A cluster analysis on students’ perceived motivational climate. Implications on
Javier Fernández-Río, Departamento de Ciencias de la Educación, Universidad de
Antonio Méndez-Giménez, Departamento de Ciencias de la Educación, Universidad de
Jose A. Cecchini Estrada, Departamento de Ciencias de la Educación, Universidad de
Universidad de Oviedo
Facultad de Formación del Profesorado y Educación
c) Aniceto Sela s/n, despacho 239
The aim of this study was to examine how students’ perceptions of the class climate
influence their basic psychological needs, motivational regulations, social goals and
outcomes such as boredom, enjoyment, effort, and pressure/tension. 507 (267 males,
240 females) secondary education students agreed to participate. They completed a
questionnaire that included the Spanish validated versions of PMCSQ-2, BPNES,
PLOC, SGS-PE, and several subscales of the IMI. A hierarchical cluster analysis
uncovered four independent class climate profiles that were confirmed by a K-Means
cluster analysis: “high ego”, “low ego-task”, “high ego-medium task”, and “high task”.
Several MANOVAs were performed using these clusters as independent variables and
the different outcomes as dependent variables. Results linked high mastery class
climates to positive consequences such as higher students’ autonomy, competence,
relatedness, intrinsic motivation, effort, enjoyment, responsibility and relationship, as
well as low levels of amotivation, boredom and pressure/tension. Students’ perceptions
of a performance class climate made the positive scores decrease significantly. Cluster 3
revealed that a mastery oriented class structure undermines the negative behavioural and
psychological effects of a performance class climate. This finding supports the buffering
hypothesis of the achievement goal theory.
Keywords: secondary education; motivation; buffering hypothesis
El objetivo del estudio fue analizar cómo las percepciones del clima de clase influyen
sobre las necesidades psicológicas básicas, las regulaciones motivacionales, las metas
sociales y diferentes consecuencias: aburrimiento, diversión, esfuerzo y presión/tensión.
Participaron 507 estudiantes de secundaria (267 varones, 240 mujeres). Completaron un
cuestionario que incluía las versiones validadas al castellano del PMCSQ-2, BPNES,
PLOC, SGS-PE y varias subescalas del IMI. Un análisis de clusters jerárquico reveló 4
perfiles de clima de clase independientes, confirmados por un análisis K-Medias de
clusters: “alto ego”, “bajo ego-tarea”, “alto ego-medio tarea” y “alto tarea”. Se
realizaron varias MANOVAS usando los clusters como variables independientes y las
Fernandez-Rio, J., Cecchini, J.A., & Mendez-Gimenez, A. (2014). A cluster analysis on students’ perceived
motivational climate. Implications on psycho-social variables. The Spanish Journal of Psychology, vol.17. (in press)
diferentes consecuencias como variables dependientes. Los resultados unieron los
climas de clase orientados a la tarea con consecuencias positivas como altos niveles de
autonomía, competencia, relación, motivación intrínseca, esfuerzo, diversión y
responsabilidad, y bajos niveles de desmotivación, aburrimiento y presión/tensión. Las
percepciones de un clima orientado al rendimiento hicieron que los resultados positivos
disminuyeran significativamente. El Cluster 3 reveló que una estructura de clase
orientada a la maestría mina los efectos negativos psicológicos y de comportamiento de
un clima de clase orientado al rendimiento. Este resultado también apuntala la hipótesis
“Buffering” de la Teoría de Meta de Logro.
Palabras clave: educación secundaria, motivación, hipótesis Buffer
Different personal and environmental elements involved in the teaching/learning
process have been researched as influential in student achievement (Braithwaite, Spray
& Warburton, 2011). There is an increasing body of evidence that connects motivation
and learning in educational settings (Ryan & Deci, 2000). The achievement goal theory
(AGT) has been the driving force to study achievement motivation in education. The
central idea is that individuals participate in any type of activity to show competence
(Nicholls, 1989). There are many different personal and situational elements that can
exert some weight on students’ motivation to learn, and the perceived motivational class
climate is one of them (Wang, Liu, Chatzisarantis & Lim, 2010).
Motivational climate can be defined as a group of implicit and/or explicit environmental
signals that determines individuals’ success or failure (Ames, 1992). Two motivational
climates have been identified in the physical domain: performance or ego-involving,
and mastery or task-involving (Ames, 1992). Performance climates emphasize
interpersonal opposition, errors are penalized, and highly normative ability is rewarded,
while mastery climates emphasize improvement and effort (Braithwhite et al., 2011).
Research has linked performance environments to maladaptive conducts such as
negative attitudes towards learning activities, cheating, disruptive behaviours, or the
belief that success is mainly the result of ability (Cervelló, Jiménez, del Villar, Ramos &
Santos-Rosa, 2004). In contrast, mastery contexts have been related to adaptive
conducts such as positive attitude towards the different tasks, active participation, and
the belief that success is a matter of effort (Wang et al., 2008).
Therefore, the existing literature connects students’ perceptions of a mastery-oriented
class climate with adaptive psychological and motivational outcomes, while it ties
maladaptive outcomes to performance-oriented class climates. From this perspective,
classroom environments are viewed as bipolar with one kind of class orientation linked
to positive and the other one to negative results. However, other researchers believe that
students can perceive several combinations of climate orientations in the classroom,
which can be related to many different motivational and achievement outcomes (Meece,
Anderman & Anderman, 2006). According to Ciani, Middleton, Summers, and Sheldon
(2010), adaptive classroom goal structures can protect against the negative effects of
performance-oriented class climates. This has been denominated the “buffering
hypothesis” of the AGT (Cianni et al., 2010), and it derives from previous researchers
who believed that adaptive climate structures can operate in an additive way
compensating the negative effects of performance-oriented climates (Duda, 2001).
