Perceptualand Motor Skills, 1992, 75, 203-210.

O Perceptual and Motor Skills 1992

N O N C O N T I N G E N T POSITIVE A N D NEGATIVE FEEDBACK DURING MAXIMAL EXERCISE ' A. FOUAD KAMAL University of Otbwa

AND

CHRISTINE BLAIS Carleton Universirjr

Summary.-A closed skdl, maximal effort on a bicycle ergometer was used to study the effects of noncontingent external feedback on performance, perceived effort, and the subject's assessment of the quality and accuracy of the feedback. The 47 subjects included elite athletes as well as physical education majors. They were asked to perform a series of 5 60-sec. trials of maximal effort, trials which included positive, negative, positive and negative, and no-feedback conditions. Although there were no significant effects on performance for the various feedback conditions, subjects judged the positive feedback favourably and as accurate, but negative feedback was judged un£avourably and considered to be inaccurate. This result is seen as supportive of the 1982 McCarrey, et al. supposition that successful athletes may have less need to internahze attributions of effect.

Many theorists regard feedback as the single most important variable controlling the acquisition and performance of motor skills (Bilodeau & Bilodeau, 1961)) either as an information source (Adarns, 1971) or as a reinforcer in terms of motivation (Rushall & Siedentop, 1972) or self-confidence (Stewart & Corbin, 1988). Ulrich and Burke (1957) found that both knowledge of success and of failure yielded a greater work output; however, knowledge of failure elicited an emotional response to thls type of stress. The effectiveness of positive versus negative feedback, praise versus criticism, has often been debated. Where praise excels over criticism is in the fact that praise informs the learner of the correct response while criticism often identifies the incorrect response without defining what is correct (Kamal & Gallahue, 1980). According to Vallerand (1983) low, moderate, or high amounts of positive verbal feedback produce increased feelings of competence as well as an increase in the motivational level, whereas relatively frequent verbal punishment usually results in performance decrements (Allison & Ayllon, 1980; Curtis, Smith, & Smoll, 1979). Despite the support for positive feedback, Kirschenbaum and Smith (1983) argued that a mixture of positive and negative feedback might sometimes yield better results than positive feedback alone, a position given partial support by Kamal and Gallahue (1980) and by Ulrich and Burke (1957). I n contrast to the above, in their study of noncontingent negative feedback Benson and K e n n e l l ~(1976) showed that, when exposed to adverse srimulation, subjects' motivation and performance level decreased. Related to 'Address requests for reprints co A. F. Kamal, Ph.D., School of Human Kinetics, University of Ottawa, 125 University, Ottawa, Ontario, Canada K I N 6N5.

204

A . F. KAMAL & C. BLAIS

this, Weary (1980) and Miller and Ross (1975) stated that people tend to accept responsibility for positive but not for negative outcomes. This arises from a greater expectation for success than failure, leading them to perceive attributions in such a manner, the effect being stronger for those more accustomed to success. McCarrey, Edwards, and Rozario (1982), however, posit that people who are used to success have less need to i n t e r n d z e attributions than those who experience less success as their egos tend to be more balanced and secure. A related issue, individual responses to feedback, is the suggestion of differential effects dependent on gender. Lenny (1977) emphasizes that females lack confidence when performing unfamiliar tasks, while Corbin, Stewart, and Blair (1981) suggest that such differences only arise if the task is sex-role inappropriate. One of the other factors considered important in terms of influencing an individual's response to feedback is the perception of the amount of exertion (Monahan, 1988). Pandolf (1982) states that perceived exertion is not just a physiological event but is actively mediated by cognitive factors as well as motivation and emotion. I n addition, Rejeski (1985) asserts that learning and past experience continually alter the way people perceive their activity. Consequently, skilled athletes are thought to be more capable of monitoring their own progress than less skilled athletes (Markland & Martinek, 1988). Rejeski (1985) also stated that athletes who focused on external cues continued to exercise longer than those who focused on internal cues, implying that those focusing on external cues may be underestimating the intensity of work relative to those using internal cues. Related to this, Morgan and Pollock (1977), working with marathon runners, noted that the more elite athletes used associative strategies which made use of the perception of their exertion rather than dissociative strategies which attempted to block out any indication of extended effort or pain. The purpose of this study was to examine how individuals' response to feedback is influenced by their previous success and how they interpret the quality and accuracy of the feedback relative to perceptions of their own performance. Hence, this study included both elite and nonelite athletes who would have experienced differing levels of success. As there has been some debate about whether there are gender differences relative to the effects of various feedback conditions, the nonelite group of women was included. Also, as previous studies (Ulrich & Burke, 1957; Kamal & Gallahue, 1980) have confounded the effects of practice with the feedback conditions, it was important to separate these two factors. METHOD Subjects

Fourteen female and 18 male students in physical education from the University of Ottawa, and 15 men from the Canadian National Waterpolo

FEEDBACK DURING MAXIMAL EXERCISE

Team served as subjects. Their average age was 23 volunteers.

