AN ANALYb 3‘3 GFTE'ZE 511337 US" 111111111: 11. :11 11 1.11111. 7'3 31111111 11111813110111 u “1’131'11‘11113. A *1 ,, TVICthuAfii‘i s10"; 4-111.” 1-0511. I {17' p\ Anu: n . «up, iv;1fi-. DULURL'Q'HVAVTI.u'thV'1fiu1 " 1975 LInRAnY BINDER. "- mmnmluml .n . \— A \l _ 3-. q I gum; Inzllyujml Lm gnu)» fl @W H j 1!qu ABSTRACT AN ANALYSIS OF THE EFFECT OF VIDEOTAPE REPLAY ON LEARNING TO SWIM THE BREASTSTROKE BY Dolores Cady Hageman The purpose of the study was to analyze the effec- tiveness of videotape replay in learning to swim the breaststroke. Students enrolled in four intermediate swim classes at Calvin College served as subjects. Classes were randomly assigned to treatment or control groups. It was hypothesized that the treatment subjects who received visual and verbal information feedback via videotape replay would improve more than the control subjects whose information was restricted to verbal feedback in learning to swim the breast- stroke. The procedures in effect during the experiment were as follows: the Kilby glide test plus time was administered during the third class session. The four classes were taught by the same instructor. Lessons were divided into two phases: an instruction and drill phase, and a feedback and workout phase. During the first phase, subjects in all classes were Dolores Cady Hageman taught the same way including directions, instructions, demonstrations and drills. During the second phase of each lesson subjects performed the prescribed drill in the deep water tank. Subjects in the treatment group were videotaped at this time. Following the drill, subjects practiced according to a posted schedule. During the practice portion of the lesson the subjects in the treatment group individ- ually viewed their videotaped performance and received verbal error information feedback whereas the control subjects received only individual verbal feedback information. All subjects were asked to refrain from practice outside of class. The Kilby glide test plus time which required the subjects to swim the breaststroke for twenty-five yards, was again administered during the seventh week of the study. Subjects were randomly assigned to a testing time during the evening testing session. Four dependent variables were obtained from the test-~number of strokes, number of errors, total count (strokes and errors), and the time required to complete that distance. Descriptive and inferential statistics were computed for the pretest, posttest, and gain score data of these measurements. The treatment group mean scores revealed that they performed better initially than did the control group. Significant differences between groups were determined through the use of t-tests. The results revealed that there were no significant differences between the treatment and Dolores Cady Hageman control group gain scores at the .05 level. Tests for significant differences between pretest and posttest scores demonstrated that improvement occurred for both groups in the technique variables, but not in the time variable. It was concluded that a combination of videotape and verbal feedback was not more effective than verbal feedback alone in learning to swim the breaststroke. Improvement scores in technique variables rather than the product variable (time) indicated that learning was affected by the instructor's objectives. AN ANALYSIS OF THE EFFECT OF VIDEOTAPE REPLAY ON LEARNING TO SWIM THE BREASTSTROKE BY Dolores Cady Hageman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Health, Physical Education, and Recreation 1976 ACKNOWLEDGMENTS To Dr. John Haubenstricker, the author extends her grateful appreciation for serving as chairman of the guidance committee. Dr. Haubenstricker's pursuit of professional and intellectual excellence has made this research study very educational. A note of gratitude also to Dr. William Heusner and Dr. Vern Seefeldt who served on the guidance committee. Dr. Heusner's guidance and help concerning statistics and data processing was invaluable. Dr. Seefeldt's capable guidance and advise added to the quality of this pursuit. A special note of gratitude is also extended to those members of the Calvin College staff who gave encouragement and help in this project. A special thanks to Dr. James Timmer for his guidance and time in editing. Thanks also to Dr. Marvin Zuidema, Chairman of the Department of Physical Education for making the subjects available for the project. Acknowledgment to Wayne Dyksen and Dr. Sanford Leetsma for their assistance with the computor work; and to Steve Veeneman who served as the videotape technician. Thanks to to the students who willingly and faithfully participated and the Physical Education majors who served as assistants. ii Finally, special thanks to my family especially my parents, Wilifred and Wilma Cady, for their confidence and encouragement. Acknowledgment is also made to my husband, Steven, for his patience and love throughout. iii TABLE OF CONTENTS LIST OF TABLES. . . . . . . Chapter I. THE PROBLEM. . . . . . Need for the Study. . . Statement of the Problem. Research Hypotheses . . Research Plan . . . . Rationale. . . . . . Delimitations . . . Significance of the Study Definitions . . . . . II. RELATED LITERATURE . . . Feedback . . . . . . Definition of Feedback. Reference Patterns . . Error Correction. . Factors Affecting Feedback Arrival Time. . . . . Level of Learning . . . Nature of the Task. . . Studies using Object Manipulation Studies using Body Manipulation. Summary . . . . . . I