Perceptual and Motor Skills, 2009, 109, 3, 831-840.

Perceptual and Motor Skills 2009

RELATIVE FREQUENCY OF KNOWLEDGE OF RESULTS AND TASK COMPLEXITY IN THE MOTOR SKILL ACQUISITION1
DALTON LUSTOSA de OLIVEIRA Universidade Presbiteriana Mackenzie Centro Universitrio Nove de Julho ROBERTO GIMENEZ Centro Universitrio Nove de Julho Universidade Cidade de So Paulo UMBERTO CESAR CORRA Universidade de So Paulo

LUCIANO BASSOANDGO TANI Universidade de So Paulo

Summary.The aim of this study was to investigate the effects of knowledge of results (KR) frequency and task complexity on motor skill acquisition. The task consisted of throwing a bocha ball to place it as close as possible to the target ball. 120 students ages 11 to 13 years were assigned to one of eight experimental groups according to knowledge of results frequency (25, 50, 75, and 100%) and task complexity (simple and complex). Subjects performed 90 trials in the acquisition phase and 10 trials in the transfer test. The results showed that knowledge of results given at a frequency of 25% resulted in an inferior absolute error than 50% and inferior variable error than 50, 75, and 100% frequencies, but no effect of task complexity was found.

It is well known that practice is fundamental for the acquisition of motor skills (Corra & Tani, 2005). However, practice must be optimized if efficiency of the learning process is the goal. The present study investigated one of the most important factors for the optimization of learning: knowledge of results (Salmoni, Schmidt, & Walter, 1984; Swinnen, 1996). It is a form of feedback that informs the learner about the result of a movement. Information related to movement execution is usually available to the learner through his own sensorial sources (intrinsic feedback). However, there are situations in which that information is absent or, when present, the learner may have difficulties in using it. In those conditions, an external source becomes necessary to provide the learner with that information (extrinsic feedback or knowledge of results). Knowledge of results can be provided in many ways by manipulating its frequency, precision, and temporal characteristics. Recently, frequency of knowledge of results (number of knowledge of results supplied in relation to the total number of trials) has been considered one of the most important variables that affect the acquisition of motor skills and, as such,
Address correspondence to Umberto Cesar Corra, Escola de Educao Fsica e Esporte, Universidade de So Paulo, Av. Prof. Mello Morais, 65, So Paulo SP, CEP 05508-030, Brazil or e-mail (umbertoc@usp.br).
1

DOI 10.2466/PMS.109.3.831-840

ISSN 0031-5125

832

D. L. de OLIVEIRA, ET AL.

has received the attention of many researchers (Chiviacowsky, 2000; Anderson, Magill, & Sekiya, 2001; Bruechert, Lai, & Shea, 2003; Tani, Meira Junior, & Gomes, 2005; Palhares, Lage, Vieira, Ugrinowitsch, & Benda, 2006; Laguna, 2008). For a long period of time, it was believed that frequent knowledge of results yielded better learning (Bilodeau & Bilodeau 1958; Bilodeau, Bilodeau, & Schumsky, 1959). However, this scenario changed drastically after the publication of a seminal paper in 1984, in which those early studies were criticized because they did not use a transfer or retention test to separate the transitory effects of performance from more permanent effects of learning (Salmoni, et al., 1984). In fact, many studies (Baird & Hughes, 1972; Castro, 1988; Wulf & Schmidt, 1989; Winstein & Schmidt, 1990; Wulf, Lee, & Schmidt, 1994; Chiviacowsky & Tani, 1997; Wrisberg & Wulf, 1997; Lai & Shea, 1998; Bruechert, et al., 2003), using the latter methodological approach, have found favorable results for low frequencies of knowledge of results, or at least have indicated that learning is not hindered by reduced frequencies. These results have been interpreted differently by means of three hypotheses: specificity, consistency, and guidance. The specificity hypothesis (Henry, 1968) refers to the similarity between the task practiced in the acquisition phase and in the retention test. In that sense, it questions the experimental design in relation to the retention test (accomplished without knowledge of results), which could facilitate the task for the subjects who are already familiar with low knowledge of results frequencies (experienced during the acquisition phase). The consistency hypothesis (Winstein & Schmidt, 1990) is based on the assumption that constant performance corrections induced by frequent knowledge of results could inhibit the acquisition of consistency in the execution of movement, which would make retention more difficult. Finally, according to the guidance hypothesis (Salmoni, et al., 1984), frequent knowledge of results could act as a guide for the learner toward the goal of the task during the acquisition trials. This orientation could generate a certain amount of dependency of the learner in relation to external information, inhibiting or interfering with other processing activities such as detection and correction of errors and elaboration of the motor plan (Schmidt, 1991). Most of the studies on knowledge of results frequency were carried out using relatively simple tasks in a laboratory environment. However, Godinho and Mendes (1996) and Swinnen (1996) have pointed out the necessity of more studies which focus on complex tasks and pay more attention to ecological validity. The relationship between frequency of knowledge of results and complexity of the task was initially explored in studies that used summary knowledge of results. Schmidt, Young, Swinnen, and Shapiro (1989), using a simple task, found better results for large numbers of trials summarized. On the other hand, Schmidt, Lange, and Young

