across a
number of adherence outcomes, including attendance (e.g., Carron,
Widmeyer, & Brawley, 1988), dropout (Spink & Carron, 1994), and
lateness (e.g., Spink & Carron, 1992, 1993).
However, as noted above, no studies have examined this rela-
tionship in an unstructured exercise setting. Given the proposition
that there may be increased interdependence among members in
an unstructured setting, it might be assumed that the member’s
perceptions of the activity group as a totality could be more
important than a member’s personal reasons when formulating
perceptions of cohesiveness. Thus, it was predicted that the group
integration (GI) factors would be more salient in unstructured ex-
ercise groups, and consequently, better predictors of adherence,
than attraction to the group (ATG) factors.
Another key assumption made in the Carron et al. (1985) con-
ceptual model of cohesion is the need to separate the task- versus
socially-oriented concerns of groups and their members. The task
concerns of a group are associated with task accomplishment while
the social concerns have to do with developing and maintaining
social relationships. In terms of the task/social distinction, the task
aspect of cohesion appears to be more salient in structured exercise
settings. Specifically, it has been found that an individual’s attrac-
tion to the group’s task (ATG-Task) typically is the cohesion factor
most strongly associated with adherence in a structured exercise
setting (Annesi, 1999; Spink & Carron, 1992, 1993). However,
whether this focus on task concerns would translate to an un-
structured setting is less clear.
In terms of possible scenarios, it could be argued that the task
elements of the group would relate to adherence as they have in
structured exercise groups, given that the members of the group
are ostensibly coming together for task reasons (e.g., to increase
fitness). However, the alternative, that the social aspects of the
group may feature more prominently, also is a possibility given that
members in unstructured exercise groups are likely to interact
more, which would provide greater opportunity for social re-
lationships to develop (Carron & Brawley, 2008). Given these
equally probable outcomes, no specifica priori hypotheses were
advanced concerning whether task or social dimensions would
underpin the cohesion perceptions most associated with adherence
in the unstructured exercise setting.
Methods
Participants and design
Individuals in this study were part of a larger online study
(N¼581) examining physical activity behavior across a number of
different settings. Only individuals (average age ¼25.1 years) who
reported participating in an unstructured exercise setting with
others in the last six months (N¼125) were included in the current
study. The majority (62.7%) of participants was female. The study
was cross-sectional in design.
Procedures
Ethical approval was granted by the University Institutional
Ethics Review Board. Participants were recruited via class
K.S. Spink et al. / Psychology of Sport and Exercise 15 (2014) 293e298294
Author's personal copy
announcements, e-mail announcements, and online notices posted
on campus and local community organization websites (e.g.,
campus-wide bulletin). Those interested in participating in the
study either accessed the survey using the webpage link provided
within the e-mail announcement or online notice, contacted the
researchers directly via e-mail, or provided their e-mail address in
class to obtain the survey link. Consent was provided prior to
participation, and the survey took approximately 10e15 min to
complete.
Measures
As a set-up for the measures, participants were asked to identify
one specific activity setting where they were active with others in
the last six months. After reporting the activity, participants also
selected the type of setting where the activity occurred from a list
of provided options: fitness class you signed up for, running club,
sports team (recreational/competitive), unstructured fitness activ-
ity done with others (e.g., weights, running), or other. Only in-
dividuals who selected the “unstructured fitness activity”option
were retained for further analyses. In addition, participants were
asked to identify the primary reason why they chose to be active in
that setting. Response options included: to be motivated/chal-
lenged, to receive assistance, to socialize, to be with friends, and
other. Participants also were asked to report the average number of
participants who participated with them each time in this un-
structured setting (see Table 1 for possible responses).
Adherence
Adherence can take many forms (Steers & Rhodes, 1978). The
adherence measure of interest in the current study was frequency
of group exercise (i.e., times active in chosen activity per month).