I I 0 METHOD 0 O O O O O 0 Population . . . . . Sample. . . . . . . iv Page vii Chapter Experimental Design. . . . Independent Variables . . Control of Extraneous Independent Variables. . Research Design . . . . Conduct of Treatment. . . Dependent Variables. . . . Measurement Tools. . . . Instrument Calibration . . Scoring and Recording Procedures . . . . . Data Collection Procedures . . . . . Treatment of Data . . . . Statistical Analysis. . . Data Presentation. . . . IV. RESULTS AND DISCUSSION . . . Statistical Results. . . . Descriptive Statistics . . Inferential Statistics . . Discussion. . . . . . . Learning Effects . . . . Differences between Methods V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS . . . . . Summary of Results . . . . Conclusions . . . . . . Recommendations . . . . . APPENDICES A. Swim Questionnaire. . . . . B. Outline of Lesson Activities . Page 46 46 46 47 48 51 52 54 56 57 57 57 58 60 60 61 63 68 68 71 76 77 79 80 82 84 Chapter Page C. Timed Kilby Stroke Count Test. . . . . . . 90 Testing Procedure . . . . . . . . . . 90 Standardized Directions . . _. . . . . . 91 Recorder's Form. . . o o o o o o o o 0 (.0 LIST OF REFERENCES . . . . . . . . . . . . 94 vi LIST OF TABLES 1. Descriptive Statistics for Pretest Measures 0 O O O O I O O 0 O O 2. Descriptive Statistics for Posttest measures. 0 O p O O O O C O O 3. Descriptive Statistics for Gain Scores . . . . . . . . . . . 4. F Values for Gain Scores . . . . . . 5. T-tests for differences Between Groups I O O 0 O O O O O O O 6. Power Values for Dependent Variables . . 7. T-Values of Learning Effects for Dependent Variables . . . . . . . 8. Frequency of Posttest Score Increases for the Dependent Variables. . . . . . vii Page 61 62 63 64 65 65 66 67 CHAPTER I THE PROBLEM One of the fundamental purposes involved in the use of movie films or videotape replay is that of providing information feedback to the performer. Through the replay, the performer receives visual information concerning the performance and can determine if the response accurately duplicates the original skill concept. This information often results in correction of errors during the next per- formance. Some investigations have focused on the value of repeatedly providing the performer with correct skill pat- terns via films of expert performers. However, the necessity of regular observance of a skilled reference pattern has not been demonstrated (Nelson, 46, wyness, 71; Brown and Messer- smith, 12; Carre, l4; Newfeld and Newfeld, 47; Caine 13; Lloyd, 41). Carre (14) suggested that if an initial under- standing of the skill pattern is adequate, repeated model showings are unnecessary. Although repeated observanCe of model performance is not essential for error corrections, review of one's own performance can result in reassessment and change in performance (Wren, 70; DeBacy, 22; Clifton and Smith, 17; Watkins, 68). The effect of visual feedback on performance is influenced by several factors. These factors include arrival time, stage of learning, and the nature of the task (Robb, 56). Films supply delayed visual feedback due to the time required for processing. Such delayed feedback is not as effective as the immediate feedback provided by videotape in learning various gross motor skills (Lloyd, 41; Jones, 35; Priebe and Burton, 54; Brown and Messersmith, 12). The effectiveness of augmented visual feedback at various levels of skill is not clearly established. Several studies have yielded significant results using beginners as subjects (Hainfeld, 31; Thompson, 66; Morgan, 45; Paulat, 49; Patterson, 48; Del Rey, 23). Others suggest that this type of information feedback is more effective after the initial learning stages (Nelson, 46; Berlin, 8; Jones, 35; Lawther, 39; Plese, 52; Green, 30; Carre, 14; Watkins, 68). Some studies suggest that learners at the middle skill levels benefit the most from such information feedback (Lloyd, 41; Taylor, 65). The nature of the task affects the value of immediate feedback as provided by videotape replay. Although skills involving object projection do not provide direct performance information, they do provide inherent knowledge of results through observation of the distance or the effect of the projectile. The acquisition of object manipulation tasks has not been shown to improve significantly with additional videotape feedback irrespective of whether skill technique or results of performance were measured. Carre (l4), Gasson (28), Sanders (60), Caine (13), and Cooper (19) found no significant improvement for groups receiving augmented visual feedback in skills such as the discus throw, badminton wall volley, volleyball wall volley, bowling, and certain basket- ball skills, respectively. However, Cooper (19) and Del Rey (23) concluded that videotape replay significantly improved performance in skills emphasizing form and accuracy. Augmented feedback enhances the learning of skills in which little intrinsic information is available. Perfor- mance of closed skills in general, and body manipulation skills specifically, has been shown to improve significantly through the use of videotape information feedback. These are skills in which the performer cannot see performance results environmentally. Plese (52) obtained improvement in parallel bar performance with the videotape replay group. He suggested that videotape replay is significant for skills in which the performer's body and/or body segment position- ing affects the skill. Other studies obtained similar results