Perceived coaching leadership of youth athletes in Singapore.
Pyun, Do Young ; Kwon, Hyungil Harry ; Koh, Koon Teck 等
Leadership is one of the most significant components in human
resource management in corporations as well as in sports teams.
Leadership in athletics refers to the process of inspiring or
influencing athletes of a team to perform their tasks enthusiastically
and competently to meet the team's goals (Bridges & Roquemore,
1996), and it has been often considered as a main reason for success and
failure of an athlete or a sport team (Ch'ng & Koh-Tan, 2006).
Accordingly, leadership has been heavily studied by sport management
scholars (e.g., Beam, Serwatka, & Wilson, 2004; Chelladurai,
Imamura, Yamaguchi, Oinuma, & Miyauchi, 1988; Jambor & Zhang,
1997; Riemer & Chelladurai, 1995). One of the most popular
approaches to study sport leadership is the Multidimensional Model of
Leadership (MML) framework developed by Chelladurai (1978).
Multidimensional Model of Leadership
Chelladurai's (1978) MML synthesized several traditional
situation theories to the athletic context, such as the contingency
model emphasizing the leader characteristics (e.g., Fiedler, 1967), the
path-goal theory for the members needs and desires (e.g., House, 1971;
House & Dessler, 1975), and the adaptive-reactive theory focusing on
the organizational demands (e.g., Osborn & Hunt, 1975). The MML
differentially and equally accounts for the characteristics of the
leader (e.g., coach), the members (e.g., athletes), and the situation
(e.g., team) in an athletic setting (Chelladurai, 1993). The model
assumes that organizational performance and member satisfaction are a
function of the congruence among three states of leader behaviors--(a)
required, (b) preferred, and (c) actual (Chelladurai, 1993). The
predictors of the three leader behaviors are (a) organizational demands
and constraints placed by situational characteristics, (b) members'
preferences for specific behaviors as a function of members'
characteristics, and (c) leaders' characteristics with personality,
ability, and experience (Chelladurai, 1993). For the measurement of the
three different leader behaviors in sport, Chelladurai and Saleh (1980)
developed the Leadership Scale for Sports (LSS).
The Leadership Scale for Sports
Upon the theoretical framework of leadership (Chelladurai, 1978),
Chelladurai and Saleh (1980) employed the three different samples from
physical education students and varsity athletes and developed a
five-dimensional LSS with 40 items. The LSS included five behavioral
dimensions of leadership (i.e., training instruction, democratic,
autocratic, social support, and positive feedback) in each of the three
different versions: athletes' perceptions, athletes'
preferences, and coaches' self-perceptions in coaching behaviors
(Chelladurai & Saleh, 1980). Training and instruction corresponds
the dimension of initiating structure. Both social support and positive
feedback reflect the dimension of consideration, which were developed by
the Ohio State University researchers (e.g., Halpin & Winer, 1957).
The difference between social support and positive feedback is that
social support is provided regardless of athletic performance but
positive feedback is given upon good performance (Chelladurai, 1984).
The other two, democratic and autocratic behaviors refer to the
coach's decision making styles (Chelladurai, 1984).
Although an internal consistency problem in autocratic behavior
subscale has been reported (e.g., Beam et al., 2004; Chelladurai et al.,
1988; Zhang, Jensen, & Mann, 1997), the LSS has been believed to
properly measure various leadership behaviors of coaches in different
types of sports. In addition, the scale has generated reliable results
in different countries, such as Finland (e.g., Liukkonen & Salminen,
1990), France (e.g., Lacoste & Laurencelle, 1989), Greece (e.g.,
Iordanoglou, 1990), Japan (e.g., Chelladurai et al., 1988), and Sweden
(e.g., Isberg & Chelladurai, 1990). Using the LSS, there have been
many studies on the MML to test the effects of member and situational
characteristics on the leadership behaviors. The member characteristics
influencing preferred and/or perceived leadership behaviors included
gender (e.g., Erie, 1981; Jambor & Zhang, 1997), maturity (e.g.,
Chelladurai & Carron, 1983), and personality (e.g., Chelladurai,
1978). Several situational characteristics such as task attributes
(e.g., Chelladurai, 1978), sports type (e.g., Chelladurai, 1984), and
culture (e.g., Chelladurai et al., 1988) were revealed to influence
members' leadership behaviors. However, no study related to
leadership has been done on athletics in Singapore.