Studies conducted in structured exercise settings often examine
percent attendance (i.e., number of exercise classes attended
divided by total possible number of classes; Spink, Wilson, & Priebe,
2010). However, self-reported attendance may be less reliable in
the unstructured exercise setting as the total number of sessions
that could be attended may not be known with any certainty. Thus,
frequency was deemed a more appropriate adherence outcome to
examine. To assess frequency, participants were asked to think
about the unstructured exercise group that they had self-identified,
and then respond to the open-ended question, “Thinking about the
last 6 months, how many times were you active in the specified
activity setting in a typical month?”
Cohesion
A modified version of the Group Environment Questionnaire
(GEQ; Carron et al., 1985) was used to assess cohesiveness. The
instrument contains 18 items that assesses four dimensions of
cohesion: Individual Attractions to the Group-Task (ATG-Task), In-
dividual Attractions to the Group-Social (ATG-Social), Group
Integration-Task (GI-Task), and Group Integration-Social (GI-So-
cial). The ATG-Task scale assesses an individual’s attraction to the
group’s task, productivity, and goals and objectives. The ATG-Social
scale reflects an individual’s personal desire to remain within the
group for social reasons (e.g., friendship opportunities). The GI-Task
scale is a measure of the individual’s perceptions of the similarity,
closeness, and bonding within the group around the task, while the
GI-Social scale reflects the perceptions of similarity, closeness, and
bonding within the group around social orientations.
Originally designed to assess cohesion in sport teams, the GEQ
has been successfully modified to reflect the nature of an exercise
group (Carron & Spink, 1992). Slight changes in the frame of
reference of this modified version highlight the fact that subjects
are providing responses for their perception of an exercise group
rather than a sports team. For example, one of the ATG-Task
questions was changed from, “This team provides me enough op-
portunities to improve my personal performance”to “This group
gives me enough opportunities to improve my personal fitness”.
This version has been used in other exercise studies and shown
good internal reliabilities (e.g., Carron & Spink, 1992;Courneya &
McAuley, 1995), and reflects similar values to those reported for
the original GEQ (Carron et al., 1985). Further, as this study was part
of a larger study involving both sport and exercise groups, we also
modified the GEQ to say “this activity group”as opposed to exercise
group.
Data analyses
In order to assess the relationship between cohesion and
adherence in unstructured exercise settings, a multiple regression
was utilized. Prior to analyses, the data were screened and all
multiple regression assumptions were met. Furthermore, alpha
values for the cohesion dimensions were calculated. Within the
regression equation, the four subscales of cohesion (ATG-Task, ATG-
Social, GI-Task, and GI-Social) were entered simultaneously as in-
dependent variables, while frequency (number of times partici-
pating in this group during a typical month) was entered as the
dependent variable.
Results
On average, participants reported exercising with their self-
identified exercise group 11.34 times (SD ¼7.30) in a typical one-
month period. Table 1 includes characteristics about the unstruc-
tured group settings reported by the participants. Individuals re-
ported engaging in a variety of unstructured activities with others,
such as working out at the gym, going for a run/walk, or lifting
weights. The total number of individuals (i.e., group size) in the self-
identified settings ranged from two to more than 10. The reasons
that participants gave for choosing that particular setting also
varied from the need to be motivated/challenged by group mem-
bers to the need to receive assistance from group members. Of
interest, reasons for choosing a particular setting did not differ
across groups of different sizes. An amalgam of the characteristics
Table 1
Descriptive statistics of participants and the unstructured group settings.
Variable Mean (SD) or frequency (%)
Participant background
Age 25.1 years (SD ¼8.0)
Gender
Male 45 (35.7%)
Female 79 (62.7%)
Setting specifics
Types of activities
Gym 48 (38.1%)
Running 26 (20.6%)
Weights 16 (12.7%)
Walking 9 (7.1%)
Biking 8 (6.3%)
Other (e.g., swimming, hiking) 12 (9.5%)
Average number of participants
2 87 (69.0%)
3e5 23 (18.3%)
6e10 7 (5.6%)
>10 9 (7.2%)
Reasons for choosing particular setting
Motivated/challenged 73 (57.9%)
Be with friends 14 (11.1%)
Socialize 5 (4%)
Receive assistance 4 (3.2%)
Other (e.g., fun) 30 (23.8%)
K.S. Spink et al. / Psychology of Sport and Exercise 15 (2014) 293e298 295
Author's personal copy
describing an unstructured setting reveals that the most typical
scenario reported by participants was attending the gym with one
other person because it was motivating and challenging (see
Table 1).