205

* 2.5 yr. AU were unpaid

Task

A bicycle ergometer with the tension fixed at 3.5 kp was used by all subjects throughout the study. Subjects were unable to view the tension setting, odometer, or timer. Wiring of the bicycle ergometer was done to create the impression that the readings were being electronically derived and recorded and so were the basis for the performance feedback. The feedback devices consisted of a standard desk-top bell and a clock buzzer. Procedure Each subject underwent five performance trials: a baseline trial plus three feedback trials and one no-feedback trial. The duration of each trial was 1 min., with a 7-min. rest between trials. Distance in kilometres per minute served as the dependent variable for each of the five trials, based on the odometer readings. Before each trial the subjects were instructed to d o their best and pedal as hard as they could. The experimenter then gave a warning command "ready" and a start command "go." The subjects were told that the first trial would represent the baseline against which their subsequent trials would be measured and that they would receive feedback on their subsequent trials relative to their baseline. A bell would signify an improvement, and a buzzer would signify a decreased or baseline performance. Every 10 sec. during each trial the subjects received, independent of actual performance, a random mixture of positive and negative feedback, or no feedback. Prior to the no-feedback trial, the subjects were informed that feedback would not be available but that they were still to do their best. Using a within-subject design and requiring all subjects to pedal without feedback on both the first trial (baseline) and on one subsequent trial, combined with the complete counterbalancing of the conditions over trials, permitted the separation of the effects of noncontingent feedback and practice. Immediately after completing all five trials subjects were asked to rate the effectiveness, accuracy, and the quality of their response to the various feedback conditions. They were also asked to identify the trials on which they pedalled fastest and slowest, and to estimate the number of revolutions pedaled on their final trial.

RESULTSAND DISCUSSION The findings are presented in three sections. The first section examines bicycle ergometer performance as a function of the feedback conditions. The second section presents the subjects' judgements of the effectiveness, accuracy, and their response to the feedback. Finally, brief characterizations of the accuracy of incidental memory for the effort required in pedalling are presented.

206

A. F. KAMAL & C. BLAIS

The data in Table 1 (distance in km/min.) are the mean and standard deviation for each feedback condition for each group. I t can be seen that performance did not vary as a function of the feedback condition. However, speed; the it is clear that the three groups differed substantially in women students pedaled the slowest and waterpolo players the fastest. An analysis of variance with feedback conditions as the within-subject factor confirms these conclusions. Neither the main effect of feedback conditions (F,,,,, = 0.83) nor the interaction between feedback condition and group (F,,,,= 1.30) achieved the .05 level of significance. However, the difference between groups was significant (F,,,,= 12.01, p < 0.01). Re-analysis of the data, upon collapsing over conditions, to examine the effects of practice showed that, while the trials following the baseline appeared to be faster than those at baseline, neither the effect of trials nor the interaction with groups was significant. TABLE 1 MEANSAND STANDARD DEVIATIONS FORPERFORMANCE (KM/MIN.) ON THE Feedback Condition

Men M

Women SD

M

SD

-

18

n

Baseline Positive Negative Mixed None

.85 .82 .84 .82 .82

--

14

.08 .10 .09 .11 .09

.67 .70 .70 .73 .69

.09 .13 .15 .19 .12

--

BICYCLEERGOMETER Waterpolo (Men) M SD 15 .82 .09 .86 .10 .87 .09 .85 .09 .86 .ll -

-

The data in Table 2 represent the proportion of subjects (after collapsing over all three groups) who selected each of the categories on the rating scale for the positive and negative feedback conditions. The percentages of subjects responding show clearly that positive feedback was judged positively and negative feedback was judged negatively, in terms of both their response to the type of feedback (feeling) and how accurately they felt it reflected their performance. Data in Table 3 show subjects also indicated that they found the feedback information influenced their performance, on average, improving their performance rather than hindering it. Lack of statistical evidence for an effect of feedback on bicycle ergometer performance indicates that the postperformance ratings are at variance. These judgments and evaluations of the various feedback conditions are not necessarily the consequence of an inabdity to monitor aspects of one's own performance. A summary of the accuracy of subjects' incidental memory for pedalhng speed is presented in Fig. 1. Separate analyses were performed for subjects' selection of the trial on which they pedaled the fastest and on which they pedaled the slowest. I n each case a discrepancy index was ob-