With more emphasis from the Singapore government to achieve
sporting excellence at the world's sporting map, besides having
excellence infrastructure and sound elite athletics development system,
there is also a need to examine leadership behaviors employed by the
successful coaches during training and competitions. This is because
good leadership exhibited by coaches will help establish good
athlete-coach-relationship, and in turn, have a potential to enhance
performance. One example is the leadership showed by the Singapore table
tennis coach, Lin Guodong, who helped the Singapore women's table
tennis team won a silver medal at the 2008 Beijing Olympic Games. This
medal is particularly meaningful to Singaporeans as its last silver
medal won at the Olympics was 48 years ago. Understanding such
antecedents of leadership behaviors will provide insightful instructions
to administrators, coaches, and athletes in Singapore and offer a
significant value to the current literature.
According to a recent research on coaches in schools conducted by
the Singapore Sports Council in 2004, 96.9% of the schools surveyed had
engaged "external coaches" to train their sports teams
("School Sports," 2005). Most of these coaches/instructors
have at least a Level ! coaching certificate, with level 3 being the
highest standard in Singapore. As more athletes in schools exposure to
proper coaching in a younger age, it is important to ensure that they
have an enjoyable sporting experience, while building up their technical
skills and knowledge. Accordingly, the leadership exhibited by coaches
in schools need to be examined so that early interventions can be
carried to sustain athletes' interest and help Singapore achieve
its goal in sport. Therefore, the purpose of the present study was to
examine youth athletes' perceptions toward their coaches'
behaviors in Singapore, upon the conceptual framework of the MML. The
specific focus of enquiry was to compare differences in perceived
leadership in terms of an individual characteristic (i.e., gender), a
situation characteristic (i.e., task dependence), and their interactions
(i.e., gender by task dependence).
Gender Difference in Athletic Leadership
Earlier studies revealed gender differences with regard to
athletes' preferences in leadership behavior of coaches and
resulted in similar findings. Male athletes expected more autocratic and
social support leadership than did female athletes (Chelladurai, 1978;
Terry, 1984; Terry & Howe, 1984). Female athletes, on the other
hand, preferred to be more involved in decision-making (e.g.,
democratic) than did male athletes (Chelladurai, 1978; Chelladurai &
Arnott, 1985; Chelladurai, Haggerty & Baxter, 1989). Erle (1981)
also reported that male intramural and intercollegiate hockey players
expected more training and instruction, more autocratic behavior, more
social support, and less democratic behavior than did female players. In
addition, a recent study by Beam et al. (2004) revealed that
intercollegiate male athletes showed significantly greater preferences
for autocratic and social support. However, unlike the aforementioned
literature, female athletes showed significantly greater preferences for
training and instruction.
In terms of perceived leadership behavior, the previous literature
revealed conflicting findings in coaches' self-perceptions. Jambor
and Zhang (1997) showed that there was no difference in the
self-perceived behaviors between male and female coaches in American
junior high schools, high schools, and college levels. On the contrary,
Finnish female coaches perceived to be more instructive, supportive, and
rewarding than male coaches (Salminen, Kiukkonen, & Telama, 1990).
Such inconsistent results seemed to be derived from different levels of
competition as well as different leader characteristics (e.g.,
personality, ability, and experience) and different demands and
constraints placed by situational characteristics such as cultural
difference (Chelladurai, 1993). In addition to the coachers'
personal characteristics, Chelladurai (1993) noted that athletes'
characteristics would differentiate their preferences for coaching
behaviors, and such preferences would influence coaches' actual
behaviors in the MML. Serpa and Antunes (1989) reported that female
elite athletes perceived their coach to be task oriented (e.g., training
and instruction and rewarding behavior). Based on the conceptual support
from the MML (Chelladurai, 1978) and the empirical evidence from the
past literature (e.g., Beam, et al., 2004; Chelladurai, et al., 1989;
Terry, 1984), it was hypothesized in the Singapore athletic context that
while male student-athletes may perceive more autocratic and social
support behaviors, female student-athletes perceive more democratic
behavior.
Task Dependence in Athletic Leadership
Task dependence is defined as "the extent to which the
successful performance of a task requires interaction with other tasks
in the team, and where the unit's success is based on the
coordination of these tasks" (Chelladurai, 1984, p. 29). In
athletic settings, an individual sports is considered as an independent
task which does not require interaction with others, and a team sports
is considered as an interdependent task which requires high interactions
with others in the team. In the MML model, task dependence is considered
as a factor of situational characteristics (e.g., organizational goals
and culture) that influences athletes' preferred and perceived
coaching behaviors (e.g., Chelladurai, 1978; Riemer & Chelladurai,
1995). Chelladurai (1993) indicated that coaches are significantly
influenced by situational characteristics. For example, coaches exhibit
different leadership behaviors based on whether they are involved in
interdependent (team) sports or independent (individual) sports.
