MODULE 10

Wilson1982

Tact10-PlayerCommunication.pdf

A social-constructivist approach in physical education: Influence of dyadic interactions on tactical choices in an instructional team sport setting

Florence Damis-Paraboschi Lucile Lafont Andre Menaut Universite Victor Segalen Bordeaux 2, France

The purpose of this study was to analyze the role of dyadic verbal peer interactions in a team sport such as handball. Participants. 20 boys and 20 girls aged between JI and 12. were assigned to two learning condition groups. The task was an instructional setting in team handball (2 attackers against I defender in each half court). The experimental group with verbal exchanges was e.xpected to obtain higher perfortnances in the post-tests than those who had not interacted verbally between playing sequences. The secondary objective was to investigate whether dyadic gender had an effect on tactical choice and cognitive skills in playing games. The experiment demonstrated the superiority of the dyadic verbal interaction group over the non- interactive group as regards tactical choices and action projects. It also showed gender differences since males performed better females with regard to action efftciency. The results are discussed within the theoretical framework of the social psychology of acquisition and development and within that of teaching for understanding. In conclusion, the socio-constntctivist approach would seem to be pertinent in team sport decision-making tasks.

The socio-constructivist theory based on the concept of socio-cognitive conflict (SCC) (Perret-Clennont, 1980, Perret-Ciermont and Nicolct, 1988; Doise and Mugny 1981. 1997} outlines the role of social interaction in cognitive development. The child can coordinate his/her actions with others and then can develop systems of organization of his/her actions. In these conditions of interindividual coordination, the child develops his/her own cognitive structures through a mechanism of decentering. In exchange, his/her cognitive progress will allow him/her to participate in new, more elaborate social interactions, which in turn will modify the structuring of his/her thought. The supposed causal link is circular and its progress

Aeitnowledgemcnls lo Gerville-Reache. L. for his advice with Ilie statistics.

172 F. DARNIS-PARABOSCHI, L. LAFONT, & A. MENAUT

corresponds to the spiral posited by Piaget (1974) to describe the vector of development (Doise, 1988). Interactions between students therefore allow cognitive development. Nevertheless, improvement happens only wlien cognitive conflict occurs between partners. The conflict must be strong and bring to the fore disagreements that have to be discussed and argued in order to create an SCC. Gilly, Fraisse, and Roux (1988) demonstrate the possible efficiency of other forms of conflict-free interaction such as acquiescent co-elaboration and co-construction. Such mechanisms are able to induce cognitive progress in problem-solving. A key element here is the interactive dynamics between partners. Both theories emphasize intra-individua! prerequisites and regulation of socio-cognitive interactions, and refer to Vygotskian theory, which highlights peer roles in the concept and knowledge building area. Fraisse (1985) stresses the disruptive role of the partner in a logic problem solving situation. In this case, working with somebody may be better than working alone when the initial level of the participants, the socio-cognitive dynamics of the dyad, and the intellectual functioning demanded by the task are taken into account. Team sports seem able to generate co-elaboration and co-construction abilities with or without SCC between partners, and a co-operative task involving opposing points of view may be a way to develop cognitive strategies.

In the domain of motor skill acquisition, the importance of working in a dyadic peer situation as a leaming method has been highlighted by Arripe-Longueville (d'), Fleurance, and Winnykamen (1995); Arripe-Longueville {d'), Gernigon, Huet, Cadopi, and Winnykamen, {2002a,b). In the first study, the task used was the forward somersault. The results demonstrated the superiority of both asymmetrical and symmetrical conditions versus the individual condition for all variables. The participants were adolescents from 14 to 16 years old. In the case of another morphokinetic skill, the breaststroke turn in swimming, Arripe-Longueville (d') et al. (2002b) demonstrated an interaction between gender and dyadie situations. Heavily asymmetrical situations were better for males whereas lightly asymmetrical conditions were better suited to females. In a pilot study, Darnis-Paraboschi, Lafont and Menaut (2002) demonstrated the efficiency of symmetrical dyadic situations on tactical choices in an invasive team sport during the deferred post-test. A weakness of the study was the control of dyad gender (female, male, mixed dyads). Moreover, Arripe-Longueville (d') (1998) highlights three peer interaction modes within dyads: guidance-tutoring for asymmetrical dyads, cooperative behavior in the case of girls and symmetrical dyads and parallel activity mainly adopted by boys. Charleswortli and Dzur (1987) showed the preference of girls for tutoring and co-operative behavior in comparison with boys, who were more inclined to social comparison. In a more recent study, Tutge (1992) reported that girls and boys differ according to collaboration, girls being more likely to regress because they are more interested in preserving good relations with their partners. Moreover, in a co-operative situation involving opposing points of view such as in an instructional setting in team sport, we hypothesize the existence of gender differences.

Dyadic interaction can also be examined from the point of view of a theoretical framework of co-operation leaming in a classroom setting in physical education (Lafont & Winnykamen, 1999). Co-operative leaming may be seen as an organization mode in small groups within the classroom. Co-operative situations are more efficient than individual and competitive leaming conditions inasmuch as three main fields are concerned: interpersonal relations, self-confidence and self-esteem, as well as academic achievement (Johnson, Muruyama, Johnson, Nelson, & Skon, 1981; Slavin, 1983). Co-operation with competition between groups appears to generate more interaction than inter-individual competition and working-alone situations (Slavin, 1983). However, co-operation and competition are often linked in sports, particularly in team sports where intra-group co-operation exists alongside inter-group eompetition (Lafont & Winnykamen, 1999). In the reduced type situation studied here, there is a co-operative goal both with the partner and the opponent. In the same theoretical perspective, Lafont, Proeres, Burvingt, Menaut, and Poitreau (1999) investigated the effects of team discussion (about playing strategies) on interpersonal relations and skill acquisition. In that study, an experimental group placed in a situation of collective discussion seemed to gain an advantage in the interaction situation with regard to individual and

LEARNING IN DYADS IN TEAM SPORT 173

collective participation in the game. Moreover, according to Bouthier (1988), verbalization seems to constitute an essential operation for working out, implementing, and regulating collective motor skills. Its introduction into teaching team sports and games should develop dynamics and might particularly involve reciprocal effects between the individual contribution of the players and the collective performance. When building knowledge and action rules, collective discussions and verbalization help players to develop tactical skills (Grehaigne & Godbout, 1995; Grehaigne, Godbout, & Bouthier, 1999).