KNOWLEDGE OF RESULTS AND TASK COMPLEXITY

833

(1990), in a study involving a more complex task, found favorable results with small knowledge of results summaries. Guadagnoli, Dornier, and Tandy (1996; Exp. 2) compared the size of the knowledge of results summary, the complexity of the task, and the learning phases. The results showed an interaction between the size of the summary and the complexity of the task only for the groups at initial phases of learning. In relation to the groups at advanced phases, the interaction between summary of knowledge of results and complexity of the task was only partially confirmed. Using a task that involved the control of several degrees of freedom, Wulf, Shea, and Matschiner (1998; Exp. 2) investigated the frequency of knowledge of results effects (control, 100, and 50%) in the learning of a slalom movement in a ski simulator. The results showed better learning for the 100% frequency group in relation to 50% and control groups. In another study with variations in task complexity and arrangements of knowledge of results controlled by the experimenter as well as self-controlled, Chiviacowsky (2000) found no interactions. However, as the amount of practice differed for the groups, the analysis of the true effects of the task complexity was difficult to measure. In summary, the effect of relative frequency of knowledge of results considering task complexity on motor skill acquisition has not been thoroughly studied yet. Therefore, the aim of the present study was to investigate the hypothesis that for tasks with high complexity, high frequencies of knowledge of results will be necessary. Since task complexity is defined by the number of elements and their interactions, complex tasks would require higher motor control, and as a consequence the possibility of performance errors increases. This could imply the need for more corrections and, therefore, demand for more information about the results of the movement (higher frequencies of knowledge of results). Method Participants One hundred and twenty children who had no experience with the experimental task, both boys (n=60) and girls (n=60), with ages between 11 and 13 years (M=11.8, SD=1.2), voluntarily participated in this experiment. The children were all students of a private school in So Paulo, Brazil. The participation required the written consent of those responsible for the children. Children in this age range were chosen as participants instead of adults in order to reduce the experience effect, but assuring that they were able to learn the experimental task. Task and Apparatus The tasks consisted of two bocha game throwing patterns. It consisted

834

D. L. de OLIVEIRA, ET AL.

of throwing the ball (bocha) and placing it as close as possible to the target (bolim). The simple task consisted of throwing the ball with a backward-forward pendulous movement of the extended arm. The complex task was performed by the same pendulous movement followed by an overhead circular movement of the arm, thus involving the control of a greater number of muscles, that is, degrees of freedom. The dependent variable measure was the distance from the target in centimeters. The balls to be thrown were made of polyester (weight=100 g; diameter=5 cm). The target consisted of a small ball weighing 500 g and measuring 8 cm in diameter. Both balls were those used in the official Bocha game. During the trials, a mobile fence was used to block the ball trajectory from the students line of vision and prevent knowledge of results. The mobile fence was made of black fabric and metal connecting rods (1.70 m high, 1.50 m wide). Procedure and Design The subjects were randomly assigned into eight groups (ns=15) according to frequency of knowledge of results (25, 50, 75, and 100%) and complexity of the task (simple and complex). In the acquisition phase the subjects performed 90 trials. The transfer test was conducted 10 min. after the end of the acquisition phase and consisted of 10 trials without knowledge of results. The interval between the trials was approximately 30 sec. The participants were requested to throw the ball, aiming to place it as close as possible to the target positioned at a distance of 12 m. After each throw, the experimenter positioned the fence in front of the subject in order to block the line of vision (see Fig. 1).