Several researchers investigated the leadership behavior
differences between athletes engaged in interdependent tasks and
independent tasks using preferred and perceived versions and have found
consistent results. Beam et al. (2004) reported that the athletes of
independent sports preferred democratic, positive feedback, and social
support behaviors more than did the athletes of interdependent sports.
Chelladurai (1978) found that independent sports athletes preferred more
democratic behavior while interdependent sports athletes preferred more
autocratic behavior and training and instruction. Liukkonen and Salminen
(1990) showed that athletes perceived their coaches in interdependent
sports to be more autocratic and less democratic, and more socially
supportive. From the Chelladurai's MML that the situational
characteristic, task dependence, influences leadership behaviors and a
drawn conclusion from the literature review, it was hypothesized that
there should be a task dependence difference, particularly for training
instruction, decision making, and social support behaviors, among
athletes in Singapore.
In summary, the present study was designed to test the several
hypotheses stated. First, the researchers attempted to see whether male
student-athletes perceived more autocratic behavior and social support,
and whether female student-athletes perceived their coaches to be more
democratic. Second, it was compared that the situational characteristic
would differentiate student-athletes' perceptions for coaching
behaviors. Specifically, the researchers hypothesized that
interdependent student-athletes perceive more training and instruction,
autocratic, and social support behaviors. Lastly, the study examined the
interaction between gender and task dependence on the five coaching
leadership behaviors.
Methods
Participants
The participants were 554 secondary school student-athletes (359
male and 185 female students) randomly selected from 21 (11.4%)
secondary schools based on the classification of schools defined by the
Ministry of Education in Singapore. All schools in Singapore are divided
into four zones, and there are a total of 185 secondary schools around
the country. The student-athletes were competing for their schools in a
wide range of sporting events organized by the Singapore Schools Sports
Council. Among the questionnaires collected, 10 were discarded due to
incompleteness. Accordingly, 544 respondents (359 male and 185 female
students) were utilized in future data analyses. The participants'
average age ranged from 13 to 18 years (M = 15.60; SD = .78), and their
average number of years representing the schools was 3.31 (SD = 1.04).
Among the respondents, 446 student-athletes participated in
interdependent sports (i.e., basketball, cricket, hockey, netball,
rugby, soccer, softball, and volleyball), and 98 student-athletes
participated in independent sports (i.e., badminton, cross-country,
fencing, track and field, and table tennis).
Measure
The athlete's perception version of the Leadership Scale for
Sports (LSS) developed by Chelladurai and Saleh (1980) was used to
evaluate student-athletes' perceived leadership behaviors of their
coaches. The LSS contains five dimensions of leader behavior with 40
items: training and instruction (13 items); democratic (nine items);
autocratic (five items); social support (eight items); positive feedback
(five items; refer to Chelladurai & Saleh, 1980 for each item
statement). Chelladurai and Riemer (1998) noted that the psychometric
properties of the LSS are adequate. A recent study tested the
reliability and validity of the LSS in the Singapore coaching contexts
(Kwon, Koh, Pyun, & Wang, in press). While the study concern the
reliability of autocratic construct, there was a fair fit of the model
to the data, [chi square] = 1493.24, [chi square]/df = 2.05, RMSEA =
.061 (CI = .057, .063,p < .01), CFI = .97, NFI = .93 (Kwon et al., in
press). In the perception version of the LSS, the items were preceded by
the statement "my coach ..." and were measured by a five-point
Likert scale, anchored with never (1), seldom (2), occasionally (3),
often (4), and always (5). In addition, the questionnaire included a
section to measure student-athletes' demographic characteristics
such as gender and task dependence for group specification.
Procedure
The schools identified were contacted via email or phone. They were
briefed on the purpose of the study and were encouraged to participate
in the study. Given permission granted by heads of PE departments, the
appointment for on-site data collection was confirmed at least one week
in advance. On the day of data collection, informed consent was obtained
from the participants and the heads, and the participants were briefed
again on the purpose of the study. Prior to questionnaire
administration, the participants were assured that their participation
in the study was completely voluntary, and they may withdraw at any time
without penalty. The third author was presented on sites to administer
the questionnaire and answer questions from the participants. All of the
participants completed the questionnaires within half an hour.
Data Analysis
The data were analyzed using the Statistical Package for the Social
Science (SPSS 15.0). For the preliminary tests, the Mahalanobis
distance, correlation, and internal consistency were conducted to test
multivariate normality, multicollinearity, and reliability of the
measures. For the main study, a multivariate analysis of variance
(MANOVA) was performed with five behavior variables as a set of
dependent variables and two independent variables of gender (male and
female) and task dependence (interdependence and independence sports).