Several models describe the relationship between knowledge and performance in the development of complex skills within rapid time constraints. French and Thomas (1987), Thomas and Thomas (1994) and Williams and Davids (1995) highlight the simultaneous importance of declarative knowledge, procedural knowledge and performance skills in team sports. French and Thomas (1987) hypothesize that a foundation of sport-speeific declarative knowledge is necessary to make appropriate decisions within the context of a game. They demonstrate that both cognitive and motor skills contribute to the development of children's basketball skills. The improvement of specific declarative knowledge might be related to better procedural knowledge and decision-making development. According to French and Thomas (1987), both declarative and procedural knowledge improve decision-making ability. To improve qualities of inference as well as "if., then... I do..." statements, declarative knowledge of the same type has to be developed. It seems that dyadic interactions and peer verbalisation are able to generate declarative knowledge as action rules on tactical choices. In Williams and Davids' framework (1995), declarative knowledge is built through training and playing, while procedural knowledge promotes the acquisition and rclention of specific declarative knowledge; i.e., knowing facilitates doing and doing facilitates knowing. To study the development of tactical efficiency in social interaction, declarative knowledge and procedural knowledge need to be explored and could have an impact on performance.

Another framework is the Teaching Games For Understanding (TGFU) model (Thorpe, Bunker, & Almond, 1986; Grehaigne, Godbout, & Bouthier, 1999; Grehaigne, Godbout, & Mahut, 1999) which highlights the debate concerning the respective importance and role' of techniques and tactics in learning games and sports. The model proposes that priority should be given to understanding and learning tactics related to a game or a sport and to the specific technical skills to be practiced when the need is perceived by the student. We suggest a tactical approach to teaching games in the light of a socio-constructivist perspective of the teaching-learning process through the notion of debate-ofidea developed by Grehaigne, Godbout, and Bouthier (2001). In a debate-of-idea setting, retrospective verbalization seems to provide information and stimulate reflection about obstacles encountered by players in their efforts to solve a problem. The construction of suitable personal tactical skills in a specific situation could be referred to as indirect teaching (Grehaigne, Godbout, & Bouthier, 1999) associated with a socio-constructivist approach which contends that knowledge constructed by the student is a result of the interaction between his/her cognitive activity, reality, and other people's. Conscious action-efficient rules emerge from peer verbaiizalions. A co-operative learning process in harmony with the internal logic of invasive team sports requires both eonceplualization of the action (Vergnaud, 1996) and achievement of the playing action. Barth (1993) developed a pedagogy focused on the student's cognitive and meta-cognitive strategies. Meta-cognitive knowledge corresponds to knowledge the student has upon his/her own knowledge and his/her cognitive operation. Meta-cognition is a knowledge and attitude transfer tool, so meta-cognitive processes seem to be developed to generate conscious action- efficient rules about decision-making in team sports and games. It seems that social interactions between peers enable the emergence of meta-cognitive knowledge required in the TGFU model.

The main purpose of tbis study therefore was to highlight the role of dyadic peer interactions to build action rules. Declarative, procedural and meta-cognitive knowledge are constitutive of tactical efficiency in a team sport. To our knowledge, apart from the research conducted by Darnis-Paraboschi, Lafont, and Menaut (2002), the present experiment with dyads is the first for this type of task. Following on from Arripe-Longueville's results (1998),

174 F. DARNIS-PARABOSCHI, L. LAFONT, & A. MENAUT

the participants in dyads were expected to obtain higher performances in the post-tests than those who had not interacted between playing sequences. The study is a quantitative one and the verbal statements of the participants are not described here. The secondary objective was to observe the effect of dyadic gender on tactical choice and cognitive skills in playing games. However, previous studies do not allow precise hypotheses to be made regarding the differences between boys and girls.

Method

Participants and design

The participants were forty French students: 20 boys and 20 girls. Their average age was 1 1.2 years (SD=0.42) and they were from middle-class and lower s o c i o - e c o n o m i c backgrounds. They were selected from a sample of 72 pupils attending three different classes of two teachers who worked together on the same project in the same school. These two teachers were not part of the research team. All participants and their parents provided informed consent before they participated in the study. They were told that they would be filmed but that their privacy would be protected. The experimental design was approved by the university's ethical committee.

All the 72 pupils volunteered to participate in the study. We selected beginners in the team sport, i.e., only those who did not practise team sport outside of the school and those were present at the three stages of the experiment. Thus, 40 participants were randomly assigned to the four independent groups of a 2x2 (Gender x Learning condition) design.

Task materials

The required task was a reduced situation of European handball. This instructional setting was established within the planned sequence of activities for the first year of secondary school and was part of the basic program for teaching beginners how to make decisions in the activity. The experiment was carried out during a physical education lesson. The class teacher took charge of the pupils not selected by the experimental design but did not participate in data collection. The experiment took place in the middle of the curriculum and the pupils knew the basic rules. Lessons were given in a 40x20 meter gymnasium with standard handball court regulations. All the participants received the same verbal instructions, and had to perform both attacking and defensive tasks.

Figure I. The motor task, handball instructional setting

The attacking instruction was as follows: "you have to start at the goal area line with one ball for two players and you have to reach the opposite goal area line to take a shot at the goal without losing the ball. In order to succeed, you have to follow handball rules."