Motivation has also been researched as one of the key elements related to learning
outcomes. A major theoretical framework that is being used to study motivation in
physical education settings is the self-determination theory (SDT). It identifies three
basic types of behavioural regulations: intrinsic motivation, extrinsic motivation, and
amotivation. Intrinsic motivation has been defined as doing an activity for its inherent
satisfaction, which represents the highest degree of self-determined motivation.
Extrinsic motivation is evident when individuals perform an activity because they value
its associated outcomes. Three types of extrinsic motivation have been researched:
identified regulation, introjected regulation and external regulation (Goudas et al.,
1994). Finally, amotivation can be described as lack of motivation. It arises from
feelings of personal incompetence, lack of activity value, and the belief that one’s
actions cannot influence one’s outcomes (Ryan & Deci, 2000). Regarding the
relationship between perceived class climates and motivation, Vallerand, Deci, and
Ryan (1985) considered that performance-oriented climates are motivationally
negative, because they tend to damage subject’s self-determination, whereas mastery-
oriented environments have been linked to higher levels of intrinsic motivation
(Papaioannou, Marsh & Theodorakis, 2004).
Motivation can be affected by three essential psychological needs that are directly
linked to the students’ social environment: autonomy, competence and relatedness (Deci
& Ryan, 2000). Autonomy is the desire to be the source of one’s own behaviour (Deci
& Ryan, 2000). Competence is the student’s perception of being able to show
effectiveness within a particular context (Deci & Ryan, 2000). Relatedness refers to the
feeling that one belongs in a particular social setting (Vlachopoulos & Michailidou,
2006). Any factor which could fulfil students’ needs for autonomy, competence, and
relatedness will facilitate the development of intrinsic motivation (Vallerand, 1997).
A recent meta-analysis (Braithwaite et al., 2011, pp: 632-633) has connected
students’ perceived motivational climate to several variables: “maladaptive outcomes
such as anxiety, boredom, competitive strategies.... were largest for.... groups exposed
to performance climate. Adaptive outcomes that were positive for groups experiencing a
mastery treatment included attitude, commitment, enjoyment, competence...”. Certainly,
the learning environment can be affected by variables such as students’ feelings of
boredom, effort or pressure/tension. Student engagement in physical education seem to
decline as students progress through secondary education, but Treasure and Roberts
(2001) found that mastery-oriented motivational climates were related to students’
beliefs that effort caused success and satisfaction. Anxiety involves feelings of tension,
uncertainty or nervousness, and Papaioannou (1995) found that students who perceived
a high learning environment had low levels of anxiety.
Previous research has showed that peers’ influence can have an impact on students’
perceptions of the class climate, especially during adolescence (Vazou, Ntoumanis &
Duda, 2006). This is particularly true in physical education where students interact
constantly through active practice. Two main social goals have been researched in
educational contexts. Social relationship refers to an individual’s desire to form and
maintain positive peer relationships in school (Patrick, Hicks & Ryan, 1997). Social
responsibility represents the desire to adhere to social rules and role expectations
(Wentzel, 1991). There is evidence of the positive correlation between students’ social
goals and task-involving class climates in physical education (Gonzalez-Cutre, Sicilia,
Moreno & Fernandez-Balboa, 2009). Similarly, there has also been observed a positive
connection between responsibility goals and desirable consequences such as effort or
persistence (Guan, Xiang, McBride & Bruene, 2006) and between relationship goals
and interest, enjoyment, intrinsic motivation, and satisfaction (Papaioannou et al.,
In the XXI century, physical education teachers face one major educational goal:
motivate their students to learn and develop lifelong physical activity habits.
Understanding what types of class climates teachers create would help them reach this
important target. This study proposes the identification of clusters in the perceptions of
class climate of a group of Spanish adolescents, and how these perceptions shape
several students’ psychological, motivational, and social variables.
Based on the aforementioned, the main purpose of this study was to uncover the
different motivational climate profiles in a large sample of physical education students
in Spain. A second goal was to examine the relationship between different motivational
climate profiles and students’ basic psychological needs, motivation, social and
behavioural outcomes. Our hypothesis was that task-learning class climates will be
correlated to high levels of self-regulated motivation, effort, enjoyment, responsibility
and relationship, and low levels of pressure/tension and boredom.
Participants and procedure
A total of 507 secondary education students from a high school in the northern part of
Spain agreed to participate (267=52.6% males, 240=47.4% females). The age of the
students ranged from 12-17 years (M= 14.37, SD= 1.69). Participants’ socioeconomic
and ethnic background was normal for Spanish’ standards (white, middle-class
students). Our aim was to analyze students’ perceptions of an average high school in
Spain. The implementation of this project involved three steps: first, permission from
the Ethics Committee of the University of Oviedo and the participating school were
obtained. Second, an informed consent was also obtained from the parents of all
students who participated. Third, all questionnaires were administered by two of the
researchers during a regularly scheduled physical education class, who monitored the
students during data collection, and answered all questions.
The Perceived Motivational Climate in Sport Questionnaire-2 (PMCSQ-2; Newton et
al., 2000) was validated for Spanish physical education settings by Gonzalez-Cutre,
Sicilia, and Moreno (2008). It consists of two high order scales, each one including
three subscales: Task Climate: Cooperative Learning, Effort/Improvement, and
Important Role; Ego Climate: Punishment for Mistakes, Unequal Recognition, and
Intra-Team Member Rivalry.