1m

12 m

Place of Throw

Mobile Fence Target

Fig. 1. Schematic illustration of the experimental environment

KNOWLEDGE OF RESULTS AND TASK COMPLEXITY

835

Results The data were analyzed in blocks of 10 trials in terms of absolute (Table 1) and variable (Table 2) errors. For the acquisition phase, analysis of variance (ANOVA) was carried out with repeated measures on the last factor, 2 (complexity)4 (frequency)9 (blocks), and for the transfer phase a factorial ANOVA, 2 (complexity)4 (frequency), was conducted.
TABLE 1 Mean of Absolute Error (cm) in Blocks of 10 Trials, in the Acquisition (Blocks 1 to 9) and Transfer Phases (T), Independent of the Complexity of the Task Group Task Simple Freq. 1 231 249 338 228 266 303 332 250 2 194 221 222 159 243 183 203 152 3 177 221 231 139 240 144 244 157 4 183 189 204 130 265 155 204 150 100% 75% 50% 25% Complex 100% 75% 50% 25% Acquisition 5 197 195 207 127 257 135 203 133 6 153 211 222 127 227 148 177 131 7 162 185 191 113 213 98 191 131 8 144 166 227 102 217 108 184 125 9 145 156 184 103 193 92 183 114 181 206 228 175 221 152 222 168 Transfer

Absolute Error For the acquisition phase, the main effects for knowledge of results frequency (F3,112=15.92, p<.01; 2=0.29), blocks (F8,896=41.30, p<.01; 2=0.26), interaction between complexity and frequency (F3,112=6.88, p<.01; 2= 0.15), and interaction between frequency and blocks (F24,896=1.90, p<.01; 2=0.05). In terms of knowledge of results frequency, the Tukey HSD test showed that 25% frequency was better than other frequencies ( p<.05), and that 75% was better than 50% frequency. Regarding differences among blocks, the Tukey HSD test showed that the absolute error diminished in the acquisition phase ( p<.05). In relation to the interaction between task complexity and knowledge of results frequency, the Tukey HSD test showed that 100% frequency in the complex task was superior to 100% frequency in the simple task ( p<.05). It was also observed that in both tasks, 25% frequency had absolute error inferior to that in the blocks with 50% and 75% frequencies ( p<.05). The Tukey HSD test also showed that in the complex task, 25% frequency had absolute error inferior to that of the blocks with 100% frequency. Finally, concerning interaction between knowledge of results frequency and blocks, the Tukey HSD test showed that all the knowledge of results frequencies diminished the absolute error ( p<.05). In the transfer test, significant effects for knowledge of results frequency were found (F3,112=2.70, p<.05; 2=0.07). The Tukey HSD test showed that results were superior in the blocks with 25% knowledge of