LEARNING rN DYADS IN TEAM SPORT 175

The defensive instruction was as follows: "you have to divide up into each half-court: one defensive player in the half-court in front of the medium-tine and the other one in the other half-court behind the medium-line. You have to try to stop the attacking players carrying the ball toward your goal by intercepting the ball or trying to stop the shot using handball rules".

The participants had to take turns in playing the role of attacker and defender. The players always rotated in the same manner, so each attacking dyad always met the same defensive dyad.

Videotape: Verbal exchanges were recorded on a Hitachi 8mni Video Camera/Recorder VM-E555LE and a directional microphone: Firstline Micro Cassette Recorder.

Procedure

The first test consisted in assessing the initial level of the participants in the instructional setting {all participants were randomly assigned to a dyad). Twenty dyads were composed, 10 female dyads and 10 male dyads assigned to two conditions: with (experimental group) or without verbalization (control group).

Pre-test. A declarative knowledge pencil and paper test ( D K l ) was performed with a 1 4 - p o i n t s c a l e for t h e o r e t i c a l t a c t i c a l c h o i c e s . T h i s c o n s i s t e d of a m u l t i p l e - c h o i c e questionnaire asking the pupils a series of if X occurs, then I do V" statements. Each question was illustrated by a diagram explaining the playing situation.

An action project questionnaire ( A P I ) using a 4-point scale described the different stages of learning the skill. A motor pre-test consisted in an instructional setting in team handball. On a handball court, one defender was on each half court with a goalkeeper. Each dyad had to move from one place to take the ball to the opposing goal area and make a shot at goal, using passes or dribbling, Each dyad had five attempts at attack and five at defense. For each attempt we assessed a choice relevance measure ( C R l ) using a 20-point scale (a ratio of good choices on all the choices observed multiplied by 20) and an action efficiency measure ( A E l ) using a 5-point scale (1 point for every shot at the goalkeeper on the goal area line).

Training in pairs

In the experimental conditions (Exp), the participants were asked to discuss (3 times, 3 minutes) and to come to an agreement with the partner in order to find a solution for winning.

In the control condition (Cont), the participants were asked to make as many passes as possible (3 times, 3 minutes). Between each training phase, one attempt was made on the court in the same conditions as the pre-test.

Immediate post-test. All participants performed five trials in the same conditions as the pre-test (2 attackers against 1 defender on each half court) and a choice relevance measure (CR2) and an action efficiency (AE2) measure were assessed. Furthermore, a declarative knowledge test (DK2) and an action project questionnaire (AP2) were administered.

Deferred post-test. All participants performed five attempts in the same conditions as the pre-test one week later. For each attempt, a choice relevance measure (CR3) and an action efficiency measure (AE3) were assessed.

Measures

Dependent measures included choice relevance (CR) and action efficiency (AE) for motor performance. In accordance with previous research (Arripe-Longueville (d'), 1998; Damis-Paraboschi, Lafont, & Menaut, 2002), action project and declarative knowledge were also measured.

176 F. DARNIS-PARABOSCHI, L. LAFONT. & A. MENAUT

The declarative knowledge test (DK) described seven variable situations all requiring decision making by the player. For each "if-then " statement, we assigned 2 points for the best solution, 1 point for a halfway solution and 0 point for the wrong solution. This test was constructed by three expert team handball teachers and was standardized in a pilot study conducted with 34 boys and girls of the same age and skill level.

According to Thomas and Thomas (1994), novices lack declarative knowledge such as offensive and defensive strategies. They often do not know which skill to execute or when to execute it. With more experienced students, the performance differences tend to be less skill- related and more knowledge-based (French & Thomas, 1987; McPherson & Thomas, 1989). However, in a situation of movement, it is possible to know when and how to perform a movement without actually being able to execute it (Thomas & Thomas, 1994). We therefore investigated choice-making declarative knowledge through this test, looking first for a link to choice relevance and secondly for the effect of the acquisition process on it.

The action project questionnaire (AP) was a written questionnaire asking the students how to perform in order to succeed in the problem situation. It stated: "In a 2 attacking players against I defensive player situation, what do you both have to do to succeed in making a shot eveiy time without losing the ball?

Describe the possibilities in this situation and foresee the solutions. " The action project questionnaire was coded 1 point for a "you have to pass the ball": 2

points for: "you have to pass^ the ball and tnove into space ".* 3 points for: "you have to pass the ball or dribble, then tnove into space": 4 points for an answer in which the various possible answers were given according to the defensive player: "If the defensive player does that.... then I do that... ". This test had been used and standardized in a pilot study with 34 boys and girls of same age and had been validated by three expert handball teachers.

According to Grehaigne, Godbout, and Bouthier (1999), efficiency during a game depends on action efficient rules which learners try to fonnulate. Tactical efficiency is linked to a generative capacity to produce infinite tactical types of behavior in response to infinite configurations of play. We therefore expected ati improvement in answers regarding the tactical rules described in the action project questionnaire with the training conditions. The experimental group had to verbalize and discuss the best solution for succeeding in the task.

For each handball team player, the (+) number was divided by the action number (+;+/-;-) and multiplied by 20. This gave a ratio of rather good choices with regard to total action function. This scale had been used and standardized in a pilot study with 34 boys and girls of same age and validated by three expert handball teachers (Darnis-Paraboschi, Lafont, & Menaut, 2002). Choice relevance (CR) using a 20-point scale was measured by three teachers who were handball experts. They observed the players in the task and noted "good choices" (+), "neutral choices" (+/-) and "poor choices" (-). A concordance test (Kendall test: ^(2,24)=0.16,/?<.05) was performed and there were no differences between the teachers.

We evaluated procedural tactical knowledge evaluation on the choice relevance scale. According to French and Thomas (1987), declarative knowledge tests and choice relevance scales are linked. Moreover, the learning condilions might influence choice relevance performance. Moreover, experimental group players who are called upon lo perform verbal exchanges in pairs should generate and produce better adapted choices for the situation. They should then improve their choice relevance score more than the control group.