The Basic Psychological Needs in Exercise (BPNES; Vlachopoulos & Michailidou,
2006) was validated to Spanish physical education contexts by Moreno et al. (2008). It
contains three subscales: Autonomy, Competence, and Relatedness.
The Perceived Locus of Causality questionnaire (PLOC; Ryan & Connell, 1989)
contains four subscales to measure motivation in the classroom: Intrinsic Motivation,
Identified Regulation, Introjected Regulation, and External Regulation. It was adapted
for physical education contexts by Goudas et al. (1994). The same authors also adapted
the Amotivation subscale of the Academic Motivation Scale (Vallerand et al., 1993).
The complete instrument was validated for Spanish physical education settings by
Moreno et al. (2009).
Three subscales of the Intrinsic Motivation Inventory (IMI; McAuley, Duncan &
Tammen, 1989) were used: Effort, Enjoyment, and Pressure/Tension. They represent
significant consequences of the different types of motivation (Ntoumanis, 2002).
Following Hambleton, Merenda and Spielberger (2005), the three subscales were
translated into Spanish by a specialist, and then again into English to test their similarity
with the original ones. Two experts assessed all the items, and they approved their
adequacy in Spanish education contexts.
The Social Goal Scale (SGS; Patrick et al., 1997) includes two subscales:
Responsibility and Relationship. It was adapted by Guan et al. (2006) for physical
education settings (SGS-PE), while Moreno, González-Cutre, and Sicilia (2007)
validated it for Spanish contexts.
A subscale developed by Duda, Fox, Biddle, and Armstrong (1992) to measure
students’ affective responses while performing physical activity was used. Again, we
followed Hambleton et al.’s (2005) procedure to probe its adequacy in Spanish
The item response format of all questionnaires was a 5-point Likert-type scale, ranging
from 1 = “totally disagree” to 5 = “totally agree”.
All data was analyzed using the statistical program SPSS 19.0 (IBM, Chicago, IL).
Psychometric properties of the instruments
The first goal was to test whether the factor structure of the scales matched the
dimensions described above and confirm that they were valid for our sample. We
carried a Confirmatory Factor Analysis of the different subscales using the robust
maximum likelihood method. Several indices were considered: χ2, ratio between Chi-
Square and Degrees of Freedom (χ2/D.F.), Goodness of Fit Index (GFI), Comparative
Fit Index (CFI), Incremental Fit Index (IFI), Tucker-Lewis Index (TLI), Root Mean
Square Error of Approximation (RMSEA), and Standardised Root Mean Square
Residual (SRMR). According to Jöreskog and Sörbom (1989), χ2 can be influenced by
the sample size (p is usually significant with large samples). Therefore, it is better to
consider χ2/D.F., which it is satisfactory when values are below 5 (Bentler, 1999).
Following Schumacker and Lomax (1996), indices such as GFI, CFI, IFI, and TLI are
adequate when their values are .90 or above. RMSEA values of .6 or below and SRMR
values of .08 or below are also acceptable (Hu & Bentler, 1999). Table 1 presents all
Confirmatory Factor Analysis’ fit indices values.
The limited global fit of the original model, coupled with the presence of several
measurement errors linked to some items (along with some undesirable cross-loadings
suggested by modification indexes provided by the statistical program) prompted some
changes in the initial model. Deleting items to improve the factor structure of an
instrument is considered a legitimate process, since it keeps the overall structure of the
model originally formulated using the right indices (Hofman, 1995). Therefore, several
items had to be disregarded to improve the original model: PMCSQ-2: one of the
cooperative learning, effort/improvement and important role subscales; BPNES: one of
the competence and relatedness subscales; PLOC: one of the identified regulation,
introjected and external regulation subscales; IMI: one of the enjoyment and
pressure/tension subscales; SGS-PE: one of the responsibility and relationship
subscales. All these changes produced a better fit of the original model (table 1), which
allowed us to use the selected instruments with our sample and analyze the results.
Descriptive analysis and bivariate correlations
Table 2 shows Cronbach’s alpha coefficients of all the subscales, means and standard
deviations of all variables, as well as bivariate correlations among them after deleting
the mentioned items. Cronbach’s alphas were above .70 in all subscales, except intra-
team rivalry (.61). However, this result could also be considered acceptable considering
the small number of items of this subscale (Nunnally & Bernstein, 1994). In the
PMCSQ-2, the highest score appeared in the effort/improvement subscale and the
lowest in the intra-team member rivalry one. In the BPNES, the highest score emerged
in relatedness, followed by competence and autonomy. In the PLOC, the lowest score
appeared in Amotivation, being the highest identified regulation. In the IMI, the highest
value emerged in effort and the lowest in pressure/tension. In the SGS-PE, both
variables obtained very similar high values. Finally, boredom achieved the lowest score
of all. The subsequent correlation analysis revealed significant connections among most
variables, which allowed us to perform the cluster analysis to see how these different
correlations grouped showing different student profiles.
It was developed to identify groups of students that responded similarly within the
different motivational climates. The six factors that shape this construct were used as
predictive variables. All different variables were standardized using Z scores (mean= 0,
standard deviation= 1).