In terms of participants’ratings about their perceptions of
cohesion, mean scores assessed on a 9-point scale ranged from 6.77
(SD ¼1.80) for the GI-Task dimension to 7.41 (SD ¼1.42) for the
ATG-Task dimension. As well, Cronbach’s alpha values for the four
cohesion subscales ranged from .68 (ATG-Task) to .88 (GI-Social).
These reliability scores are consistent with the norm values re-
ported for the GEQ instrument when it was being developed for use
in the sport setting (Widmeyer, Brawley, & Carron, 1985)aswellas
its later use in the exercise setting (e.g., Carron & Spink, 1992).
Regression analyses revealed that perception of cohesiveness
was a significant predictor of exercise frequency in unstructured
exercise settings, F(4,121) ¼3.13, p¼.02, accounting for 9% of the
variance. Examination of the standardized beta values (see Table 2)
indicated that GI-Task and GI-Social were significant predictors of
adherence. Of interest, GI-Task was a positive predictor of exercise
frequency (
b
¼.44, p¼.004, semipartial correlation ¼.25) while
GI-Social was a negative predictor (
b
¼.31, p¼.03, semipartial
correlation ¼.19). Both the ATG-Task (
b
¼.21, p¼.13) and ATG-
Social (
b
¼.30, p¼.06) dimensions of cohesion were not signif-
icant predictors of adherence, although ATG-Social approached
significance.
Discussion
The aim of this study was to examine the relationship between
perceptions of cohesion and exercise adherence in unstructured
exercise groups. This study sought to replicate and extend previous
research that has examined the construct of cohesion in the
structured exercise setting. Findings indicated that, in line with
previous cohesion research, individuals’perceptions about the
cohesiveness of their exercise group were related to adherence in
that group context.
It should be noted, however, that the results of this study added
to the extant literature by examining groups in unstructured ex-
ercise settings. To our knowledge, all studies to date examining
cohesion in exercise groups using young adults have focused on the
structured setting. Given that individuals exercise with others in
both structured and unstructured group settings (Canadian Fitness
& Lifestyle Research Institute, 1997), and the effect of group vari-
ables may be contextually-oriented (cf. Spink & Carron, 1994), un-
derstanding the relationship between cohesion and exercise
adherence in the unstructured group context also was deemed
important. Thus, not only did this study proffer continued support
for the role of cohesion in exercise involvement, but also it made
the additional contribution of exploring the cohesioneexercise
adherence relationship in this important new context.
Although the results of this study found that cohesion was
related to exercise participation, there were two key differences
that emerged relative to previous cohesion research. The first re-
lates to the distinction between capturing the member’s percep-
tions of cohesion from the perspective of the individual versus the
group as a totality. Previous studies with young adults have shown
that individuals’attractions to the group (ATG) are most salient to
their participation in structured exercise classes (e.g., Annesi, 1999;
Spink & Carron, 1992, 1993). However, as hypothesized, this study
found that perceptions about the group’s integration (GI) were
more important. Specifically, results indicated that both GI factors
(GI-Task and GI-Social) were related to how frequently participants
reported exercising in their unstructured group. As previously
described, it was assumed that ongoing interdependence would be
needed to ensure that unstructured groups function effectively, and
this might serve to unite the members around the group as a to-
tality. Given that GI factors reflect the closeness and bonding of the
group as a whole, it may not be surprising that this aspect of
cohesion would be more likely to emerge in unstructured exercise
groups. By way of contrast, coordinating around aspects of group
formation and maintenance may be less important in a structured
exercise group since the time, place, and activities of the group are
set by someone outside of the group (Spink et al., 2006). As such,
the emergence of GI cohesion factors in this context may be less
likely.