The action efficiency measure (AE) used a 5-point scale. For each attempt the attacking dyad players received one point if they succeeded in making a shot at the goal line. They had five attempts, and then a 5-point scale was used to measure playing efficiency. We expected a relationship between choice relevance and action efficiency as well as a better training condition effect.

Data analyses

First, correlational analyses were computed among relevant variables. To test the hypothesis that verbal interaction would lead to higher performances than the control

LEARNING IN DYADS IN TEAM SPORT 177

condition (without interaetion) and to observe the potential gender differences, a 2x2x2 (Gender x Learning condition x Period) analysis of variance with repeated measures on the last factor (RMANOVA) was conducted for Declarative Knowledge and Action Project. For Choice Relevance and Action Efficiency, a 2x2x3 (Gender x Learning condition x Period) analysis of variance with repeated measures on the last factor (RMANOVA) was conducted. Comparisons were conducted by univariate ANOVA. Eta squared was computed to examine the effect size of the significant differences. Following Cohen (1988), ES is interpreted as 0.8 representing a large effect, 0.5 as a moderated effect and 0.2 as a small effect.

With a view to comparing the four groups according to the acquisition procedure, their equivalence on the pre-test was verified. No difference was found between the four groups for all measures.

Results

Correlation analyses

Table 1

Correlational matrix

DK1 DK2 API AP2 CRI CR2 CR3 AEI AE2 AE3

DKl DK2 .70*** API .27 .25 AP2 .18 .13 .57*** CRI .35* .35' .26 .30* CR2 .45** .49*** .25 .37* .67*** CR3 .33* .42** .28 .48** .80*** .82*** AEI .16 .15 .09 .06 .58*** .33* .41** AE2 .36 .31* .15 .02 .57*** .62*** .48** .69*** AE3 .28 .25 .18 .14 .62«** .55*** .56*** .88*** .87***

Note. •/7<.05,*V<.01,"**/J<.001.

The correlational analysis showed that all the measures were related between the pre-test and the post-tests. Moreover, declarative knowledge level was positively related to choice rel- evance on the pre-test and on the immediate post-test (r=.35,p<.05 and r^.49,/j<.0l, respec- tively). Declarative knowledge was also positively correlated to action efficiency only on the immediate post-test {r=3l, p<.05). Action project was positively related to choice relevance only on the immediate post-test (r=.37, p<.02). Choice relevance was positively related to action efficiency on the pre-test, the immediate post-test and the deferred post-test (r=.58, /K.00I;r=.62,p<.001 and r-.56,/7<.00l respectively).

Learning condition differences

Table 2 displays the means and standard deviations of all measures for the two learning conditions.

No difference was found when analyzing remainders. Observing the scores diagrams of the descriptive analysis did not reveal any outliers.

178 F. DARNIS-PARABOSCHI, L. LAFONT. & A. MENAUT

Table 2

Means (and standard deviations) of all measures for the two learning conditions

Group

Experiment.

Control

Declarative knowledge

10.2 (3.43)

10.1 (2.92)

10.65 (3.42)

9.50 (3.50)

Action project questiontiaire

1 2

1.75 2.50 (0.97) (1)

1.80 1.90 (1) (1.12)

1288 (2.80)

12.22 (3.50)

Choice relevance

14.92 (3.63)

12.49 (4.02)

14.85 (2.9)

12.31 (3.4)

2.50 (1.47)

3.3 (1.22)

Action efficiency

3.1 (1.68)

3.8 (1.44)

3 (1.38)

3.4 (1.05)

Dyad gender and learning condition differences

Table 3 displays means and standard deviations for leaming conditions by gender.

Table 3 Means (and standard deviations) of all measures for each group

Group

Feme I exp.

Femal control

Male exp.

Male control

Declarative knowledge

8.60 (3.17)

10.60 (2.54)

11.80 (3.01)

9.60 (3.30)

10.10 (2.92)

9.60 (3.50)

11.20 (3.94)

9.40 (3.68)

Action project questionnaire

!.7O (0.82)

1.80 (1.23)

1.80 (1.13)

1.80 (0.79)

2.40 (0,70)

1.90 (f.28)

2.60 (1.26)

1.90 (0.99)

12.06 (2.53)

11.45 (1.23)

13.70 (2.92)

12.99 (2.56)

Choice relevance

13.33 (1.98)

12.67 (5.20)

16.50 (4.28)

12.32 (2.65)

13.80 (1.40)

12.03 (4.05)

15.90 (3.66)

12.60 (2.79)

1.80 (1.23)

3.20 (1.23)

3.20 (1.40)

3.40 (1.26)

Action efficiency

2.00 (0.66)

3.60 (1.95)

4.20 (1.68)

4.00 (0.66)

2.20 (1.42)

3.40 (1.26)

3.80 (1.55)

3.40 (0.84)

No difference was found when analyzing remainders. Observing the scores diagrams of the descriptive analysis did not reveal any outliers.

Declarative knowledge

For declarative knowledge, the RMANOVA showed no effect for Gender or learning condition (F(l,36)=0.7 and F(l,36)=0.42, ;7>.O5). There was no significant progress (F(],36)=0.04) for declarative knowledge.

Action project

For action project, the RMANOVA revealed a main effect for the Assessment period, F(l,36)=8.26, ;7<.O1 (£5^.19). The analyses also indicated a Leaming condition x Assessment period interaction F(l,36)=4.82, p<.05 {ES=.\2). Bonferroni's post-hoc test indicated no significant difference between groups. Nevertheless, for the Assessment period, only the experimental group of boys improved significantly (F(],9)^16, p<.01) (£5'-.64). No significant or interactional effects between Gender and Learning condition emerged (F(l,36)-.03,;7>.05).

LEARNING IN DYADS IN TEAM SPORT 179

Choice relevance (CR)

There was a significant effect of the Assessment period (from pre-test to immediate post- test and deferred post-test), F(2,72)=5.84,;;<.01 (£5-.14).