for modern dance (Patterson, 48) and swimming, (Morgan, 45; Green, 30; Newfeld and Newfeld, 47; and Taylor, 65). The effectiveness of videotape feedback as a teach- ing method for body manipulation skills is apparent from these results. Due to the nature of the environment in which it is performed, swimming is a body manipulation skill for which there are limited means of providing immediate feedback. Therefore, the effectiveness of videotape replay in teaching a swim stroke is a viable problem. Need for the Study The nature of swimming is such that the participant cannot view the execution of the skills inherent in this activity. Classes in swimming skills provide instruction in stroke execution--technique. Since swimming skills emphasize form and body manipulation, the use of augmented feedback may be beneficial (Cratty, 21). Most studies using augmented feedback have involved beginning level learners. Many results indicate, however, that beginners have too many variables to be concerned with to benefit from television replay (Plese, 52; Morgan, 45; Berlin, 8; Lloyd, 41; Carre, l4; Lawther, 39). It has been suggested that augmented feedback is more helpful after beginners are more familiar with the task. The present study dealt with intermediate swimmers. Most intermediate swimming classes include instruc- tion in the breaststroke. Although intermediate swimmers may not display correct technique, most of them can demon- strate a minimal knowledge of what the breaststroke looks like and may have attempted to perform it. Therefore, the skill involved in the study was not a novel skill to the learners. On the other hand, few intermediate swimmers are able to demonstrate an accurate or efficient breaststroke. Few studies have focused on the value of videotape replay in helping the swimmer refine or even relearn a skill. Several authors have used underwater photography to analyze swim strokes (Alterveen, 2; McCatty, 44), but few have used underwater videotape filming to provide information feedback. Taylor (65) videotaped his subjects from the deck of the pool. Green (30) taped below the surface three times but did not determine if these tapes made any difference in the rate of learning pre-beginning and beginning swim skills. The nature of the environment in which swimming strokes are performed causes some unique problems with surface filming. Refraction causes distortion whereas splashing and surface waves cause concealment (McCatty, 44). On the other hand, underwater photography allows the student to see arm and leg movements underneath or below the body. Also, most of the instructional models seen on charts and diagrams use under- water views of swimmers. Therefore a study using underwater filming may provide useful information in the effectiveness of augmented feedback in learning a swimming stroke. Statement of the Problem The purpose of this study was to investigate the effectiveness of videotape replay as a teaching device in learning to swim the breaststroke. Specifically, this study compared the effectiveness of augmented (supplementary), terminal, visual and verbal feedback via videotape replay with the verbal feedback provided by the instructor alone. The objective was to compare learning of the breaststroke between two groups of college, intermediate swimmers. One group received videotape replay of the breaststroke perfor- mance during practice time while the other group received practice time without videotape replay. The difference between the pre- and post test scores on the Kilby glide test plus time determined the effectiveness of the videotape replay method. Research Hypotheses It was hypothesized that teaching with videotape replay would result in significantly more learning than teaching without replay for intermediate swimmers learning the breaststroke. Thus, it was hypothesized that students in the swim classes taught the breaststroke with the use of videotape replay would have better performance scores (the difference between the pre- and post test scores) than those taught without the replay. Research Plan The students enrolled in four coed Swim II classes held at Calvin College, Grand Rapids, Michigan, served as subjects. The Kilby glide test plus time was given to obtain initial performance scores. Classes were randomly assigned to treatment or control groups. Both groups received conventional instructions followed by a prescribed practice period. Stroke demonstrations were videotaped and shown to all classes. The instructional phase of each lesson (lasting approximately twenty minutes) concluded with each swimmer performing the stroke or part of the stroke in the deep water tank. The order for this drill was randomly assigned and rotated daily. During this drill, the experi- mental group was videotaped. The instructor made notes of each individual's problems with technique at this time. The subjects then worked through the posted work-out schedule for the last fifteen--twenty minutes of class. During this time, the subjects in the treatment group individually reviewed their tapes with the instructor. The control group subjects received only a verbal review of their performance from the instructor. Each lesson included a review and/or practice of the breaststroke or some part of the stroke. Students were asked to refrain from practicing the stroke except in class. The Kilby glide test plus time was administered again during the seventh week. Gain scores were used to account for initial group differences. T-tests