The second inconsistency between this study and previous
research relates to how the GI factors were specifically related to
exercise frequency. The current results indicated that social cohe-
sion (GI-Social) was negatively related to exercise frequency. Given
that group cohesion has typically been portrayed as having a pos-
itive relationship with exercise adherence (Burke et al., 2008), the
fact that the participants in this study who felt more socially in-
tegrated with their fellow exercise group members (GI-Social) re-
ported the lowest frequency of exercise with their group was
revealing.
One potential explanation for this negative relationship could
relate to one of the distinguishing features of the unstructured
exercise group compared to its structured counterpart. As members
of an unstructured exercise group make the decisions about when
and with whom they will meet, and what forms of exercise will be
done, it is possible that the participants in this study exercised with
their identified group primarily for task reasons. Previous research
has reported that one’s motives for exercising with others can be
either task (e.g., for challenge, to receive assistance) or social (e.g.,
to make friends, to socialize) in nature (Bostick, Spink, Bruner,
Watson, & Wilson, 2003). As groups form for a reason, and it
might be assumed that participants who are part of an unstruc-
tured exercise group come together for task reasons, it is plausible
that those who perceived their group to be overly social were less
inclined to participate in that setting (i.e., socialization distracted
from the exercise task, so they attended less). This speculation was
consistent with the results of the current study. Specifically, the
majority of participants reported what could be considered a task
reason (i.e., to be motivated or challenged) as the primary motive
for participating in the identified setting over social reasons such as
“being with friends”or “to socialize”(see Table 1). Further, it should
be noted that these motives did not differ based on the size of the
group that participants had identified. Should future research be
able to replicate the finding that social cohesion relates to
decreased adherence behavior in unstructured exercise groups,
consideration of member’s motives for being in the group might be
important.
Although this study found some differences between what has
been reported previously in structured exercise groups, one key
finding that aligned with past research is that task cohesion was
positively related to exercise frequency. That is, individuals who felt
more integrated with fellow group members around the task ele-
ments of their unstructured exercise group reported participating
in more exercise sessions with the group. That task cohesion
emerged in this study was not surprising given previous exercise
research has reported a positive relationship between perceptions
Table 2
Linear regression analysis for cohesion predicting exercise frequency.
Independent variable BBSE
b
95% CI p
ATG-Task 1.07 .71 .21 [.33, 2.48] .13
ATG-Social 1.30 .69 .30 [2.66, .06] .06
GI-Task 1.80 .61 .44 [.59, 3.02] .004
GI-Social 1.27 .58 .31 [2.43, .12] .03
K.S. Spink et al. / Psychology of Sport and Exercise 15 (2014) 293e298296
Author's personal copy
of task cohesion and various measures of adherence (e.g., Annesi,
1999; Spink & Carron, 1992, 1993).
Although this study makes a unique contribution to the litera-
ture by extending our understanding of the cohesioneadherence
relationship to unstructured exercise groups, there are limitations
worth noting. One concerns the generalizability of study findings.
At one level, we used a convenience sample. Given that participants
knowingly volunteered for a study about participating in physical
activity with others, it is possible that these results may not
generalize to the entire population. For example, because inclusion
criteria required participants to have participated with others in an
unstructured exercise activity at least once in the previous month,
it is possible that only those who had an enjoyable experience
volunteered to participate. This could have introduced a systematic
bias that might limit generalizability to the entire population. At
another level, the effects found may be gender specific as a majority
of the participants was female, so additional research with males
may be warranted.
Another issue related to generalizability concerns the size of the
unstructured exercise groups examined. Even though the groups
that participants identified ranged in size from two to more than 10
members, the majority (69%) of the unstructured settings identified
involved dyads (see Table 1). Given that more than two-thirds of
participants reported on a dyad, it may be difficult to generalize the
current findings to larger unstructured groups at this time. How-
ever, it is worth noting that we conducted a secondary analysis that
examined whether group size (i.e., dyad/non-dyad) would moder-
ate the cohesioneadherence relationship. Results revealed no sig-
nificant interactions (all p’s>.1) across any of the dimensions of
cohesion.