Between pre-test and immediate post-test, RMANOVA indicated significant progress ( F ( ! , 3 6 ) ^ 6 . 8 3 , ;3<.O5 (£"5=.16). The analyses also indicated a Learning condition x Assessment period interaction, F(I,36)=3.96, p=.O5 (ES=.\O). Bonferroni's post-hoc test indicated no significant differences between groups. Nevertheless, on the immediate post-test, ihere was a significant effect of group F(l,36)=4.I5,p<.05 (£5=.10). The experimental group obtained a higher performance than the controls {M exp.= 14.92 versus M contr.= 12.50). Furthermore, only the experimental group of boys progressed significantly between the pre- test and immediate post-test F{ 1,9)=8.14, / x . 0 5 (£S=.48).

Between the pre-test and deferred post-test, RMANOVA indicated a significant effect of Assessment period, F(l,36)=12.64, p<.01 (ES=.26), and a significant interaction between the training condition and assessment period F(l,36)^10.42, p<.0\ {ES=.22). On the deferred post-test, Bonferroni's post-hoc test indicated that the experimental group of boys obtained a higher p e r f o r m a n c e than female c o n t r o l s (M e x p . M a l e s = 1 5 . 0 9 versus M c o n t . Females=l2.03) (p^.O5). On the deferred post-test, ANOVA indicated a significant effect of group: F(l,36)=6.49, p<.05 {ES^.\5). The experimental group obtained a higher perfonnance than the eontrols (M exp.= 14.85 versus A/contr.=^12.31). Moreover, only experimental groups progressed significantly, respectively: F( 1,9)^6.38,/j<.05 (£"5=.41) (for females) and F(l,9)=20.17,/)<.01 (ES=.69) (for males). No other effects were significant.

Action efficiency (AE)

There was a significant effect of the Assessment period, F{2,72)=7J2, p<.0\ (from pre- test to immediate post-test and deferred post-test) (£5^.18).

Between the pre-test and immediate post-test, RMANOVA indicated a significant effect of the Assessment period, F(I,36)=8.64,/7<.O1 ( £ 5 - . 1 9 ) . The analyses also indicated a significant effect of Gender, F(l,36)=7.78,/7<.O1 {ES=.\ii). Boys performed significantly better than girls on the immediate post-test (M males=4.1 versus M females=2.8). The Bonferroni's post-hoc test indicated that the experimental group of boys had higher performances than the experimental group of girls: M exp. males=4.2 versus M exp. Females-2 (p=.006).

Between the pre-test and deferred post-test, RMANOVA indicated a significant effect of Assessment period, F(l,36)=9, p<.0\ iES=.2O) and Gender, F(l,36)=4,8, p<.05 (ES=.12). Boys performed significantly better than girls {M males=3.6 versus M females=2.8) on the deferred post-test. The Bonferroni's post-hoc test showed that the experimental group of boys had higher performances than the experimental group of girls: M exp. maies=3.8 versus M exp. Females=2.2 (p=.0\) on the deferred post-test. There was a significant interaction between Leaming condition and Assessment period F{\,36)=4, p<.05 (£5=. 10). RMANOVA by groups showed that only the experimental group of boys progressed significantly F{\,9)=\3.5,p<.0\ (£5=.60). No other effects or interaction effects were significant.

Discussion

The main purpose of the present study was to explore the effects of dyadic verbal interaction on cognitive, motor and tactical variables. Additionally, gender effect was assessed. The results demonstrate the effect of training in symmetrical dyads with verbalization for action projects and for choice relevance on the deferred post-test. This demonstrates the role in understanding of verbalization for tactical representation and for tactical choices. Action-efficient strategies developed because co-operation by verbal

180 F. DARNIS-PARABOSCHI, L. LAFONT. & A. MENAUT

interaction between peers allowed to effective tactical action in the long term. These results are consistent with those found by Lafont et al. (1999), for a team sport and by Arripe- Longueville (d'), Fleurance and Winnykamen (1995); Arripe-Longueville (d') et al. (2002a,b) for other sports and other motor skills. The results also confirm the importance of verbal exchanges (Bouthier, 1988; Grehaigne, 1999).

Following Cohen (1988) we observed moderated effects for the progress of experimental group, but small effects for learning condition. These moderate or small effects may be due to several factors. First, although they were selected as novices, some participants obtained quite good motor scores at the beginning of the tests. Moreover, the interactive session was rather brief. Finally, the motor task was a complex one used in physical education. For these reasons, the ecological validity of the situation may have attenuated the observed effects.

In the present study, no differences were found between boys and girls for choice relevance, declarative knowledge or action projects. In a preliminary study, 144 boys and girls performed equally well in tactical choices (Damis-Paraboschi, 1998). Boys were better than girls only with regard to action efficiency. This may be due to differences in socialization and experience: boys are generally more skilled than girls in team sports because they have greater practical experience of ball games (Anderson, Gebhart, Pease, & Rupnow, 1983). Moreover, studies have shown that males are generally more motivated for competition than females (White & Duda, 1994; Ryckman & Hamel, 1995). White (1991) showed that competitive coaction enhances motor performance for men, but not for women. Consequently, boys may increase their action efficiency more significantly than girls because the goal of the task and the verbal instructions may elicit competitive orientation. Concerning interaction differences between boy and girl dyads, the present findings are quite different from those reported by Arripe-Longueville (d') et al. (2002b) for morphokinetic skills. They showed that girls prefer verbal exchanges whereas boys score higher in physical involvement. Certain reasons may explain why gender differences have not emerged. However, the experimental designs and interactive situations were rather different between authors study and the present one. Nevertheless, a qualitative analysis is to be conducted on the whole corpus in reference to "conversational analysis" and to peer interactive modes within the dyads.