FEEDBACK DURING MAXIMAL EXERCISE

TABLE 2 SUBJECTS' POSTPERFORMANCE RATINGS IN POSITIVE AND NEGATIVE FEEDBACK BASEDON RATINGS OF ALL SUBJECTS CONDITIONS: PERCENTAGES Aspect Feeling

Accuracy

Category

Feedback Condition Positive (Bell) Negative (Buzzer)

Very annoyed Moderately annoyed Indifferent Moderately pleased Very P leased Always contrary More contrary than in accord As often contrary as in accord More in accord than contrary Alwavs in accord

TABLE 3

Category

8

Hindered greatly Hindered slightly No effect Improved slightly Improved greatly

0.0 8.6

12.7

66.0 12.7

tained based on the difference between the actual or correct trial and the judged trial. If the fastest trial was Trial 3 but the subject judged it to be Trial 2, this would represent a discrepancy of minus one. Since nearly identical distributions of results were obtained for the judgements of fastest and slowest trials, the responses were averaged over both judgements. From Fig. 1 we can see that the majority of instances fall into the zero category, indicating accurate judgement of pedaling speed (fastest or slowest trial), at least at an ordinal level. I t is also clear that, when errors in judgement occur, they reflect the proximity of the judged trial to the actual trial one of the actual trial. Fig. 2, with the majority of errors being within the incidental memory for perceived effort, exhibits the classic serial position curve typically found in studies of verbal learning and short-term memory. The curve gives the percentage of correct judgements as a function of which trial was actually correct. Subjects were most accurate in identifying the relative pedaling speed on the last of the five trials (a recency effect); and accuracy was next highest for the very first trial (a primacy effect), indicating that most of the judgement errors occurred when the fastest or slowest trial was one of the intermediate trials.

*

A . F. KAMAL & C. BLAIS

Discrepancy Index: Actual - Judged FIG.1. Discrepancy index (actual vs judged) for judgement of fastest and slowest trials

Finally, with a view toward examining the ability of subjects to quantify aspects of their perceived effort, the relationship between actual pedaling speed and estimated pedaling speed (in terms of the number of revolutions) on the final trial, was examined. Generally, a least squares linear regression analysis showed subjects tended to overestimate the number of revolutions pedaled and also provided evidence for a linear relationship between the actual number of revolutions pedalled and the number estimated. The Pearson product-moment correlation of .44 was reliably different from zero (F,,4,= 10.06, ~ ~ 0 . 0 5 ) . From the review of literature there should have been a clear opportunity for negative effects (Benson & Kennelly, 1976; Allison & Ayllon, 1980) or positive effects (Kirschenbaum & Smith, 1983), but this was not the case. The failure to find any reliable effect of noncontingent feedback on bicycle ergometer performance in the present experiment stands in clear contrast to earlier work. Ulrich and Burke (1957) and, more recently, Kamal and Gallahue (1980) found both positive and negative noncontingent feedback resulted in improvements in performance relative to an initial baseline trial. Unfortunately, both of these studies confounded the effects of practice with

FEEDBACK DURING MAXIMAL EXERCISE

Trial FIG. 2. Serial position curve for percentage of correct judgements (fastest or slowest trial)

feedback and, in light of the present results, their interpretation appears suspect. However, the Kamal and Gallahue study employed young children as ~ subjects while the present study used college students. ~ o t w i t h s t a n d i nthe differences in the design, the longer history of successful participation in sports common to subjects of the present study may have led them to be less sensitive to differential feedback techniques. The inconsistency in findings strongly suggests the need for further study with children, carefully controlling for practice effects, as in the design of the present experiment. Although the literature suggests that athletes who concentrate on external cues would tend to underestimate exertion (Rejeski, 1985), such was not the case in this study. Subjects were, in fact, quite accurate recalling the level of exertion for the first and last trials and tended to overestimate exertion on trials in between. Perhaps the magnitude and frequency of feedback provided in this study was insufficient to distract performers successfully from internal cues; it is clear that their internal representations of the skill were quite accurate. Alternatively, if higher levels of skill are paralleled by more accurate perceptions of effort, then we would not expect to find support for Rejeski's proposition in slulled athletes. From this study, using a task which requires low skill, rhe expertise represented by the subjects in the present