In addition, interaction effect between two independent variables was
examined. The Wilk's Lambda was employed for the test of the
multivariate null hypothesis in the MANOVA. When the null hypothesis was
rejected, follow up univariate analyses of variance (ANOVA) were carried
out for each of the dependent variables. For the gender by task
dependence variable, the Tukey HSD tests were conducted to examine the
pairwise comparisons among the four groups.
Results
Preliminary Test
The aim of this study was to see whether secondary school
student-athletes in Singapore showed gender and task dependence
differences in five subscales of the perceived version of LSS. Before
the MANOVA was performed, the researchers checked whether the data set
satisfied the assumptions of the analyses (i.e., multivariate normality,
internal consistency, and multicollinearity for the five dependent
variables).
Multivariate normality test. A multivariate normality is a basic
assumption required for significance testing in MANOVA which is quite
sensitive to outliers (Pallant, 2005). To test for multivariate
normality the Mahalanobis distances across the five dependent variables
were calculated. In the residuals statistics, the maximum value of the
Mahalanobis distance (i.e., 25.17) was greater than the determined
critical value (i.e., [chi square] = 20.52, df = 5) at the .001 alpha
value (see Tabachnick & Fidell, 2001), suggesting that at least one
of the current cases exceeded the critical value. The further
investigation was followed to detect problematic cases by comparing the
Mahalanobis distance values of each case with the critical value. As a
result, a total of six outliers ranging from 21.10 to 25.17 removed from
the data file.
Multicollinearity test. The correlation matrix revealed that the
five dependent variables were moderately correlated, ranging from -.02
to .75. While the correlation coefficient between democratic and
positive feedback behaviors was as high as .75, it was not considered to
indicate redundant dependent measures and decrease statistical
efficiency (Pallant, 2005).
Internal consistency test. Internal consistency among the
respective items of the five leadership behaviors were tested with
Cronbach's alphas. The alpha coefficients ranged from .47 to .91:
training and instruction ([alpha] = .91), democratic ([alpha] = .85),
autocratic ([alpha] = .47), social support ([alpha] = .78), and positive
feedback ([alpha] = .79). The alpha coefficient for autocratic behavior
did not meet the .70 cut-off (Nunnally & Bernstein, 1994). The
results relating to autocratic behavior should be viewed with caution.
Multivariate Analysis of Variance
After deleting six problematic cases from the data screening, the
remaining 538 responses were used in the main analysis. The means and
standard deviations for the five leadership behavior variables in
gender, task dependence, and gender by task dependence were summarized
in Table 1. According to Table 1, the subjects' mean scores in
terms of gender, task dependence, and gender by task dependence were
slightly above the midpoint for all variables except autocratic and
social behaviors. For hypothesis testing, first, the multivariate
significance test was conducted to determine whether there were
statistically significant differences among the groups of interest on a
linear combination of the five behavior variables. For gender, the
multivariate null hypothesis of equality of the means over male
student-athletes and female student-athletes for all variables was
failed to be rejected at the .05 significance level, Wilk's
[LAMBDA] = .98, F (5, 532) = 1.94, p = .09. For task dependence, the
multivariate null hypothesis of equality of the means over
interdependent sports and independent sports student-athletes for all
variables was rejected at the .05 significance level, Wilk's
[LAMBDA] = .95, F(5,532) = 5.75, p < .001. For gender by task
dependence, the multivariate null hypothesis of equality of the means
over the four group--male interdependent sports, male independent
sports, female interdependent sports, and female independent
student-athletes--for all variables was rejected at the .05 significance
level, Wilk's [LAMBDA] = .92, F (5,532) = 2.86, p < .001. Since
the multivariate test was not significant for the gender, follow up
univariate tests were performed only on task dependence and the
interaction of gender and task dependence (see Table 2).
Task dependence. Univariate ANOVA tests for task dependence
revealed that social support behavior showed statistical difference (p
< .001). The athletes of interdependent sports perceived their
coaches to be more socially supportive than their counterpart. Although
statistically significant, the effect size ([[eta].sup.2]) was very
small (i.e., 3%).
Interaction effect. The ANOVA tests for the interaction showed
significant differences in training and instruction (p = .01), social
support (p < .001), and positive feedback (p = .02) variables. The
effect sizes ([[eta].sup.2]) for the training/instruction, social
support, and positive feedback were .02, .03, and .02, respectively. As
a next step, post hoc tests of all pairwise comparisons for the three
behaviors using the Tukey HSD were followed. First, for training and
instruction, the female interdependent sports athletes had a
significantly higher mean score than the male interdependent sports
athletes (p = .02). Second, for social support, the male interdependent
sports athletes reported a significantly higher mean score than the male
(p = .05) and female (p = .05) independent sports athletes. The female
interdependent sports athletes also had a significantly higher mean
score than the male (p = .01) and female independent sports (p = .01).