836

D. L. de OLIVEIRA, ET AL.

results frequency than in those with 50% knowledge of results frequency ( p<.05). Variable Error In the acquisition phase, there were significant main effects for kno wledge of results frequency (F3,112=39.91, p<.01; 2=0.51), blocks (F8,896= 22.79, p<.01; 2=0.16 ), interaction between task complexity and knowledge of results frequency (F3,112=5.85, p<.01; 2=0.14), and interaction between knowledge of results frequency and blocks (F24,896=1.56, p<.05; 2=0.04). The Tukey HSD test showed that performance in blocks with 25% knowledge of results frequency was better than that in blocks with other frequencies (p<.05), and that 75% knowledge of results frequency was better than 100% and 50% frequencies. In relation to blocks, the Tukey HSD test indicated that variable error in Block 1 was superior to that in the remaining blocks, and Blocks 2 and 3 had variable error superior to Blocks 7 and 9 (p<.05). There was a significant interaction between know ledge of results frequency and task complexity. In the simple task, performance was better under the 25% knowledge of results frequency than at other frequencies, and in the complex task, the 25% and 75% knowledge of results frequencies had variable error inferior to that at the 100% and 50% frequencies. With respect to the interaction between knowledge of results frequency and blocks, at all knowledge of results frequencies but 100%, variable error decreased over blocks (Table 2).
TABLE 2 Mean of Variable Error (cm) in Blocks of 10 Trials, in the Acquisition (Blocks 1 to 9) and Transfer Phases (T), Independent of the Complexity of the Task Group Task Simple Freq. 1 152 168 230 92 180 207 243 126 2 119 135 142 53 146 106 160 63 3 111 133 145 47 140 73 207 52 4 108 119 124 49 172 82 120 58 100% 75% 50% 25% Complex 100% 75% 50% 25% Acquisition 5 129 117 131 46 162 76 125 51 6 97 125 148 53 133 77 112 52 7 90 113 115 43 126 40 121 49 8 90 98 142 39 135 49 129 48 9 90 78 127 34 114 43 109 42 101 145 149 63 132 83 142 65 Transfer

In the transfer test, the main effect for knowledge of results frequency was significant (F3,112=7.98, p<.01; 2=0.17). The Tukey HSD test showed inferior variable error at 25% knowledge of results frequency than at other frequencies. Discussion The objective of this study was to investigate the effects of relative fre-

KNOWLEDGE OF RESULTS AND TASK COMPLEXITY

837

quency of knowledge of results and task complexity on motor skill acquisition. Overall, the results did not show an effect of task complexity. In the field of motor learning, task complexity has been traditionally classified in various ways; for example, perceptual complexity, decision complexity, and effector complexity (Billing, 1980). The concept of complexity is usually defined as the number of parts or components of a task (Nussenzveig, 1999; Ward, 2002). In this study, throwing the ball with a pendulous movement followed by an overhead circular movement of the arm was considered a more complex task than throwing the ball using only a pendulous movement of the arm, because it involved the control of a greater number of muscles. However, the results allow the speculation that the difference in complexity was not evident, that is, adding a circular movement actually increased the effector complexity of the task but it did not result in a greater motor control demand because it was a well practiced movement. In relation to knowledge of results frequency, the results showed that the subjects who practiced with 25% frequency obtained better performance in the transfer test. The subjects of this group had higher accuracy than those of the 50% knowledge of results frequency group and more consistency than those of the groups experiencing 50%, 75%, and 100% frequencies. A possible explanation is that when knowledge of results is not always available, the learner involves a more active process of searching for information to relate it to intrinsic feedback and construct a reference for correction of the task. In other words, reduced knowledge of results frequency exerts a small guiding function, contributing to the enhancement of error detection and correction. What possibly occurs is that after the execution of a movement, the learner involves himself in a cognitive process to relate the following information: action goal, initial conditions, response specifications, sensory consequences, and movement outcome (Schmidt, 1975). The participant retains in short-term memory the results of the established relationships between these information sources when knowledge of results is available and then develops further control in the trials without knowledge of results in the following way: information related to initial conditions and action goal does not change from trial to trial; the relation between specifications of the response and sensory consequences strengthens in the trials with knowledge of results, which contributes to the refinement of corrections, making learning more effective. When there is no knowledge of results available, the learner tends to maintain the relation between response and consequences in the next trials, until receiving knowledge of results again when able to evaluate whether the relationship is correct or not (Tani, et al., 2005). Another important aspect that should be considered in this study is the refutation of the orientation hypothesis (Salmoni, et al., 1984). The ori-