A c c o r d i n g to French and Thomas (1987), declarative k n o w l e d g e and procedural knowledge measured by choice relevance level are positively correlated. To improve our results we hypothesized that logical thought and the subject's operative level infiuence the ability to put inferences into words as "if... then" statements. The subject's hypothetical- deductive thought might influence his/her declarative and procedural knowledge regarding tactical choice. As expected, there was a positive correlation between choice relevance and action efficiency. Thus, procedural knowledge and motor skills are linked in team sports. Declarative knowledge was positively related to action efficiency only on the post-test. These results are consistent with those of French and Thomas (1987), Thomas and Thomas (1994) and Williams and Davids (1995) who highlighted the simultaneous importance of declarative knowledge, procedural knowledge, and motor skills in team sport learning. We found a positive correlation between action projects and choice relevance on the immediate post-test, as well as a significant effect of the assessment period. These results are in agreement with those of Grehaigne and Godbout (1995) and Grehaigne (1999) showing that tactical efficiency (measured by choice relevance) implies the capacity to produce conscious action-efficient rules {measured by action project).

Socio-cognitive learning frameworks (Perret-Clennont, 1980; Perret-Clermont & Nicolet, 1988; Doise & Mugny, 1981, 1997) demonstrate greater efficiency for interactive work over individual work, in particular for Piagetian tasks. Our study demonstrates the value of a socio- constructivist approach to team game decision-making tasks. Dyadic verbal interactions make co-constructive project actions possible as well as collaborative establishment of action- efficient rules. Since team sports are characterized by the need to cooperate, co-ordinated points of view are required to establish one or several strategies aiming at confrontation. Therefore, social interaction is necessary at all stages of the game. Verbal interactions involve the formulation of word sets, and thus conceptualization and thought-making. Awareness

LEARNING IN DYADS IN TEAM SPORT 181

processes are at the heart of meta-cognitive processes and generate the "understanding of learning" (Grehaigne & Godbout, 1995; Grehaigne, Godbout, & Mahut, 1999). Piaget (1974) drew attention to the difference between achievement and understanding. He believed that exercise achievement was task-dependent and was subsumed by intuitive or tacit know-how, whereas understanding was process-dependent. While social development psychologists underline intelligence building in social interaction (Perret-Clermont, 1980), the team sport models discussed here place much emphasis on tactical choice building in social interaction.

A socio-constructivist teaching model should enhance students' construction of tactical knowledge and the development of their decision-making skills. A reduced situation highlighting tactical problems should comprise an instructional setting in which verbalizations between partners should be encouraged in order to promote the emergence of action-efficient rules. The present study shows the importance of alternating between reduced situations, verbalization and debate about idea sequences. Indeed, it might encourage teachers to set up dyadic interaction within the classroom since peers interactions seem to develop relevant tactical choices in problem-solving situations.

These findings open up new perspectives. First, it may be of interest to create a new experimental design with longer interaction within the dyads. Various reduced situations highlighting tactical problems could be investigated with regard to verbalizations between partners. Second, it would be interesting to examine the degree of symmetry of the dyads to assess its influence on improving tactical choices. Arripe-Longueville (d") (1998) demonstrated the greater efficiency of slightly asymmetrical dyads compared to symmetrical ones for morphokiTietic tasks. Future research should confirm these results in the case of tactical choices for a team sport. Although verbalizations have not been systematically studied, it appears that interactions between peers constitute co-elaboration and co-constructions (Gilly, Fraisse, & Roux, 1988). It would be of interest to analyze verbal interactions in the experimental group with progress made in cognitive and tactical performances. Finally, future research might examine the role of contextual characteristics upon the kind of interactive process. Particularly, the infiuence of competitive contexts and the role of initial instnictions and goal task could be measured with regard to dyadic interactions in physical education and sports contexts.

References

Anderson, D.F., Gebhart, J.A., Pease D.G., & Rupnow, A.A. (1983). EITecls ofage, sex, antl placcmeni of a model on children's performance on a ball-siriking task. Perceptual and Motor Skiils, 57, I 187-1 190.

AiTipe-Longueville, K. (d') (1998). Contribution a I'identification de processus cognitifs et de mecanisnies interaclifa imptiques iors de /'acquisition d'une hatiiiete inotrice comptexe. Unpublished doctoral dissertation in STAPS, University of Paris X-Nanterre ("ranee.

Arripe-Longueville, F. (d'), Fleurance, P., & Winnykamen, F. (1995). Effects ofthe degree of competence synimetry- assymmetry in the acquisition of a motor skill in a dyad. Journat of Human Movement Studies. 28, 255-273.

Arripe-Longueville, F. (d'), Gernigon, C , Huet. M.L., Cadopi, M., & Winnykamen, F. {2002a). Peer tutoring in a physical education setting: Influence of tutor skill level on novice learners' motivation and performiince. Journal of Teaching in Physical Education, 22, 105-123.

Arripe-Longueville, F. (d'), Gemigon, C , Huet, M.L., Cadopi, M., & Winnykamen, F. {2002b). Peer assisied learning in the physical activity domain; Dyad type and gender differences. yonnia/o/i/KJ/'/tfi Exercise Psychology, 24, 219-238.

B3,X\)\,^.M.{\99'i) Le savoir en construction. Paris: Relz.

Bouthier, D. (1988). Les conditions cognitives de ia formation d'actinns sportives collectives. Unpublished doctoral dissertation in STAPS, Universily of Paris V, EPME. France.

Charlesworth, W.R., & Dzur, D. (1987). Gender comparisons of preschoolers' behavior and resource utilization in group problem solving. Chiid Development, 58, 191-200.

182 F. DARNIS-PARABOSCHI, L. LAFONT, & A. MENAUT

Cohen, J. {1988). Statistical power analysis for ttie behavioral sciences Ord eA.). Hillsdale, NJ: Erlbaum.

Damis-Paraboschi, F. (1998). Processus cognitifs de type operatoire et action tactic/ue. Unpublished DEA, University of

Bordeaux 2, France.