Lastly, for positive feedback, the female interdependent sports athletes
had a significantly higher mean score than the male interdependent
sports athletes (p = .01; see Table 1 for each mean score).
Discussion
The current study extended past leadership research by applying the
MML (Chelladurai, 1978) into the context of Singapore. The study
utilized a well-developed scale of the LSS (Chelladurai & Saleh,
1980) and compared the student-athletes' perceptions on leadership
behaviors of their coaches based on gender and task dependence. In
addition, considering importance of interaction between individual and
situational characteristics associated with leadership behavior
(Fiedler, 1967; House, 197 I), the study compared the
student-athletes' perceived leadership behaviors on gender by task
dependence. Understanding student-athletes' perceptions on
leadership behaviors could help coaches and sport administrators
increase athletes' satisfaction and performance (Chelladurai, 1978)
and eventually achieve sporting excellence in Singapore.
Individual Characteristic: Gender
Previous studies (e.g., Chelladurai, 1978; Serpa & Antunes,
1989; Terry, 1984; Terry & Howe, 1984) used a preference version of
the LSS and found that male athletes expected autocratic behavior
whereas female athletes expected democratic behavior from their coaches.
The present study found that male and female student-athletes did not
statistically differ in how they perceived the decision making behaviors
of their coaches although the direction of the mean scores for male and
female student-athletes turned out as hypothesized. One possible reason
could be due to the perception of athletes toward their coaches. In
Singapore, athletes usually believe that coaches employed by schools are
experts in their fields. They have trust in their coaches and hence just
have to "follow instructions" and "do what have been
told". Unlike athletes in the western countries, Singapore athletes
are not used to questioning their coaches such as why certain things are
done in certain ways, challenging the authority etc. Hence, autocratic
behaviors seem dominating most of the coaching sessions.
Situational Characteristic: Task Dependence
The past literature found that situational characteristics such as
organizational goal (e.g., Erie, 1981), task type (Beam et al., 2004),
task variability (e.g., Chelladurai, 1978) and culture (e.g.,
Chelladurai et al., 1988) were significantly associated with leader
behaviors. Among the characteristics, the task attributes of dependence
showed a strong influence on actual coaching behaviors (Chelladurai,
1978). The study also found that the interdependent and independent
sports student-athletes differed in social support behavior;
interdependent athletes showed higher perceptions on social support.
Chelladurai (1993) noted that as task dependence increases, coaches need
to increase social support for their athletes. Similarly, Liukkonen and
Salminen (1990) demonstrated that Finnish interdependent sports athletes
reported higher perceptions of social support. In the same line, closer
interpersonal relationships between athletes and coaches were found for
interdependent sports teams in Singapore.
In addition, Chelladurai (1978) noted that while coaches of
interdependent sports need to demonstrate more social support for team
athletes, it would be also effective if they could be able to exhibit
social support differentially based on team members' preferences
and needs. Thus, further research using preferred version of LSS is also
expected to provide Singapore coaches in with meaningful information
that can enhance their coaching effectiveness.
Interaction between Individual and Situational Characteristics:
Gender by Task Dependence
The examination of interaction between gender and task dependence
provides coaches with more detailed information on how to manage their
athletes when it is related to a combination of gender and type of
sports. The results showed there were significant interactions between
gender and task dependence on training and instruction, social support,
and positive behaviors. First, for training and instruction behavior,
while there was no gender difference, a group difference was found when
gender was combined with task dependence (female interdependence = 4.02;
male interdependent = 3.82, p < .05). Hersey and Blanchard's
(1984) situational leadership theory can be employed to discuss this
finding. One aspect of the theory is about the relationship between
members' job maturity and leaders' task-oriented behavior.
They indicated that when members of the organization possess low level
of skill maturity, managers need to be more task-oriented to increase
the effectiveness than relationship oriented. When athletes possess low
level of technical ability, coaches who exhibit task-oriented behaviors
are expected to be more effective. Thus, in case of female student
athletes, whose technical ability is expected to be lower than their
male counterparts, the coaches of female interdependent sports athletes
seemed to focus on task oriented leader behaviour such as providing
specific directions to each athlete as to what she should do in every
situation (Chelladurai & Saleh, 1980).
Second, for social support behavior, the result indicates that the
male interdependent sports athletes perceived more social support than
both the male and female independent sport athletes. In addition, the
female interdependent sports athletes perceived more social support
behavior than both the male and female independent sports athletes.
However, there was no difference between male and female interdependent
sports athletes. The current findings might be derived from the
different perceptions in social support behavior between interdependent
and dependent sports athletes, discussed in the previous section. Based
on the support from the previous literature (e.g., Chelladurai, 1993;
Liukkonen & Salminen, 1990), coaches' social support behaviors
might be more appropriate for athletes in the interdependent sport teams
in Singapore, regardless of athletes' gender.