838

D. L. de OLIVEIRA, ET AL.

entation hypothesis states that frequent knowledge of results could guide the performance in the direction of the goal during acquisition and tends to deteriorate during transfer. The present study did not confirm that hypothesis. This is evident when the performance of the groups given 100% and 25% knowledge of results frequency in the acquisition phase is analyzed. As previously described, the group receiving 100% knowledge of results had poor performance during acquisition and maintained that performance in the transfer test; the group receiving 25% knowledge of results, on other hand, had high performance in both acquisition and transfer tests. The results of the present study did not show a linear relationship for the utilized knowledge of results frequencies (25%, 50%, 75%, and 100%). Linearity probably does not occur because, besides the frequency, the uncertainty (amount of information) is another important element that can affect the use of knowledge of results. In agreement with Shannon and Weaver (1949), the more unlikely the occurrence of a sign in a message, the more informative a posteriori is its occurrence. In this sense, the group receiving 25% knowledge of results frequency worked under a lot of uncertainty (the learner received knowledge of results only once in every four trials); the group receiving 50% knowledge of results frequency had more information and so less uncertainty, and so forth. In other words, knowledge of results issued at different frequencies supplies different amounts of information. Finally, the present results should be interpreted with some caution because a delayed transfer test was not used. Despite some evidence regarding similar effects of knowledge of results frequency and other variables related to extrinsic feedback on immediate and delayed transfer tests (Vander Linden, Cauraugh, & Greene, 1993; Weeks & Kordus, 1998; Anderson, et al., 2001; Badets & Blandin, 2004; Ishikura, 2005; Palhares, et al., 2006), the results are still controversial, suggesting that the time interval before retention tests is a crucial manipulation when studying permanent changes in behavior.
REFERENCES

Anderson, D. I., Magill, R. A., & Sekiya, H. (2001)Motor learning as a function of knowledge of results schedule and characteristics of task-intrinsic feedback. Journal of Motor Behavior, 33, 59-66. Badets, A., & Blandin, Y. (2004)The role of knowledge of results frequency in learning through observation. Journal of Motor Behavior, 36, 62-70. Baird, I. S., & Hughes, G. H. (1972)Effects of frequency and specificity of information feedback on acquisition and extinction of a positioning task. Perceptual and Motor Skills, 34, 567-572. Billing, J. (1980)An overview of task complexity. Motor Skills: Theory Into Practice, 4, 18-23.

KNOWLEDGE OF RESULTS AND TASK COMPLEXITY

839

Bilodeau, E. A., & Bilodeau, I. M. (1958)Variable frequency of knowledge of results and the learning of a simple skill. Journal of Experimental Psychology, 55, 379-383. Bilodeau, E. A., Bilodeau, I. M., & Schumsky, D. A. (1959)Some effects of introducing and withdrawing knowledge of results early and late in practice. Journal of Experimental Psychology, 58, 142-144. Bruechert, L. M., Lai, Q., & Shea, C. H. (2003)Reduced knowledge of results frequency enhances error detection. Research Quarterly for Exercise and Sport, 74, 467472. Castro, I. J. (1988)Efeitos da freqncia relativa do feedback extrnseco na aprendizagem de uma habilidade motora discreta simples. Dissertao de Mestrado. Unpublished master thesis, University of So Paulo, Brazil. Chiviacowsky, S. (2000)Efeitos da freqncia do conhecimento de resultados controlada pelo experimentador e auto-controlada pelos sujeitos na aprendizagem de tarefas motoras com diferentes complexidades. Tese de Doutorado. Unpublished doctoral dissertation, Faculdade de Motricidade Humana, Universidade Tcnica de Lisboa, Portugal. Chiviacowsky, S., & Tani, G. (1997)Efeitos da freqncia do conhecimento de resultados na aprendizagem de diferentes programas motores generalizados. Revista Paulista de Educao Fsica, 11, 15-26. Corra, U.C., & Tani, G. (2005)Estrutura de prtica e processo adaptativo em aprendizagem motora: por uma nova abordagem da prtica. In G. Tani (Ed.), Comportamento motor: desenvolvimento e aprendizagem. Rio de Janeiro: Guanabara Koogan. Pp. 141-161. Godinho, M., & Mendes, R. (1996)Aprendizagem motora: informao de retorno sobre o resultado. Lisbon, Portugal: Edies Faculdad de Motricidade Humana. Guadagnoli, M. A., Dornier, L. A., & Tandy, R. D. (1996) Optimal length for summary knowledge of results: the influence of task-related experience and complexity. Research Quarterly for Exercise and Sport, 67, 239-248. Henry, F. M. (1968)Specificity vs generality in learning motor skill. In R. C. Brown & G. S. Kenyon (Eds.), Classical studies on physical activity. Upper Saddle River, NJ: Prentice Hall. Pp. 331-340. Ishikura, T. (2005)Average knowledge of results schedule in learning of timing: influence of length for summary knowledge of results and task complexity. Perceptual and Motor Skills, 101, 911-924. Laguna, P. L. (2008)Task complexity and sources of task-related information during the observational learning process. Journal of Sports Sciences, 26, 1097-1113. Lai, Q., & Shea, C. H. (1998)Generalized motor program (GMP) learning: effects of reduced frequency of knowledge of results and practice variability. Journal of Motor Behavior, 30, 51-59. Nussenzveig, H. M. (1999)Complexidade e caos. Rio de Janeiro, Brazil: Editora da Universidade Federal do Rio de Janeiro. Palhares, L. R., Lage, M., Vieira, M. M., Ugrinowitsch, H., & Benda, R. (2006) Knowledge of results-delay interval effects in the acquisition of a serial skill of different compatibility levels. Journal of Human Movement Studies, 51, 47-61. Salmoni, A. W., Schmidt, R. A., & Walter, C. B. (1984)Knowledge of results and motor learning: a review and critical reappraisal. Psychological Bulletin, 95, 355-386.