Damis-Paraboschi, F., Lafont, L., & Menaul, A. (2002). Effet des interactions dyadiques sur la competence tactique en sports colleetifs chez des sujets de 11-12 ans. Actes des Journees nalionales d'etudes de la Societe Frangaise de Psychologie du Sport {pp. 25-28). Limoges, France.

Doise, W. (1988). Pourquoi le marquage social? In A.N. Perrel-Clemiont & M. Nicolet, (Eds.), tnteragir et connaltre. Enjeuxet regulations sociales dans le developpement cognitifipp, 103-105). Fribourg; DelVal.

Doise, W.,& Mugny, G. (1981). Le developpement social de I'inteiligence. Paris: inter-^ditions.

Doise, W.,& Mugny, G. (1997). Psychotogie sociale et developpement cognitif. Paris: Amiand Colin.

Fraisse, J. (1985). Interactions sociales entre pairs el decouverte d'une strategle cognitive chez des enfants de I i ans. Unpublished docloral dissertation in psychology. tJniversity of Provence, Franee.

French. K.E., & Thomas, J.R. (1987). The relation of knowledge development lo children's basketball performance.

Journat of sport psychology, 9. 15-32.

Gilly, M., Fraisse, J . C , & Roux, J.P. (1988). Resolution de problemes en dyades et progres cognitiTs chez des enfants de 11 a 13 ans: Dynamiques interaclives el mecanismes sociocognitifs. In A.N. Perret-Clermont & M. Nicolet (Eds.), Interagir et connaiire. Enjeux et reguititions sociales dans le aeveloppement cognitif (pp. 73-92). Fribourg: Del Val.

Grehaigne, J.F. {1999). L'enseignement des sports colleetifs a I'ecole: Conception, construction et evaluation. Paris Bruxelles: t)e Boeck Universite.

Grehaigne, J.F., & Godbout P. (1995). Tactical knowledge in leam sports from a socio-constmctivist and constructivist perspective. QUEST. 47,490-505.

Grehaigne, J.F., Godbout, P., & Bouthier, D. (1999). The foundations of tactics and strategy in team sports. Journal of

Teaching in Physical Education, IS, 159-174.

Grehaigne, J.F., Godbout, P, & Boulhier, D. {2001) The teaching and learning of decision making in team sports. QUEST. 5 i . 59-76.

Grehaigne, J.F., Godbout, P., & Mahut, N. {1999). L'enseignemeni des jeux par la comprehension: Une revue de ques-

tion. 5r/tP5.-(*. 81-93.

Johnson, tJ.W., Muruyama, G.. Johnson, R., Nelson, D., & Skon, L. (1981) Effects of cooperative, competitive, and individualistic goal structures on achievement: A meta-analysis. Psychological Bulletin, 89,47-62.

Lafont, L., & Winnykamen F. {1999). Co-operation and competition in children and adolescents. In Y. Vanden Auweele, F. Bakker, M. Durand, & R. Seiler {Eds.), Texbook on Psychology for Physical Educators {pp. 379- 404). Champaign, II.: Human Kinetics.

Lafont, L., Proeres. M.. Burvingt, F., Menaut, A.. & Poitreau, D. (1999) t,es situations de cooperation en ilPS. In Identifier les effets de I'intervention en motricite huniains. Actes du collogue AFRAPS-EDPM {CD-ROM) Louvain La Neuve.

McPherson, S.L., & Thomas, J.P. (1989). Relation of knowledge and performance in boy's tennis: Age and expertise. Journal of Experimental Chiid Psychology. 48, 190-211.

Perret-Clermont, A.N. {1980). Social interaction and cognitive development in children, London: Academic Press. {Original work published in 1979)

Perret-Clermont A.N.. & Nicolet M. (1988). Interagir et connaltre. Enjeux es regulations sociates dans le developpement cognitif. Fribourg: DelVal.

Piaget, J. {1974). Reussir et compremtre. Paris: Presses Universitaircs de France.

Ryckman, R.M., & Hamel, J. (1995). Male and female adolescents motive related to involvement in organized team sports. International Journal of Sport Psychology, 26, 383-39''.

Slavin. R.E. (1983). When docs cooperative leaming increase student achievement? Psychological Bulletin, W(3), 375-394.

Thomas, K.T., & Thomas, J.R. {1994). Developing expertise in sport: The relation of knowledge and performance.

Journal of Sport Psychology'. 25, 295-312.

LEARNING IN DYADS IN TEAM SPORT 183

Thorpe. R., Bunker, D.,& Almond. L. (1986). Rethinking games teaching. Lough bo rough: tJniversity orTechnology.

Tutge, J.R.H. (1992). Processes and consequences of peer collaboration: A Vygolskian analysis. Child Development. 63, 1364-1379.

Vergnaud, G. (1996). Au fond de I'aclion, la conceplualisalion. In J.M. Barbier (Ed.), Savoirs iheoriiiucs ct savoirs d'actlon (pp. 275-305). Paris: Presses Universilaires de France.

While, S.A. (1991). Effed of gender and compeiitive coaclion on motor performance. Perceptuel & Motor Skills, 73, 581-582.

White, S.A., & Duda, J.L. (1994). The relationship of gender, level of sport involvement, and participation motivation to lask and ego orientation. International Journal of Sport Psychology. 25.4-18.

Williams, A.M., & Davids, K. (1995). Declarative knowledge in sport: A by-produci of experience or a characteristic of expertise? Journal of Sport & Exercise Psychology, 17, 259-275.