210

A. F. KAMAL & C. BLAIS

study may well have placed them at a point where the external feedback was not relevant. However, Markland and Martinek (1988) did suggest that skded athletes were better at monitoring their effort, and Morgan and Pollock (1977) also implied that athletes at higher levels made use of internd cues. Taken together, the implication may be that more highly slulled athletes are better able to monitor their internal states and so are less responsive to external factors-at least where the latter are viewed as offering little additional useful information. REFERENCES ADAMS,J. A. (1971) A closed-loop theory of motor learning. Journal of Motor Behavior, 3, 111-150.

ALLISON.M. G., & AYLLON,T. (1980) Behavioral coaching in the development of skills in football, gymnastics and tennis. lournal of Applied Behavior Analysis, 13, 297-314. BENSON, J. S.. & KENNELLY, K. J. (1976) Learned helplessness: the result of uncontrollable reinforcement or uncontrollable adverse stimuli? Journal of Personality and Social Psychofo,q?,34, 138-145. Buooe~uE , A , & BILODEAU, I. (1961) Motor skill learning. Annual Review of Psychology, 42, 243-280

C O R B ~C.,, STEWANT, M , & BLAIR,W. (1981) Self-confidence and motor performance of preadolescent boys and girls studied i n different feedback conditions. journal of Sport Psychology, 3, 90-34. Cmns, B., SMITH, R. E., & SMOLL,F. (1979) Scrutinizing the skipper: a study of leadership behavior i n the dugout. Journal of Applied Psychology, 64, 391-400. KAMAL,F., & GALLAHUE,D. (1980) Feedback: a developmental overview. The Physical Educator, 37, 215-219. KIRSCHENBAUM, D., & SMITH, R. (1983) A preliminary study of sequencing effects in simulated coach feedback. Journal of Sport Psychology, 5 , 332-342. LENNY,E. (1977) Women's self-confidence in achievement situations. Psychological Bulktin, 84, 1-13.

MARKLAND, R., & MARTINEK,T. (1988) Descriptive analysis of coach augmented feedback given to high school varsity female volleyball players. Journal of Teaching in Physical Education, 7, 289-301. MCCARREY, M., EDWARDS, H., & ROZARIO,W. (1982) Ego-relevant feedback, affect, and selfserving attributional bias. Personality and Social Psychology Bulletin, 8, 139-194. MILLER,D. T., & ROSS,M. (1975) Self-serving biases in the attribution of causality: fact or ficcion? Psychological Bulletin, 82, 213-225. MONAHAN, T. (1988) Perceived exertion: an old exercise tool finds new applications. Physician and Sports Medicine, 16, 174-176, 178-179. MORGAN, W. P., & POLLOCK,M. L. (1977) Psychologic characterization of the elite distance runner. Annals of the N . Y Academy of Sciences, 301, 382-403. PANDOLF, K. B. (1982) Differentiated ratings of perceived exertion during physical exercise. Medicine and Science in Sports, 14, 397-405. REJESKI, W. J. (1985) Perceived exertion: an active or passive process? Journal of Sport Prychology, 7, 371-378. RUS~IALL, B. S., & SIEDENTOP,D. (1972) The development and control of behavior in sport and physical education. Philadelphia, PA: Lea & Febiger. STEWART,M., BL CORBM,C. (1988) Feedback dependence among low confidence preadolescent boys and girls. Research Quarterly for Exercise and Sport, 59, 160-164. ULRICH,C., & BURKE,R. (1957) Effect of motivational stress upon physical performance. Research Qz~arterly,28, 403-412. VALLERAND, R. J. (1983) The effect of hfferential amounts of positive verbal feedback on the intrinsic motivation of male hockey players. Journal of Sport Psychology, 5, 100-107. WEARY,G. (1980) Examination of affect and egotism as mediators of bias in causal attributions. Journal of Personality and Social Psychology, 38, 348-357.

Accepted June 11, 1992.

Noncontingent positive and negative feedback during maximal exercise.

A closed skill, maximal effort on a bicycle ergometer was used to study the effects of noncontingent external feedback on performance, perceived effor...
283KB Sizes 0 Downloads 0 Views