Following Hair, Anderson, Tatham, and Black’s procedure (1998), the sample was
randomly divided in two subsamples (n= 253, n= 254). A hierarchical cluster analysis
was conducted on the first subsample to identify the clusters emerging from it. Ward’s
method was used to minimize the within-clusters differences, and to avoid long chains
of observations (Aldenderfer & Blashfield, 1984). Since we seek a solution where
clusters are different from each other and, at the same time, the elements are close
within each cluster, the best solution would be one where the corresponding lines will
take time before coming to a close. In our case, the solution was four clusters, the one
that created a major shift in the coefficients (9.8). This indicated that, from this point,
different clusters were merging. Consequently, it was determined that the solution of
four clusters or groups was more appropriate (figure 1). This decision was also
supported by the corresponding dendrogram.
This hierarchical cluster analysis can be considered highly explorative. Therefore, in
order to verify the results obtained, a K-mean cluster analysis was performed on the
other subsample. According to Aldenderfer and Bashfield (1984), this cross-validation
procedure is very important. If the same cluster groups are found in different samples of
the same population, it is conceivable to assume that the solution has a certain degree of
generality. In this K-mean cluster analysis, 4 groups were also identified and means,
standard deviations and standardised scores were very similar to the 4 clusters identified
in the first subsample (table 3). Therefore, a final K-mean cluster analysis using the
whole sample was performed.
Figure 1 shows the four profiles identified through the cluster analysis. Cluster 1,
labelled “high ego”, was characterized by a high ego climate profile in which all scores
(punishment for mistakes, unequal recognition, and intra-member rivalry) were around
Z = 1.00, and a very low task climate profile with all the scores (cooperative learning,
effort/improvement, and important role) around Z = -1.50. It was composed of 67
students (52.2% males, 47.8% females). Cluster 2, labelled “low ego-task”, consisted of
136 students (46.3% males, 53.7% females) with a medium-low ego and task profiles
with all the scores around Z = -0.50. Students in cluster 3, labelled “high ego-medium
task”, showed a high ego profile in which all scores were around Z = 1.00, but also a
medium task profile with all scores above Z = 0.00. This group had 166 subjects (52.4%
males, 47.6% females). Finally, cluster 4, labelled “high task”, included 138 students
(59.4% males, 40.6% females) and it showed a very low ego profile with all scores
around Z = -0.50, and a very high task profile with all scores above Z = .0.5.
A one-way MANOVA was carried out using the basic psychological needs as
dependent variables and the different clusters as independent variables (figure 2). It
yielded a multivariate significant effect, Wilks’ Lambda = .851, F (9, 497) = 9.36, p <.001,
η2 = .05. The following univariate analysis showed significant differences in all
variables: Relatedness: F (3, 503) = 18.01, p <.001, η2 = .01, Competence: F (3, 503) = 13.66,
p <.001, η2 = .07, and Autonomy: F (3, 503) = 15.36, p <.001, η2 = .08 (Fig. 2). Post hoc
comparisons within groups were conducted using Newman-Keuls’ procedure (Table 2).
Clusters 3 and 4 showed higher levels of basic psychological needs, but there were no
significant differences between them on any of the variables. Nevertheless, there were
significant differences between cluster 1 and 2 (p < .001) and between these and clusters
3-4 (p < .001) in autonomy. Finally, there were also significant differences between
clusters 1-2 and clusters 3-4 in competence and relatedness (p < .001).
In the next step, we performed a second one-way MANOVA using the different types
of motivation (intrinsic motivation, identified regulation, introjected regulation, external
regulation, and amotivation) as dependent variables, and the different clusters as
independent variables (figure 3). It yielded a multivariate significant effect, Wilks’
Lambda = .729, F (15, 491) = 11.12, p <.001, η2 = .10. The following univariate analysis
showed significant differences in all variables: Intrinsic Motivation: F (3, 503) = 29.82, p
<.001, η2 = .15, Identified Regulation: F (3, 503) = 21.12, p <.001, η2 = .11, Introjected
Regulation: F (3, 503) = 13.78, p <.001, η2 = .08, External Regulation: F (3, 503) = 11.22 p <
.001, η2 = .06, and Amotivation: F (3, 503) = 27.55, p < .001, η2 = .14 (Fig. 3). Post hoc
comparisons within groups were conducted using Newman-Keuls’ procedure (Table 2).
Cluster 4 showed the highest levels of intrinsic motivation and identified regulation, and
the lowest level of amotivation. On the contrary, cluster 1 showed the lowest levels of
intrinsic motivation and identified regulation, and the highest level of amotivation.
Finally, cluster 3 showed intermediate levels of intrinsic motivation, while cluster 2
scored low in all variables except amotivation (intermediate level).
Finally, a third one-way MANOVA was carried out using social goals (responsibility
and relationship), boredom and the different outcomes measured (enjoyment, effort,
pressure/tension) as dependent variables, and the different clusters as independent
variables (figure 4). It yielded a multivariate significant effect, Wilks’ Lambda = .701, F
(18, 488) = 10.36, p < .001, η2 = .11. The following univariate analysis showed significant
differences in all variables: Effort: F (3, 503) = 17.47, p <. 001, η2 = .11, Boredom: F (3, 503)
= 20.40, p <. 001, η2 = .11, Responsibility: F (3, 503) = 14.30, p <. 001, η2 = .08,
Relationship: F (3, 503) = 15.86, p <. 001, η2 = .09, Pressure/Tension: F (3, 503) = 9.78, p <.