The fact that the majority of participants self-identified a dyad
also warrants further comment. While there is some debate among
researchers about whether or not dyads can indeed be classified as
groups (cf. Moreland, 2010), we embrace Williams’(2010) sugges-
tion that dyads are a special form of group. Although they may differ
somewhat in structure (i.e., group size), Williams (2010) argued that
many of the principles used to understand and explain group pro-
cesses in groups of three or more can also be applied to dyads. To
date, few exercise researchers have considered groups with as few
as two members. Although Spink and Carron (1995) identified
group size to be an important correlate of group cohesion in exer-
cise groups, such that members of smaller structured groups tended
to report greater perceptions of cohesiveness, they broadly oper-
ationalized “small”groups as those with fewer than 20 members.
Considering that a majority of the adults in the current study re-
ported exercising within a dyadic group, we suggest that increased
attention be paid to the role of dyads in future group exercise
research in the unstructured setting. One possibility in this regard
might involve examining perceptions about the “other half”of the
dyad. Future studies interested in examining dyadic exercise groups
could benefit from examining both members of the dyad in order to
glean more information about the group processes of interest.
A second limitation involves the use of a self-report measure of
adherence. It should be acknowledged that our strategy of having
participants retrospectively report on activity involvement with
their self-identified exercise group might have introduced potential
issues of recall or response bias (Prince et al., 2008). However, given
that this study was the first attempt to examine physical activity in
unstructured exercise groups, and securing objective attendance
data in an unstructured setting would be very difficult, it was felt
that a self-report measure would be appropriate for gaining insight
into this novel population.
A third limitation concerns the inability of this study to infer
directionality. Although the results offered support for the cohe-
sioneadherence relationship in unstructured exercise groups, the
concurrent study design precludes any inference of a causeeeffect
relationship. To address causality, future research could examine
the effects of group cohesion on member adherence via an exper-
imental study design.
Despite these limitations, this study contributes to our under-
standing of the role of cohesion in exercise adherence by broad-
ening our research focus to groups in the unstructured exercise
context. As an aside, it was unexpected that most of the participants
in this study elected to report on the nature of their membership in
a dyadic exercise group. While we have argued that differences
between the current findings and what has been reported in earlier
cohesion studies focused on structured exercise groups reflected
contextual differences (i.e., structured/unstructured groups), the
emergence of so many dyads in our study begs the question of
whether some of the explanation for the present differences may
have resulted from a non-dyadic/dyadic distinction. As this possi-
bility cannot be ruled out owing to our design, it presents some
intriguing questions for future researchers to consider.
Acknowledgments
This work was supported by Doctoral Canada Graduate Schol-
arships to the second and third authors from the Social Sciences
and Humanities Research Council of Canada.
References
Annesi, J. J. (1999). Effects of minimal group promotion on cohesion and exercise
adherence. Small Group Research, 30, 542e557.
Bostick, J., Spink, K. S., Bruner, M., Watson, J., & Wilson, K. (October, 2003). Exer-
cising in groups: what’s the buzz? In Paper presented at the 2003 Canadian
Society for psychomotor learning and sport psychology conference. Hamilton, ON.
Bruner, M. W., & Spink, K. S. (2011). Team building and adherence in youth exercise
participants. Group Dynamics: Theory, Research, and Practice, 15,161e172.
Burke, S. M., Carron, A. V., & Shapcott, K. M. (2008). Cohesion in exercise groups: an
overview. International Review of Sport and Exercise Psychology, 1,107e123.
Canadian Fitness & Lifestyle Research Institute. (1997). How structured is our
physical activity (Bulletin 22). Retrieved from http://www.cflri.ca/media/node/
167/files/pip22.pdf.