for significant differences between the means of the gain scores were used to determine if there were significant differences between groups. Rationale The purpose of this study was to investigate the effectiveness of videotape feedback in learning to swim the breaststroke. A two group design was chosen because it is the most direct method of comparing one method of instruction with the traditional method. The breaststroke is the best single stroke to use to catagorize swim students (Torney and Clayton, 67). While advanced swimmers demonstrate acceptable breaststroke form, students with a poor kick and poor arm pull form are gen- erally classified as beginning level swimmers. The breaststroke therefore is a skill taught primarily to intermediate swimmers and thus was the stroke used in the study. No optimal schedule of practice time has been established for learning a given task (Robb, 58). Taylor (65) found no significant differences between groups after five instructional periods when teaching the whip kick portion of the breaststroke. Plese (52) used a fourteen lesson plan in teaching a gymnastics routine to demonstrate the effectiveness of videotape replay. Morgan (45) used only three instructional/practice periods to teach beginners the butterfly arm pull with breathing. The learning of a skill involves not only acquiring a general picture of the sequential organization of the task--as provided during the instruction phase--but also temporal patterning, refinement of coordination, and eventually automatic execution of the skill (Robb, 58). Subjects need enough time to form and organize subroutines, to learn the correct neuromuscular patterns, and to practice the complete stroke as well as parts of the stroke, especially if they were only allowed to practice during class time. Seven weeks were planned for the study to allow for these phases of learning. This time included one week of preliminary class orientation; one week of testing, procedure orientation, and instruction in and practice of swimming mechanics; two weeks of instruction and practice devoted to the breaststroke alone; and, three weeks of instruction in other swim skills with brief reviews of the breaststroke. Since spring vacation occurred after the first seven weeks of the semester, the final testing was done before the break to avoid the influence of extraneous independent variables. The lesson procedures are based on the investigator's previous experience in teaching the breaststroke and in using videotape for instructional purposes. The procedures and situations are similar to regular class situations in order to make the conclusions relevant to a realistic class setting. For example, all Swim II classes are coed, therefore both sexes were included in the study and in each group. The control group had eight males, and thirteen females; the treatment group had eleven males, and fifteen females. All the subjects in a class are usually instructed in a similar manner, therefore, treatment was assigned to classes rather than to individuals within classes. Individual videotapep review has been found by the investigator to be more helpful and less time consuming than group review, therefore this method was chosen. 10 As noted previously, this stroke is not novel to intermediate swimmers and each class contains a range of ability in such a skill. To meet the objective of this study, a pre-test, post-test design was chosen. Most inter- mediate swimmers have had enough swimming to be aware of the general breaststroke motions and can attempt to imitate these with sufficient proficiency to swim 25 yards, therefore this pretest was used. Many Swim II subjects were not sufficiently efficient to swim much more than twenty-five yards. However, as the subject learned the skill, less extraneous motions were used and the skill became more efficient. A timed, 25-yard stroke count test was used as the pretest and post- test since it basically measured the efficiency of the stroke. A test which also considered errors in performance was chosen because such a factor increased the validity of the test. Students could swim twenty-five yards but use a scissors kick (or any improper kick) initially. This type of test penalized illegal kicks or pulls and therefore made the test more valid. The differences between the pretest and posttest scores indicated the amount of learning that took place during the study. Assumptions: The following assumptions were made: l. Skill learning could be measured by the performance test; 2. Subjects did not receive visual feedback of their performance by any means other than the treatment; 11 3. Students refrained from practicing the stroke outside of class and therefore learning was the result of class participation; 5 4. Training is specific and therefore activities other than the breaststroke performed outside of class did not affect learning of this skill; 5. All students had the same input to produce a mental image of the breaststroke. This image served as a reference pattern by which they made error corrections in their stroke; 6. The instructor gave unbiased direction and affective feedback. Delimitations The problem was delimited to students attending Calvin College, Grand Rapids, Michigan, who were classified as intermediate swimmers. Students enrolled in four Swim II classes, taught by the same instructor, comprised the sample. The study was limited to a seven week period. Class periods were forty minutes long and each of the classes met twice a