001, η2 = .06, and Enjoyment: F (3, 503) = 39.93, p <. 001, η2 = .19. Post hoc comparisons
within groups were conducted using Newman-Keuls’ procedure (Table 2). Cluster 4
showed the highest scores in effort, responsibility, relationship, and enjoyment, and the
lowest scores in boredom and pressure/tension. All these scores were significantly
different from the other clusters’ scores. Cluster 1 showed the lowest scores in effort,
responsibility, relationship, and enjoyment, and the highest scores in boredom and
pressure/tension. Cluster 3 showed moderately high scores in all variables. Finally,
cluster 2 showed intermediate scores in boredom and pressure/tension, and moderately
low scores in the other variables.
Discussion, practical implications, limitations, conclusions and future directions
The purpose of the present study was to examine the structure of perceived motivational
climate clusters detected in a large sample of Spanish high school students enrolled in
physical education classes, and asses their relations to several psychological and
motivational outcomes. Four clusters were finally identified.
Cluster 4 was a very high task climate group with a low ego profile, and it was linked
to the most positive set of outcomes: high levels of autonomy, competence, relatedness,
intrinsic motivation, identified regulation, effort, enjoyment, responsibility, and
relationship. Scores turned negative with the least self-determined types of motivation
(introjected and external regulation), amotivation, pressure/tension and boredom. This
cluster is similar to one named “self-determined”, identified in a sample of British
physical education students (Ntoumanis, 2002). Previous studies have also shown that a
task-involving environment is the most desirable class climate in educational settings,
since it has been associated to higher levels of students’ intrinsic motivation,
persistence, effort, interest and participation (Morgan & Carpenter, 2002). Certainly,
highly self-determined students are intrinsically motivated to participate in class
(Vallerand et al., 1997). This idea was also reinforced by the students’ low levels of
amotivation. They seem to feel that it is important to participate and try hard in class for
the intrinsic pleasure of performing the different tasks designed by the teacher, for the
activity’s sake, to learn new things, to develop their competence and to have fun.
According to Vallerand and Losier (1999), self-determined motivation is enhanced
when cooperation is promoted. In our study, students in cluster 4 showed high levels of
intrinsic motivation and cooperative learning. Furthermore, Ames (1992) believed that
motivational climates that encourage students to help each other learn will increase their
feelings of competence, which, in turn, will guide them to higher levels of self-
determination (Vallerand, 1997). Our results support all these different connections,
since students in cluster 4 also reported high levels of competence. According to
Goudas and Biddle (1994), students who perceive their physical education class climate
as task-oriented show higher levels of intrinsic motivation and perceived competence.
Therefore, this cluster’s results tie task climates with high levels of self-determined
motivation and high feelings of personal competence and autonomy. Furthermore, our
findings in cluster 4 bond these ideas with high levels of enjoyment, and low levels of
boredom, too. Certainly, students tend to have fun when they find themselves competent
or skilled, when they can help other classmates learn, when they participate because
they feel it is important. When all these happen, feelings of boredom disappear, because
students have a good time in class. They also feel more autonomous, because they see
themselves capable of doing things without the direct supervision of the teacher.
According to Ntoumanis (2002), physical education students become more interested
in the class when its climate is task oriented. Students in cluster 4 rated significantly
higher the effort they felt they displayed in class, and they also reported significantly
lower levels of amotivation, pressure/tension, and boredom. This could mean that this
group of students had fun, felt lower levels of pressure/tension, and felt competent while
participating in class, so they tried hard. In a previous study, Papaioannou (1995) found
that students who perceived a high learning environment had low levels of anxiety,
which means low feelings of tension. Our results reinforce the idea that mastery
climates tend to produce less pressure/tension in the students.
Results from cluster 4 also showed a link between high levels of cooperative
learning, relatedness, relationship, and responsibility. Previous works have reported that
cooperative learning facilitates the quality and quantity of students’ interactions,
encouraging the development of interpersonal skills (Dyson, 2002). When teachers use
cooperative learning strategies, students work together in groups, interacting with other
students. These processes seem to lead to feelings of connectedness among them, and to
the development of social skills. Cooperative learning also seems to develop feelings of
responsibility among group members, because each one of them feels responsible for, at
least, one part of the group task (Dyson, 2002).
On the other end, cluster 1 represented the perfect example of a high ego, low task
student profile. Subjects in this cluster showed the lowest scores of the whole sample in
autonomy, competence and relatedness, intrinsic motivation, identified regulation,
introjected regulation, effort, enjoyment, responsibility and relationship, and the highest
scores on external regulation, amotivation, boredom, and pressure/tension. This is the
most undesirable class climate and motivational profile in educational settings. This
type of students with low levels of self-referenced motivation and high levels of
external regulation and amotivation are negative predictors of future participation in
education (Vallerand et al., 1997). Wang and Biddle (2001) found a similar
motivational cluster, and they also had the lowest rates of physical activity and the
lowest scores of physical self-worth. According to Ntoumanis (2002), this type of
students can be considered motivationally at risk, because high levels of the least self-
referenced types of motivation can lead those youngsters out of the school system
(Vallerand et al., 1997). As described earlier, these profiles are also related to negative
affective and behavioural outcomes. Results from cluster 1 indicate that this group of
students had low confidence on being able to improve and succeed in school. It is very
possible that these students did not try hard because they felt incompetent to perform
the different tasks proposed by the teacher. Consequently, they did not have fun, and
they were bored. When an individual is not able to achieve success, he/she tends to
dislike the activity and, eventually, stops doing it. Moreover, these results also relate
this students’ profile to low levels of social outcomes such as relationship or
relatedness. This finding was also reinforced by the low scores in cooperative learning
and other variables related to a task class climate found in this cluster. Certainly, these
students did not seem to believe in the group. Maybe they thought that their classmates
could not help them improve. Therefore, their connection with other students was
damaged, loosing that important aid in a person global development. Fortunately, this
group of students was the smallest in the sample.