Lastly, for positive feedback behavior, female interdependent
sports athletes perceived higher levels of positive feedback than did
male interdependent sports athletes. There has been no study revealing a
significant difference in perceived or preferred positive feedback
behavior in terms of athletes' gender, task dependence, or gender
by task dependence. This aspect of coaching behaviour needs to be
studied with more intensity because Chelladurai and Saleh (1980) assert
that positive feedback is essential to maintain athletes'
motivational levels. According to the cognitive evaluation theory in
education (Ryan, Connell, & Deci, 1985), one of the most important
factors influencing individuals' level of intrinsic motivation
toward particular achievement activity is positive feedback given in
response to students' performance errors. An empirical study in
athletics by Amorose and Horn (2000) revealed that college athletes with
higher levels of intrinsic motivation perceived that their coach
provided high frequencies of positive feedback. Coaches' frequent
exhibition of positive feedback (e.g., compliment, appreciation, credit,
and reward) to athletes during practice and games will increase
intrinsic motivation in athletes and eventually promote their
achievement.
Limitations and Future Research
There are some limitations of the present study that need to be
mentioned. The first limitation is related with the cell sizes of the
study. The sample sizes among the groups compared were not consistent or
similar A couple of cells had relatively small sample sizes (e.g., n =
68 for the male interdependent sport athletes group; n = 29 for the
female interdependent sports athletes group). A small sample size for a
certain group could produce an unexpectedly high standard error when
testing a difference between means. As a standard error works as a
denominator in a significance test formula, an overestimated standard
error could have resulted in decreased statistics (e.g., T or F value).
Thus, the test might have increased type II error which was the
probability of incorrectly failing to reject the null hypothesis when it
needed to reject the hypothesis. For example, the difference in training
and instruction between female interdependent (n = 155) and male
interdependent (n = 286) sports athletes was significant (a mean
difference = .20, SE = .07). However, even a bigger mean difference in
training and instruction between female interdependent (n = 155) and
female independent (n = 29) sports athletes was not significant (a mean
difference = .33, SE =. 14), due to the high standard error. In
addition, the small sample size of the groups reduces the effect size in
the tests. Accordingly, more effort to increase sample size of the
groups is necessary to enhance power of tests in a future study. Hair,
Anderson, Tatham, & Black (1998, p. 352) suggested that "...
with sample sizes of fewer than 50 members, obtaining desired power
levels can be quite problematic. Increasing sample sizes in each group
has marked effects until group sizes of approximately 150 are
reached".
Second, the reliability of autocratic behavior subscale may hurt
the validity of the results. The results revealed that the alpha
coefficient of autocratic behavior (.47) was lower than the suggested
value of .70. Most past studies utilizing the LSS have experienced the
internal consistency problem in autocratic behavior (e.g., Chelladurai
et al., 1988; Isberg & Chelladumi, 1990; Riemer & Chelladurai,
1995). Future researchers need to carefully redefine the psychometric
characteristics of autocratic behavior and to generate more
representative items for the domain. Considering this dimension's
conceptual and empirical meaningfulness as a significant dimension of
coaching behavior, Chelladurai and Riemer (1998) suggested to include
more homogeneous items representing the domain for its' internal
consistency.
Lastly, the other limitation is related to the generalizability of
the study. The current findings may not be applicable when the
observations are replicated to other types or levels of sports.
Particularly, the researchers attempted to select more popular sports in
Singapore but failed to seize several key sports which have good
potential to perform better in international competitions.
Traditionally, Singapore has shown world-level performance in the
selected sports, such as, bodybuilding, bowling, sailing, and swimming.
Recently, Singapore earned a total of six gold medals in the 2006 Doha
Asian Games. Particularly, the five medals were from sailing; the other
three were from bodybuilding, bowling, and swimming. It is deemed that
there is a high probability that Singapore would win a second Olympic
medal from one of those sports. Understanding perceptions of Singapore
elite athletes participating in the target sports in a future study will
accelerate Singapore to achieve medal glory at the Olympics in the near
future.
References
Amorose, A. J., & Horn, T. (2000). Intrinsic motivation:
Relationships with collegiate athletes' gender, scholarship status,
and perceptions of their coachers' behavior. Journal of Sport
Exercise Psychology, 22, 63-84.
Beam, J. W., Serwatka, T. S., & Wilson, W. J. (2004). Preferred
leadership of NCAA Division I and II intercollegiate student-athletes.
Journal of Sport Behavior, 27, 3-15.
Bridges, F. J., & Roquemore, L. L. (1996). Management for
athletic/sport administration: Theory and practice (2nd ed.). Decatur,
GA: ESM Books.