week. Two pools were used--the deep water diving tank and the competitive pool which is 25-yards long and 4 feet in depth. The pools are located next to each other. Throughout the experiment, the following pieces of videotape equipment were used: Videotape recorder--Sony model CV-2200, serial number 6538, Monochrome video monitor (19 inch screen)—-Sony model CVM-lBOU, serial number 12 16604494, Monochrome video camera with viewfinder--Sony model CVC-ZlOOA, serial number 15871 and Sony model CVF-4, serial number 14231. The camera and viewfinder were located directly in front of an underwater viewing window, using a 12.5 mm, fl:1.4 lens Opened to an f-stOp of 5.6 and focused at an arbitrary 15 feet setting. The composite video signal generated in this arrangement was sent via a single cable to the videotape recorder located along with the 19 inch monitor on a balcony-like area close to the pool itself. Each sub- ject individually viewed the replay in black and white image on the television screen located about one foot above eye level (six feet). The videotape replay was run at the same speed at which the skill was performed. Limitations: The investigator recognized the following weaknesses in the study: 1. Human judgement was involved in determining perfor- mance errors. 2. Only students enrolled in these Swim II classes were able to participate. 3. Personality types and degree of motivation could not be controlled in the study. 4. The expertise of the television operator and the vulnerability of the equipment could have affected the results. 5. Class size and sex ratios within the classes could not be controlled. 13 6. The experimenter could not control mental practice or prevent subjects from observing the breaststroke performance of other students during class or out- side of class. Significance of the Study A study of this nature provides important information for teachers concerning methods. The study may also supply valuable guidelines for teacher preparation. Videotape equipment is expensive. Studying the effectiveness of videotaping in teaching skills will provide quatitative data to use for comparing the value of this method in terms of cost with its value in terms of production. In addition, the value of time spent in videotaping and reviewing as Opposed to that expended in actual practice is an important issue for instructors. Studying the effective- ness of videotaping in learning a sport such as swimming allows teachers to make knowledgeable decisions concerning the value of videotaping situations that involve motor learning. Instructors of intermediate level performers are often faced with the problem of breaking Old habits or changing misconceptions the student has concerning skills. For example, many intermediate swimmers use a scissors kick to prOpel themselves in the breaststroke. Using the Kilby test, these kicks are counted as errors. An indication of the effectiveness of videotaping in changing this habit may 14 be obtained by comparing pre-error and post-error scores. The value of this means of visual feedback in assisting in the relearning process could be important to instructors working with students at this ability level. Professional preparation programs may be encouraged to place more emphasis on this medium if the technique is beneficial. The use of videotape emphasizes the importance of knowledge of performance or the process of skill execution as Opposed to emphasizing the results of performance i.e. distance, time, etc. If videotape replay is helpful, instruc- tion in skill analysis with an emphasis on technique may play a more important role in teacher preparation, especially preparation for instruction of certain skills such as body manipulation tasks. The results of the study may also suggest important factors concerning the techniques of videotaping. For example, the importance of underwater viewing as Opposed to surface viewing may be a consideration. Other technical considerations might be the value of individualized or group review, the necessity of using a trained technician, how often review is necessary. Definitions Information feedback--know1edge of performance or results provided to the learner that gives cues and which if correctly translated guides the learner to correct responses toward a desired reference pattern (Bilodeau, 10). 15 Intermediate swimmer--a student who is nonfearful of water and who can propel through the water for a distance of twenty-five yards using a front stroke, twenty-five yards using a back stroke, survival float and tread water for one minute (Calvin College minimal swim test). Practice--when students perform the skill by swimming various times and/or distances. During this time the instructor will give individual correction concerning stroke performance. Videotape replay--the process of replaying the videotape taken of each student's performance. The replay is a visual review of the performance. Closed skills--movements that are relatively stereotyped and predictable, performed in a relatively stable environment (Robb, 59). Examples of closed skills include swimming, diving, gymnastics, and dance skills. Knowledge of performance--information concerning the process of executing a task--understanding how a skill is performed. Knowledge of results--information concerning the outcome of performance--how far, accurate, etc., the performance was i.e., knowledge of products of performance. Open skills-movements in which sterotyped patterns are limited and in which environmental requirements are unpre- dictable (Robb, 59). Examples of Open skills include soccer, dodgeball, and hockey skills. 