Cette etude se propose de mettre en evidence le role des interactions verbales en dyades parilaires dans t'apprentissage d'un sport collectif comme le handball. Les participants, 20 gar^-ons et 20 fdles. dgees de 11 a 12 ans. etaient repartis en deux groupes selon la condition d'appreniissage. La tdche est une situation dejeu reduite du handball: 2 attaquants contre 1 defenseur dans chaque moitie du terrain. Les participants du groupe experimental, beneficiant d'interactions verbales entre pairs, etaient supposes obtenir de meilleurs resultats aitx post-tests que ceux du groupe te/noin, n 'ayant pas interagi verbalemenl entre les sequences dejeu. Le second objectif etait d 'observer si le genre avait un effet sur les performances tactiques et sur les connaissances en jeu. L'experimentation demontre la superiorile du groupe beneftciant d 'interactions verbales en dyades par rapport au groupe n 'ayant pas interagi au regard des choix lactiques et du projet d'action. En outre des differences de genre out ete trouvees dans le sens oit les gar<,'ons obtiennent de meilleures performances que lesfilles au niveau de I 'efficaclte de I 'action. Ces resultats sont discutes dans la perspective theorique des travaux de psychologic sociale des acquisitions et du developpemenl et dans le courant de I'apprentissage par la comprehension. Cette etude experimentale demontre un interet pour une approche socio-constructivisle de la prise de decision en sports collectifs.

Key words: Dyads, Team handball. Verbal interaction.

Received: May 2003

Revision received: May 2004

Florence Darnis-Paraboschi. University Victor Segalen Bordeaux 2, Faculte des Sciences du Sport et de TEP, Laboratoire VSTIl, EA 498, 12 Av. Camille Jullian, 33607 Pessac Cedex, France. Professeur agrege d'E.P.S., docteur en STAPS a ITnstltut Universitaire de Fomiatlon des Maitres d'Aquitaine, 49 rucdc TEcoleNormale, BP219, 33021 Bordeaux Cedex, France; E-mail: [email protected]

Current theme of research:

Cognitive and socio-cogniiive factors on the foundations of tactical abilities in team sports.

184 F. DARNIS-PARABOSCHI, L. LAFONT, & A. MENAUT

Most relevant publications in the field of Psycholog}' of Education:

Damis, F., Lafoni, L., & Menaut, A. (2002). Effels des interactions dyadiques sur la competence tactique en sports colleclifs chez des sujets de 11-12 ans. Actes du congres de la SFPS (pp. 25-28), 2 1 . 22 mars 2002. Limoges, Krance.

Damis-Paraboshi, F. Lafont, L., & Menaut, A. (2004). Verbalisations en dyades dans une silualion de cooperation et d'affrontement en E.P.S. Communieation orale au Congres International: Faut-il parler pour apprendre? Dialogues. yerbali.tation et apprentissages en situation de travad a I'ecole: Acquis et questions vives. 17, 18, 19 mars 2004, Arras, France.

Darnis, F. (2004). Place et role du developpement ccgnitif et de ^interaction sociale dans I'acte tactique et jeii.x et sports collectifs. Unpublished doctoral dissertation in STAPS, University of Bordeaux 2, France.

Lucile Lafont. Maitre de conferences HDR, Universite Victor Segalcn Bordeaux 2, Faculte des Sciences du Sport el de I'EP, Laboratoire VST2I. EA 498, 12 Avenue Camille Jullian, 33607 Pessae Cedex; B-mail: [email protected]

Current theme of research:

Interactive imitation. Peers interaction. Tutoring. CoDperative learning. Modeling.

Most relevant publications in the field of Psychology of Education:

Lafont, L., & Winnykamen, F. (1999). Cooperation and competition in children and adolescents: Current issues and perspectives in physical education. In Y. Vanden Auweele, F Bakker, S. Biddle. M. Durand, & R. Seiler (Eds.), Textbook on psychology for physical educators (pp. 379-404). Champaign, IL: Human Kinetics.

Lafont L. (2002). Efficacit^ compar^e de la demonstration explicitee el de I'imitation-modelisation interactive pour l'acquisition d'une sequence dansee chez des adolescentes de 12 ^ 15 ans. STAPS, S8,69-79.

Lafont L., Bedouret T., Py Y., Ricordeau L., & Guilbeau H. (2003). Le tutorat en action: Observations et analyses. Recherche et Eormation, 43, 65-81.

Damis-Parabosbi, F. Lafont, L., & Menaut, A. (2004). Verbalisations en dyades dans une situation de cooperation et d'affrontetnent en E.P.S. Communication orale aii Congri-s International: Faut-il parler pour upprendre? Dialogues, verhalisalion et apprentissages en situation de travail a I'ecole: Acquis et questions vives, 17, 18, 19 mars 2004, Arras, France.

Munoz, E., & Lafont, L. (submitted). Adult tutoring for motor-skill acquisition: The role of different guidance methods in novices' motivation and performance.

Andr£ Menaut. Universite Victor Segalen Bordeaux 2, Faculte des Sciences du Sport et de I'Education Physique, EA 498 Vic Sportive: Tradition, Innovalicm, Intervenlion (VST2I}, 12 avenue Camiile Jullian, 33607 Pessac Cedex, France; E-mail: andre.menaut(@u-bordeaux2.fr

Current theme of research:

Cognition, strategies des acteurs et environnements sportifs. Cognition et phenom^nes collectifs dans les pratiques physiques et sportives.

Most relevant publications in the field of Psychology of Education:

Menaut, A. (1993). A propos d'un type particulier de conduite cognitive: La pens6e tactique. Bulletin de Psychologie. 4l2[XiM\). 684-693 (septembre-octobre, n" special "Cognition Education Langage").

Menaut, A. (1998). Le reel et le possible dans la pensee tactique. Contribution a une Iheorie dujeu sportif Bordeaux: PUB, Nov. 1998.

Menaut, A. (2002). L'art d'entrainer et la quete du sens: Etudes praxeologiques, L'exemple du football professionnel. Colloque International de Praxeologie Atotrice (20, 21 et 22 juin 2002). Paris.

Vallet, C , Goodall, C . & Menaut, A. (2000), L'ajustement spatio-teniporel du danseur face au stress. Science et

Motricite. 41,61-76.