Cluster 2 represented those students who perceived a low task-ego motivational
climate in their classes. They showed low or very low levels in almost all variables.
Surprisingly, these results were very similar to those obtained by students in cluster 1
(very high ego and very low climate profiles). However, the distinctive elements
between both groups were: significantly higher levels of autonomy, identified
regulation, and enjoyment, and significantly lower levels of external regulation,
amotivation, and boredom in cluster 2. These outcomes could be explained by the fact
that this group of students perceived a significantly higher task climate in their classes.
This perception could have positively affected the mentioned variables, and make this
group of students more self-referenced, which is connected to more desirable
behavioural and affective outcomes (Ntoumanis, 2002).
Regarding this idea, cluster 3 depicted the most interesting student profile of all: high
ego, but also medium-high task. Previous research has showed that students can hold
multiple combinations of goals in classroom situations (e.g., high mastery and high
performance) which, in turn, may be connected to different motivational and
achievement outcomes (Meece et al., 2006). Bearing in mind cluster 4 (high task),
students in cluster 3 showed similar levels of autonomy, competence, relatedness,
identified regulation and relationship. However, the distinguishing elements between
both groups were: significantly higher levels of introjected and external motivation,
amotivation, boredom, and pressure/tension, as well as significantly lower levels of
intrinsic motivation, effort, enjoyment and responsibility in cluster 3. Undoubtedly, this
is a remarkable finding that deserves additional consideration. This group of subjects
perceived a high performance class climate and, consequently, many variables’
outcomes reflected that perception. Recent reviews overwhelmingly links students’
perceptions of a performance-oriented classroom context to maladaptive outcomes
(Meece et al., 2006). However, our students’ scores in autonomy, competence,
relatedness, amotivation, boredom, effort, enjoyment, pressure/tension, responsibility,
and relationship were significantly better/higher than the ones obtained by the other
group of students who perceived a high ego classroom environment (cluster 1). These
results showed that students in cluster 3 reflected many of the adaptive psychological
and motivational outcomes expected from task-involving environments. A possible
explanation for this shift can be found in the buffering hypothesis of the AGT. It
suggests that an adaptive classroom goal structure can weaken the undesirable effect of
a maladaptive classroom goal structure (Ciani et al., 2012). In our case, students’
perception of a medium-high task class climate could have mitigated the effects of a
high ego class climate to produce significantly better psychological and motivational
outcomes. These findings are very important, because the largest number of students of
the total sample belonged to this group. Students’ perceptions of both class climates
could have produced an additive effect on the different outcomes measured
(Linnenbrink, 2005). That is, the perceptions of a medium-high task class environment
seemed to have buffered the negative effects of the high ego class climate perceptions.
Although we believe that our findings can be of help to understand educational
environments in physical education, the present study also holds some limitations. The
first one concerns its representativity for the entire population of secondary education
students. The sample used in this study was limited because all subjects came from just
one high school. Consequently, the results obtained cannot be generalized. A second
limitation refers to the age-range of the sample (12-17). It could be considered very
large, since it covers 6 years. Smaller age-ranges could have yielded different results,
and a better picture of how the different variables change across adolescence. Finally,
another limitation is that the results were not evaluated based on gender. A
differentiated analysis could have helped us find if girls and boys have different
perceptions of the class climate and the related variables studied.
Future investigations should try to establish links between students’ perceived class
climate and students’ achievement goals. How they interact to produce more or less
adaptive responses in adolescents. Moreover, how this interaction might affect high task
and performance class structures within the buffering hypothesis of the AGT. Another
important shift for researchers would be to move away from subjective perceptions of
the goal context to more objective measures such as observations or experimental
In conclusion, in the search for a better understanding of the different elements that
shape educational contexts, students’ perceptions of the class climate that teachers help
create should be very carefully considered. High mastery class climates have been tied
to positive psychological and motivational outcomes, while performance oriented class
climate have been related to less motivationally self-determined and bored individuals.
However, a third group of students emerged from our sample. It represented those
subjects that perceived a high ego class climate, but also a medium-high task oriented
class environment. The buffering hypothesis of the AGT indicates that the most
adaptive classroom structure weakens the effects of the less adaptive one (additive
effect). This group of students had similar autonomy, competence, relatedness,
identified regulation, and relationship values to students in cluster 4 (high task),
significantly higher scores than students in cluster 1 (high ego) in intrinsic motivation,
introjected regulation, effort, enjoyment, and responsibility, and significantly lower in
amotivation, boredom, and effort/tension. Therefore, teachers should try to develop
class environments where students could make choices while performing a task, where
they have to work in close contact with their peers and help them improve. Educators
must try to create learning contexts where all the students could feel that they have a
significant role to play, that their performance is valued by the teacher. Similarly,
teachers ought to generate physical education settings where trying hard is rewarded,
where students could feel successful when they improve their skills, not when they
outperform others. Task-involving class climates posses all these positive traits, but the
additive effect of an ego class climate should not be neglected, either.
To sum-up, physical educators should try to create task-involving learning
classes where trying hard and working and helping peers is rewarded, where all students
feel that they have an important role to play, that they can make choices, and that their
performance is appreciated by the teacher. These learning contexts have been connected
with self-regulated motivation and high levels of autonomy, competence, relatedness,
enjoyment, effort, responsibility and relationship. Task class climates also seem to exert
less pressure/tension on students. However, ego class climates hold an additive, positive
effect to task climates.
Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Beverly Hills, CA:
Ames, C. (1992). Achievement goals, motivational climate, and motivational processes.
In G. Roberts (Ed.), Motivation in Sport and Exercise (161-176). Champaign, IL:
Bentler, P. M. (1989). Comparative Fit Indices in Structural Models. Psychological
Bulletin, 107(2), 238-246.
Braithwaite, R., Spray, C. M., & Warburton, V. E. (2011). Motivational climate
interventions in physical education: A meta-analysis. Psychology of Sport and
Exercise, 12, 628-638.
Cervelló, E. M., Jiménez, R., del Villar, F., Ramos, L., & Santos-Rosa, F. J. (2004).
Goal orientations, motivational climate, equality, and discipline of Spanish physical
education students. Perceptual & Motor Skills, 99(1), 271-284.
Ciani, K. D., Middleton, M. J., Summers, J. J., & Sheldon, K. M. (2010). Buffering
against performance classroom goal structures: The importance of autonomy support
and classroom community. Contemporary Educational Psychology, 35, 88–99.
Deci. E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human
needs and the self-determination of behaviour. Psychological Inquiry, 11, 227-268.
Duda, J. L. (2001) Goal perspective research in sport: Pushing the boundaries and
clarifying some misunderstandings. In G. Roberts (Ed.), Advances in motivation in
sport and exercise (129-182). Champaign, IL.: Human Kinetics.
Duda, J. L., Fox, K., Biddle, S. J. H., & Armstrong, N. (1992). Children’s achievement
goals and believes about success in sport. British Journal of Educational Psychology,
Dyson, B. (2002). Cooperative learning in an elementary physical education program.
Journal of Teaching in Physical Education, 22, 69-85.
González-Cutre, D., Sicilia, A., y Moreno, J. A. (2008). Modelo cognitivo-social de la
motivación de logro en educación física [Socio-cognitive model of goal motivation
in physical education]. Psicothema, 20(4), 642-651.
Gonzalez-Cutre, D., Sicilia, A., Moreno, J. A., & Fernandez-Balboa, J. M. (2009).
Dispositional Flow in Physical Education: Relationships with Motivational Climate,
Social Goals, and Perceived Competence. Journal of Teaching in Physical
Education, 28, 422-440.
Goudas, M., Biddle, S. J. H., & Fox, K. R. (1994). Perceived locus of causality, goal
orientations, and perceived competence in school PE classes. British Journal of
Educational Psychology, 64, 453-463. doi:10.1111/j.2044-8279.1994.tb01116.x
Guan, J., McBride, R., & Xiang, P. (2006). Reliability and Validity Evidence for the
Social Goal Scale-Physical Education (SGS-PE) in High School Settings. Journal of
Teaching in Physical Education, 25, 226-238.
Guan, J., Xiang, P., McBride, R., & Bruene, A. (2006). Achievement goals, social goals
and students’ reported persistence and effort in high school physical education.
Journal of Teaching in Physical Education, 25, 58–74.
Hair, J. F. J., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data
analysis (5th ed.). Englewood Cliffs, NJ: Prentice-Hall International.
Hambleton, R., Merenda, P., & Spielberger, C. (Eds.) (2005). Adapting educational and
psychological tests for cross-cultural assessment. Mahwah, NJ: Erlbaum.
Hofman, R. (1995). Establishing factor validity using variable reduction in confirmatory
factor analysis. Educational and Psychological Measurement, 55(4), 572-582.
Hu, L. T., & Bentler, P. M. (1999). Cut-off criteria for fit indexes in covariance
structure analysis: Conventional criteria versus new alternatives. Structural Equation
Modelling: A Multidisciplinary Journal, 6(1), 1-55.doi:10.1080/10705519909540118
Jöreskog, K. G, & Sörbom, D. (1989). LISREL7: User’s reference guide. Mooresville.
IN: Scientific Software.
Linnenbrink, E. A. (2005). The dilemma of performance-approach goals: The use of
multiple goal contexts to promote students’ motivation and learning. Journal of
Educational Psychology, 97, 197–213.
Lloyd, J., & Fox, K. R. (1992). Achievement goals and motivation to exercise in
adolescent girls: A preliminary intervention study. British Journal of Physical
Education Research Supplement, 11, 12-16.
MacAuley, E., Duncan, T. E., & Tammen, V. V. (1989). Psychometric properties of the
intrinsic motivation inventory in a competitive sport setting: A confirmatory factor-
analysis. Research Quarterly for Exercise and Sport, 60, 48-58.
Meece, J. L., Anderman, E. M., & Anderman, L. H. (2006). Classroom goal structure,
student motivation, and academic achievement. Annual Review of Psychology, 57,
Moreno, J. A., González-Cutre, D., & Chillón, M. (2009). Preliminary validation in
Spanish of a scale designed to measure motivation in physical education classes: the
Perceived Locus of Causality (PLOC) Scale. The Spanish Journal of Psychology,
Moreno, J. A., González-Cutre, D., Chillón, M., y Parra, N. (2008). Adaptación a la
educación física de la escala de las necesidades psicológicas básicas en el ejercicio
[Adaptation of the basic psychological needs scale in exercise to physical education].
Revista Mexicana de Psicología, 25(2), 295-303.
Moreno, J. A., González-Cutre, D., y Sicilia, A. (2007). Metas sociales en las clases de
educación física [Social goals in physical education classes]. Análisis y Modificación
de Conducta, 33, 351-368.