16 Breaststroke--a swimming skill swum on the surface of the water with the shoulders horizontal to the surface. The arm- pull and leg kick are made underwater. A portion of the swimmer's head breaks the surface before the next stroke is started. The whip kick and short arm pull with the hands leading the pull going down and out will be used (DGWS Aquatic Guide, 1925-1975). Stroke count--the number of strokes it takes the subject to swim twenty-five yards from a stationary prone position. One stroke includes one arm pull and one leg kick. Gross motor skill--a skill in which large muscle groups are involved or large movements are made. Fine motor skill--a skill in which small or few'muscle groups are activitated and movements are small. Augmented feedback--information supplementary to that pro- vided by the task (Robb, 55). Intrinsic feedback--information inherent in the task (Robb, 55). External feedback--information provided by the sense organs concerning the environment outside the body (Robb, 55). Internal feedback--information provided by the sense organs concerning the internal environment of the body (Robb, 55). l7 Concurrent feedback--information present during task execu- tion (Drowatsky, 26). Terminal feedback--information available after the response is completed (Drowatsky, 26). Referencejpatterns--performance models or the learner's con- ception of the skill patterns (Bilodeau, 10). CHAPTER II RELATED LITERATURE The literature dealing with the concept of informa- tion feedback and the factors effecting it is voluminous. For example, the reader may refer to the writings of (Robb (57); Adams, Goetz, and Marshall (1); and Bell (6) for information concerning the arrival times of feedback; to Annett (4); Bernasconi (9); Bell (7); Berlin (8); Wills (69); Chansky (16); and Jordan (36) for information comparing types of feedback; and to Judd (37); Spencer (63); Lloyd (40); and Pierson and Rasch (51) for information concerning the value of feedback. This chapter is a review of the literature involving information feedback and its relationship to gross motor skill learning. Feedback will be identified and defined. Feedback via film or videotape replay will be examined. The factors of arrival time, level of learning, and nature of the task will be reviewed as they relate to visual supple- mentary feedback. 18 19 Feedback Bilodeau (10) has identified several general terms in the research literature which indicate the functions of feedback. These include information feedback, knowledge of results, reinforcement and reward. The terms information feedback and knowledge of results identify the nature of feedback whereas reinforcement and reward identify its uses. Knowledge of results emphasizes the subject's awareness of the outcome of the response. The outcome may be in terms of points scored, distance covered or target hits. Information feedback involves stimuli under the experimenter's control which provide the performer with information concerning dis- crepencies between the goal and the actual results (Bourne, 11). This is the type of feedback videotape replay and films provide. Definition of Feedback The information of a skill begins with some concep- tion of that skill (Lockhart, 42). The performer hypothe- sizes what the skill is like and how to perform it. Once the performer has an idea of what the skill is, an attempt may be made to perform it in accordance with this concept. The performance is then compared to the concept to determine if the skill was accurately completed. Feedback thus is the information used by the performer to compare the response to the reference skill (Robb, 55; Robb, 56; Bilodeau, 60). 20 Feedback can be described as a return of part of the output, or response, to the input; this may lead either to a revision of the response just made or to a confirmation of the response as being correct (Drowatsky, 26; pp. 85). Information feedback, therefore, helps direct or regulate the performance (Stallings, 64; Bilodeau, 10). Reference Patterns In order for the performer to use the information provided by replay to correct errors, a conception of the skill pattern is necessary. Several studies have dealt with the use of filmed models to provide the framework for skill patterns. Most of these investigations demonstrated that repeated model useage does not significantly improve perfor— mance (Nelson, 46; Berlin, 8; Priebe and Burton, 54; Caine, l3; Wyness, 71; Carre, 14; Brown and Messeremith, 12; Newfeld, 47; Lloyd, 41). There is limited evidence that subjects may benefit more from repeated viewing of the reference pattern after initial learning stages rather than during beginning stages of learning. Lockhart (43) used a film designed especially for use with beginning bowling classes. The treatment group viewed either parts of or the entire film throughout the instructional unit. Critical ratios and learning curves were computed and revealed that the rate of improvement in the movie group was more consistant than that of the control, especially after the third week. Nelson (46) employed 100p 21 films of male and female golfers filmed at various angles while using the 7-iron. Using gain scores based on points earned by closeness to a target, he found no statistically significant difference between experimental and control groups. He did note, however, that slow motion lOOp films seemed to favor the later