840

D. L. de OLIVEIRA, ET AL.

Schmidt, R. A. (1975)A schema theory of discrete motor skill learning. Psychological Review, 82, 225-260. Schmidt, R. A. (1991)Frequent augmented feedback can degrade learning: evidence and interpretations. In J. Requin & G. E. Stelmach (Eds.), Tutorials in motor neuroscience. Dordrecht, The Netherlands: Kluwer Academic. Pp. 59-75. Schmidt, R. A., Lange, C., & Young, D. E. (1990)Optimizing summary knowledge of results for skill learning. Human Movement Science, 9, 325-348. Schmidt, R. A., Young, D. E., Swinnen, S., & Shapiro, D. C. (1989)Summary knowledge of results for acquisition: support for the guidance hypothesis. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 352-359. Shannon, C. E., & Weaver, W. (1949)The mathematical theory of communication. Champaign, IL: Univer. of Illinois Press. Swinnen, S. P. (1996)Information feedback for motor skill learning: a review. In H. N. Zelazninik (Ed.), Advances in motor learning and control. Champaign, IL: Human Kinetics. Pp. 37-66. Tani, G., Meira Junior, C. M., & Gomes, F. R. F. (2005)Frequncia, preciso e localizao temporal de conhecimento de resultados e o processo adaptativo na aquisio de uma habilidade motora de controle de fora manual. Revista Portuguesa de Cincias do Desporto, 5, 59-68. Vander Linden, D. H., Cauraugh, J. H., & Greene, T. A. (1993)The effect of frequency of kinetic feedback on learning an isometric force production task in nondisabled subjects. Physical Therapy, 2, 79-87. Ward, L. M. (2002)Dynamical cognitive science. Cambridge, MA: The MIT Press. Weeks, D. L., & Kordus, R. (1998)Relative frequency of knowledge of performance and motor skill learning. Research Quarterly for Exercise and Sport, 69, 224-230. Winstein, C. J., & Schmidt, R. A. (1990)Reduced frequency of knowledge of results enhances motor skill learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 677-691. Wrisberg, C. A., & Wulf, G. (1997)Diminishing the effects of reduced frequency of knowledge of results on generalized motor program learning. Journal of Motor Behavior, 29, 17-26. Wulf, G., Lee, T., & Schmidt, R. A. (1994)Reducing knowledge of results about relative versus absolute timing: differential effects on learning. Journal of Motor Behavior, 26, 362-369. Wulf, G., & Schmidt, R. A. (1989)The use of generalized motor programs: reducing the relative frequency of knowledge of results enhances memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 15, 748-757. Wulf, G., Shea, C. H., & Matschiner, S. (1998)Frequent feedback enhances complex skill learning. Journal of Motor Behavior, 30, 180-192. Accepted November 13, 2009.