Chelladurai, P. (I 978). A contingency model of leadership in
athletics. Unpublished doctoral dissertation. University of Waterloo,
Waterloo, Ontario, Canada.
Chelladurai, P. (1984). Discrepancy between preferences and
perceptions of leadership behavior and satisfaction of athletes in
varying sports. Journal of Sport Psychology, 6, 27-41.
Chelladurai, P. (1993). Leadership. In R.N. Singer, M. Murphy,
& L. K. Tennant (Eds.), Handbook of research on sport psychology
(pp. 647-67 l). New York: McMillan.
Chelladurai, P., & Arnott, M. (1985). Decision styles in
coaching: Preferences of basketball players. Research Quarterly for
Exercise and Sport, 56(1), 15-24.
Chelladurai, P., & Carran, A. V. (1983). Athletic maturity and
preferred leadership. Journal of Sport Psychology, 5, 371-380.
Chelladurai, P., Imamura, H., Yamaguchi, Y., Oinuma, Y., &
Miyauchi, T. 0988). Sport leadership in a cross-national setting: The
case of Japanese and Canadian university athletes. Journal of Sport and
Exercise Psychology, 10, 374-389.
Chelladurai, P. Malloy, D., Imamura, H., & Yamaguchi, Y.
(1987). A cross-cultural study of preferred leadership in Sports.
Canadian Journal of Sports Sciences, 12, 106-110.
Chelladurai, P., Haggerty, T. R., & Baxter, P. R. (1989).
Decision style choices of university basketball coaches and players.
Journal of Sport and Exercise Psychology, 11, 201215.
Chelladurai, P., & Riemer, H. A. (1998). Measurement of
Leadership in Sport. In J. L. Duda (Ed.). Advances in sport and exercise
psychology measurement (pp. 227-253). Morgantown, WV: Fitness
Information Technology.
Chelladurai, P., & Saleh, S. D. (1980). Dimensions of leader
behavior in sports: Development of a leadership scale. Journal of Sport
Psychology, 2, 34-45.
Ch'ng, A., & Koh-Tan, A. (2006). Managing sport: Concepts
and issues of non-profit organizations. Singapore: Prentice Hall.
Erle, F. J. (1981). Leadership in competitive and recreational
sport. Unpublished master's thesis. University of Western Ontario,
London, Canada.
Fiedler, F. E. (1967). A theory of leadership effectiveness. New
York: McGraw-Hill.
Halpin, A. W., & Winder, B. J. (1957). A factorial study of the
leader behavior descriptions. In R. M. Stogdill & A. E. Coons
(Eds.), Leader behaviors: It's description and measurement (pp.
39-51). Columbus, OH: Ohio State University, Bureau of Business
Research.
Hair, J. E, Anderson, R. E., Tatham, R. L., & Black, W. C.
(1998). Multivariate data analysis (5th ed.). Upper Saddle River, N J:
Prentice-Hall, Inc.
Hersey, P., & Blanchard, K. H. (1984). Management of
organizational behavior (4th ed.). Englewood Cliffs, N J: Prentice Hall.
House, R. J. (1971). A path-goal theory of the leader
effectiveness. Administrative Science Quarterly, 16, 321-338.
House, R. J., & Dessler, G. (1974). The path-goal theory of
leadership: Some post hoc and a priori tests. In J. G. Hunt & L. L.
Larson (Eds.), Contingency approaches to leadership (pp. 29-55).
Carbondale, IL: Southern Illinois University Press.
Iordanoglou, D. (1990). Perceived leadership in Greek soccer: A
preliminary investigation. Unpublished manuscript, University of
Manchester, Department of Education.
Isberg, L., & Chelladurai, P. (1990). The Leadership Scale for
Sports: Its applicability to the Swedish context. Unpublished
manuscript, University College of Falun/Borlenge, Sweden.
Jambor, E. A., & Zhang, J. J. (1997). Investing leadership,
gender, and coaching level using the revised leadership for sport scale.
Journal of Sport Behavior, 20, 313-321.
Kwon, H. H., Koh, K. T., Pyun, D. Y., & Wang, J. (in press).
Psychometric properties of leadership scale for sport: The case of
Singaporean Secondary School Athletes. International Journal of Sport
Management.
Lacoste, P. L., & Laurencelle, L. (1989). The French validation
of the Leadership Scale for Sports. Unpublished abstract, Universite du
Quebec a Trois-Rivieres, Trois-Rivieres, Canada.
Liukkonen, J., & Salminen, S. (1990, June). The athletes
'perception of leader behavior of Finnish coaches. Paper presented
at the World congress on Sport for All, Tampere, Finland.