stages of learning. Berlin (8) used the golf swing, soccer punt, fencing lunge, tennis serve, and lacrosse catch to compare the use of various teaching methods. She also concluded that after initial learning, subjects are helped by various techniques includ- ing film aids. With such meager supportive evidence, the value of repeatedly viewing a reference pattern might be questioned. Carre (l4) maintained that as long as the subject had a well developed idea of the skill, repeated viewing of an expert performing via film was not needed. Rather, once the subject had an idea of what was required, time was more efficiently used in trial-and-error attempts at performing the skill. Augumented feedback allows the subject to assess performance to see if it was what it was intended to be. DeBacy (22) studied the effect of videotape replay of golf on self-assessment. She used a two group design. The sub- jects were exposed to treatment after seven weeks of instruction. During the treatment session, the subjects from both groups came individually, read standard directions, warmed up, hit four balls, viewed four model performances, and rated their swings as poorer than, better than, or equal 22 to the model performance. The experimental group subjects then reviewed their own performance and the performance of the four models again. The control subjects observed the models' performance again but did not see a videotape of their performance. Both groups then reassessed their performance. Five judges also rated each subject using the same scale. The experimental group reduced their over-assessment by 37 percent while the control group reduced theirs only 5 percent. Viewing model and personal performance improved assessment to the extent that the experimental group's assessments were not significantly different from the judges' ratings. It should be noted that videotape replay was not used until after completion of the initial learning phase of seven weeks. Error Correction Augmented visual feedback eliminates much ambiguity in communication between instructor and student because both can review the student's performance at the same time (Cratty, 21). Reviewing play via films or television has helped the coach or teacher see what needed to be improved and has assisted the player in understanding what the teacher wanted (Speed, 62; Hughs, 34; Coffey, 18). Thus, films provide the individual with direct information by which personal performance can be assessed and errors in performance can be observed and corrected. 23 The role of films in effecting change in self assess- ment of throwing performance was shown by Clifton and Smith (17). Sixty college men and women volunteers were randomly placed in an experimental or a control group. During the first phase, the subjects performed fundamental skills such as walking, running, throwing, catching, and long jumping. They were then given a checklist to rate their performance as favorable or unfavorable. After four weeks with no additional practice, the experimental subjects viewed loop films of their performance while the control group performed the skills again. All then re-evaluated their performance. The authors con- cluded that viewing the motion picture of throwing performance resulted in positive changes in assessment scores. Videotape replay has also been effective in improv- ing balance performance from one day to the next without practice. A study by Wren (70) involved seventy-two second, fourth, and sixth grade children given task of standing on a platform, on signal moving the platform to a level position, and maintaining this position without assistance. After twelve trials on the first day, members of the experimental group viewed their videotaped performance. Both groups were given five trials on the second day. Results indicated a significant difference in performance between the two groups with the experimental group superior at each grade level. It was assumed that the subjects receiving videotape replay assessed their performance because of what they saw and changed their performance the next day. Although learning, 24 i.e., more permanent change in performance (Drowatsky, 27), was not involved in this study, it demonstrated that feedback affects change in performance. It is inferred that this change is due to the error corrective information provided by feedback. The fact that performers correct errors after seeing their performance was also demonstrated by Watkins (68). He used motion pictures as an aid in correcting baseball batting faults. He filmed at the beginning of the experiment, at the end of the third week, and at the end of the fifth week. The experimental subjects reviewed the most recent films without receiving corrections and then were reshown the films and given corrections. The experimental group thus received an additional three minutes of instruction while the control group did not see films or receive extra instruction. Instruction during regular practice was the same for both groups. The experimental group significantly decreased in number of batting faults (p = .01) as rated by a judge using the Vana batting criteria. The treatment group showed the greatest amount of improvement during the first three weeks. Since advanced skill level subjects were used in this study, the objective of the film review was to correct a skill already learned. Factors Affecting Feedback Robb (56) has identified three specific factors which affect the role of feedback: (a) the stage of learning, 