Carron, A. V., & Brawley, L. R. (2008). Group dynamics in sport and physical activity.
In T. S. Horn (Ed.), Advances in sport psychology (pp. 213e237). Champaign, IL:
Human Kinetics.
Carron, A. V., & Spink, K. S. (1992). Internal consistency of the Group Environment
Questionnaire modified for an exercise setting. Perceptual and Motor Skills, 74,
304e306.
Carron, A. V., Widmeyer, W. N., & Brawley, L. R. (1985). The development of an
instrument to assess cohesion in sport teams: the Group Environment Ques-
tionnaire. Journal of Sport Psychology, 7, 244e266.
Carron, A. V., Widmeyer, W. N., & Brawley, L. R. (1988). Group cohesion and indi-
vidual adherence to physical activity. Journal of Sport & Exercise Psychology, 10,
127e138 .
Courneya, K. S., & McAuley, E. (1995). Cognitive mediators of the social influence-
exercise adherence relationship: a test of the theory of planned behavior.
Journal of Behavioral Medicine, 18,499e515.
Kwak, L., Kremers, S., Walsh, A., & Brug, H. (2006) . How is your walking group
running? Health Education, 106,21e31.
Moreland, R. L. (2010). Are dyads really groups? Small Group Research, 41,251e267.
Prince, S. A., Adamo, K. B., Hamel, M. E., Hardt, J., Connor Gorber, S., & Tremblay, M.
(2008). A comparison of direct versus self-report measures for assessing
physical activity in adults: a systematic review. International Journal of Behav-
ioral Nutrition and Physical Activity, 5,56.
Reis, H. T. (2008). Reinvigorating the concept of situation in social psychology.
Personality and Social Psychology Review, 12,311e329.
Spink, K. S., & Carron, A. V. (1992). Group cohesion and adherence in exercise
classes. Journal of Sport & Exercise Psychology, 14,78e86.
Spink, K. S., & Carron, A. V. (1993). The effects of team building on the adherence
patterns of female exercise participants. Journal of Sport & Exercise Psychology,
15,39e49.
Spink, K. S., & Carron, A. V. (1994). Group cohesion effects in exercise groups. Small
Group Research, 25,26e42.
Spink, K. S., & Carron, A. V. (1995). The group size-cohesion relationship in minimal
groups. Small Group Research, 26,86e105 .
Spink, K. S., Shields, C. A., Chad, K., Odnokon, P., Muhajarine, N., & Humbert, L.
(2006). Correlates of structured and unstructured activity among sufficiently
active youth and adolescents: a new approach to understanding physical ac-
tivity. Pediatric Exercise Science, 18, 203e215.
K.S. Spink et al. / Psychology of Sport and Exercise 15 (2014) 293e298 297
Author's personal copy
Spink, K. S., Wilson, K. S., & Priebe, C. S. (2010). Groupness and adherence in
structured exercise settings. Group Dynamics: Theory, Research, and Practice, 14,
163 e173 .
Steers, R. M., & Rhodes, S. R. (1978). Major influences on employee attendance: a
process model. Journal of Applied Psychology, 63,391e407.
United States Department of Labor. (2008). Spotlight on statistics: Sports and exercise.
Report of the Bureau of Labor Statistics. Retrieved from http://www.bls.gov/
spotlight/2008/sports/pdf/sports_bls_spotlight.pdf.
Watson, J. D., Martin Ginis, K. A., & Spink, K. S. (2004). Team building in an exercise
class for the elderly. Activities, Adaptation & Aging, 28,35e47.
Widmeyer, W. N., Brawley, L. R., & Carron, A. V. (1985). The measurement of cohesion
in sport teams: The Group Environment Questionnaire. London, ON: Sport
Dynamics.
Williams, K. D. (2010). Dyads can be groups (and often are). Small Group Research,
41, 268e274.
Zander, A. F. (1971). Motives and goals in groups. New York: Academic Press.
K.S. Spink et al. / Psychology of Sport and Exercise 15 (2014) 293e298298