Osborn, R. N., & Hunt, J. G. (1975). An adaptive-reactive
theory of leadership: The role of macro variables in leadership
research. In J. G. Hunt & L. L. Larson (Eds.), Leadership frontiers
(pp. 27-44). Kent, OH: Kent State University.
Pallant, J. (2005). SPSS survival manual: A step by step guide to
data analysis using SPSS for Windows (2nd ed.). Berkshire, United
Kingdom: Open University Press.
Paton, G. (1987). Sport management research: What progress has been
made? Journal of Sport Management, 1, 25-31.
Riemer, H. A., & Chelladurai, P. (1995). Leadership and
satisfaction in athletics. Journal of Sport and Exercise Psychology, 17,
276-293.
Ryan, R. M., Connell, J. P., & Deci, E. L. (1985). A
motivational analysis of self-determination and self-regulation in
education. In C. Ames & R. E. Ames (Eds.), Research on motivation in
education: The classroom milieu (pp. 13-51). New York: Academic Press.
Salminen, S., Liukkonen, J., & Telama, R. (1990, July). The
differences in coaches' and athletes' perception of leader
behavior of Finnish coaches. Paper presented at the AIESEP Congress,
Loughborough, England.
Serpa, S., & Antunes, I. (1989). Leadership styles of elite
Portuguese women's volleyball coaches. Paper presented at the 6th
International Congress on Sport Psychology. Lahti, Finland.
School Sports--How good are school coaches? Sports Council survey
shows about a quarter of outside help have no training in coaching
(2005, April 29). The Straits Times.
Tabachnick, B. G, & Fidell, L. S. (2001). Using multivariate
statistics (4th ed.). New York: HarperCollins.
Terry, P. C. (1984). The coaching preferences of athletes at
Universiade '83. The Canadian of Applied Sport Sciences, 9,
201-208.
Terry, P. C., & Howe, B. L. (1984). The coaching preferences of
athletes. The Canadian of Applied Sport Sciences, 9, 188-193.
Do Young Pyun
National Institute of Education, Nanyang Technological University,
Singapore
Hyungil Harry Kwon
Chung-Ang University, Korea
Koon Teck Koh and Chee Keng John Wang
National Institute of Education, Nanyang Technological University,
Singapore
Address Correspondence to: Do Young Pyun, Assistant Professor,
Department of Physical Education and Sports Science, National Institute
of Education, Nanyang Technological University, 1 Nanyang Walk,
Singapore, 637616. E-mail:
[email protected].
Table 1. Descriptive Statistics (a)
Gender Task dependence
Male Female Interdependent
(n = 354) (n =184) sports (n = 441)
Training and 3.83 (.71) 3.97 (.64) 3.89 (.69)
instruction
Democratic 3.37 (.75) 3.47 (.78) 3.41 (.77)
Autocratic 2.70 (.66) 2.62 (.62) 2.68 (.64)
Social support 3.07 (.74) 3.13 (.78) 3.15 (.76)
Positive feedback 3.61 (.78) 3.80 (.74) 3.69 (.78)
Task dependence
Independent sports
(n = 97)
Training and 3.79 (.71)
instruction
Democratic 3.38 (.69)
Autocratic 2.65 (.66)
Social support 2.83 (.68)
Positive feedback 3.63 (.66)
(a) Table includes means and (in parentheses) standard deviations.
Gender by Task dependence
Male Male- Female
interdependent independent interdependent
(n=286) (n=68) (n=155)
Training and 3.82 (.72) 3.84 (.71) 4.02 (.61)
instruction
Democratic 3.36 (.76) 3.40 (.71) 3.49 (.80)
Autocratic 2.71 (.64) 2.67 (.72) 2.62 (.64)
Social support 3.12 (.75) 2.86 (.66) 3.21 (.77)
Positive feedback 3.61 (.79) 3.64 (.67) 3.84 (.74)
Gender by Task dependence
Female independent
(n=29)
Training and 3.69 (.71)
instruction
Democratic 3.33 (.68)
Autocratic 2.59 (.49)
Social support 2.75 (.71)
Positive feedback 3.60 (.67)
(a) Table includes means and (in parentheses) standard deviations.
Table 2. Univariate ANOVA Tests for Task Dependence and Gender
by Task Dependence.
Task dependence
F [[eta].sup.2]
Training and instruction 1.59 .003
Democratic .13 .00
Autocratic .19 .00
Social support 14.82 * .03
Positive feedback .50 .001
Gender * Task dependence
F [[eta].sup.2]
Training and instruction 3.55 * .02
Democratic 1.04 .01
Autocratic .71 .004
Social support 5.54 * .03
Positive feedback 3.28 * .02
* Significant at the .05 level