Morgan, K., & Carpenter, P. J. (2002). Effects of manipulating the motivational climate
in physical education lessons. European Journal of Physical Education, 8, 207–229.
Newton, M., Duda J. L., & Ying, Z. (2000). Examination of the psychometric properties
of the Perceived Motivational Climate in Sport Questionnaire-2 in a sample of
female athletes. Journal of Sport Sciences, 18, 275-290.
Nicholls, J. G. (1989). The Competitive Ethos and Democratic Education. Cambridge,
MA: Harvard University Press.
Ntoumanis, N. (2002). Motivational clusters in a sample of British physical education
classes. Psychology of Sport and Exercise, 3, 177-194.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory. Nueva York:
Papaioannou, A. (1995). Differential perceptual and motivational patterns when
different goals are adopted. Journal of Sport and Exercise Psychology, 17, 18-34.
Papaioannou, A., Marsh, H. W., & Theodorakis, Y. (2004). A Multilevel Approach to
Motivational Climate in Physical Education and Sport Settings: An Individual or a
Group Level Construct? Journal of Sport & Exercise Psychology, 26, 90-118.
Patrick, H., Hicks, L., & Ryan, A. M. (1997). Relations of perceived social efficacy and
social goal pursuit to self-efficacy for academic work. Journal of Early Adolescence,
Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization:
Examining reasons for acting in two domains. Journal of Personality and Social
Psychology, 57, 749-761.
Ryan, R. M., & Deci, E. L. (2000). Self-Determination Theory and the facilitation of
intrinsic motivation, social development and well-being. American Psychologist, 55,
Schumacher, R. E., & Lomax, R. G. (1996). A beginner’s guide to SEM. Manwah, New
Jersey: Lawrence Erlbaum Associates.
Treasure, D. C. & Roberts, G. C. (2001). Students' perceptions of the motivational
climate, achievement beliefs, and satisfaction in physical education. Research
Quarterly for Exercise & Sport, 72(2), 165-175.
Vallerand, R. J. (1997). Toward a Hierarchical Model of Intrinsic and Extrinsic
Motivation. In M. P. Zanna (Ed.) Advances in Motivation in Sport Psychology (271–
360). San Diego, CA: Academic Press.
Vallerand, R. J., & Losier, G. F. (1999) An Integrative Analysis of Intrinsic and
Extrinsic Motivation in Sport. Journal of Applied Sport Psychology, 11, 142–69.
Vallerand, R. J., Deci, E. L., & Ryan, R. M. (1987). Intrinsic motivation in sport.
Exercise and Sport Science Reviews, 15, 389-426.
Vallerand, R. J., Fortier, M. S., & Guay, F. (1997). Self-determination and persistence
in a real-life setting: Toward a motivational model of high school dropout. Journal of
Personality and Social Psychology, 72, 1116-1176.
Vallerand, R. J., Pelletier, L. G., Blais, M. R., Briere, N. M., Senecal, C., & Vallieres, E.
F. (1993). On the assessment of intrinsic, extrinsic and amotivation in education:
Evidence on the concurrent and construct validity of the academic motivation scale.
Educational and Psychological Measurement, 53, 159-172.
Vazou, S., Ntoumanis, N., & Duda, J. L. (2006). Predicting young athletes motivational
indices as a function of their perceptions of the coach- and peer-created climate.
Psychology of Sport & Exercise, 7, 215-233.
Vlachopoulos, S. P., & Michailidou, S. (2006). Development and initial validation of a
measure of autonomy, competence, and relatedness in exercise: The Basic
Psychological Needs in Exercise Scale. Measurement in Physical Education and
Exercise Science, 10(3), 179-201. doi:10.1207/s15327841mpee1003_4
Wang, C. K. J., & Biddle, S. J. H. (2001). Young people’s motivational profiles in
physical activity: A cluster analysis. Journal of Sport and Exercise Psychology, 23,
Wang, C. K. J., Lim, B. S. C., Aplin, N. G., Chia, M., McNeill, M., & Tan, W. K. C.
(2008). Students’ Perceived Purposes of Physical Education in Singapore:
Perspectives from a 2 x 2 Achievement Goals Framework. European Physical
Education Review, 14, 51–70.
Wang, C. K. J., Liu, W. C., Chatzisarantis, N. L. D., & Lim, C. B. S. (2010). Influence
of Perceived Motivational Climate on Achievement Goals in Physical Education: A
Structural Equation Mixture Modelling Analysis. Journal of Sport and Exercise
Psychology, 32, 324-338.
Wentzel, K. R. (1991). Social competence at school: Relation between social
responsibility and academic achievement. Review of Educational Research, 61, 1-24.
Table 1. Confirmatory Factor Analysis’ fit indices values
4.Punishment for Mistakes
Table 2. Cronbach’s alphas of the different subscales; means, standard deviations, and correlations among all variables.* p<.05; **p<.01
Cluster 1 (N = 67)
Cluster 2 (N = 136)
Cluster 3 (N = 166)
Cluster 4 (N = 138)
Punishment for Mistakes
Intra-Team Member Rivalry
Males n (%)
Females n (%)
Table 3. Profiles of the four-cluster solution from the K-Means cluster analysis.
Note. Means in the same row that do not share superscripts differ at p <.01 in the Newman-Keuls post hoc test. 0
Figure 1. Perceived motivational climate of the four clusters.
Figure 2. Basic psychological needs of the four clusters.
Figura 3. Motivational profiles of the four clusters.
Figure 4. Consequences measured on the four clusters.