25 (b) the task to be learned, and (c) the arrival time of feed- back. Some studies show that augmented feedback is more effective during the later stages of learning. The difference in level and speed of learning may be due to differences in tasks since feedback is task specific (Ammons, 3). Arrival time is important in that the longer the delay in feedback, the less effect the information has on error correction. The major means of providing the performer with visual, augmented feedback have been films (motion pictures, 100p films, graphcheck pictures) or videotape replay. Both types of media provide the learner with error correction information. However, videotape replay provides immediate feedback while films usually involve delayed feedback. Arrival Time Although films can provide an excellent means of augmenting information feedback, the film development pro- cess creates a time lag which results in delay of informa- tion feedback. Most studies in which films have been used to provide augmented visual feedback have failed to obtain significant results regardless of the type of task involved. For example, Wyness (71) failed to obtain significance in his study which involved learning to put a sixteen pound shot. Jones (35) used long and short badminton serves and obtained no significant difference in learning between the group reviewing films of personal performance and those not reviewing performance. Lloyd (41) obtained similar results 26 using tennis ground strokes. Priebe and Burton (54) and Brown and Messersmith (12) also found that film feedback did not significantly enhance the learning of body manipulation tasks. Pictures of models and/or self performance do not increase over-all learning more than teaching without such aids (Wyness, 71; Lloyd, 41). Wyness (71) divided his sub- jects into four groups to learn to put the shot: a control group, a treatment group Observing pictures of themselves twice a week, a treatment group reviewing an adept model twice a week, and a treatment group reviewing pictures of themselves once a week and of a model once a week. Analysis of covariance revealed no significant difference between groups. Lloyd (41) studied the stage of learning at which feedback and model comparisons were useful in learning ten- nis ground strokes. His study also involved four groups. Group I viewed loop films of personal performance (taken the previous week) and films of experts during the fifth week of the study. Group II viewed its performance and lOOps of experts during week eight. Group III viewed similar films during week eleven. Group IV was the control. Lloyd found significant differences in gain scores among the groups at the end of sixteen weeks. It was concluded that 81 w motion pictures imposed at various stages in learning did not affect ultimate learning although viewing during the middle stages of instruction provided a temporary stimulus to learning rate. This early learning advantage was not maintained. 27 Similar results were obtained by Jones (35) who tested the learning of two badminton serves and by Priebe and Burton (54) and Brown and Messersmith (12) with high jumping and tumbling skill learning; respectively. Subjec- tive and objective evaluations were Obtained at the conclu- sion of Jones' study, but the range of ability of the two groups was too great to indicate statistically significant gains in learning. However, the experimental group did make gains between the first and third test periods, while the controls did not. In both body manipulation studies the treatment group saw films of experts during the early part of the studies and reviewed films of themselves later in the study. Priebe and Burton compared the average high jumping heights of both groups and found that the experimental group outjumped the control after the third, fourth, and fifth week. However, the control group outjumped the experimental group after the final (thirteenth) week. They concluded that slow motion pictures were effective at first but did not make a difference as learning progressed over extended time periods. Brown and Messersmith used judges' ratings for assessing performance in their study. They questioned the use of films for beginning tumbling in view of the cost and the lack Of positive results Obtained. Instead, they sug- gested that class time might better be used for practice. The failure of studies utilizing films to obtain significant results was possibly due to the inherent delay of the information feedback. Studies in which this delay 28 was minimized have been more positive. For example, Hainfeld (31) found improvement in beginning swimming skills after providing information feedback the day after the filming occurred. Since the learner received this feedback before any additional instruction was given, the feedback was rela- tively immediate. A higher percentage of the beginning swimmers who saw their performance were able to jump into the pool, level off, and swim twenty-five yards than those not observing their performance. The filming occurred during day three and was replayed before lesson four. The subjects in the study were classified as non-swimmers but the investi- gator did not identify the criteria for that classification or note whether fear of water was considered, a factor which might affect the results. The study lasted only six days so it is doubtful how much learning could have occurred in such a short period of time. Immediate feedback can also be provided with polariod pictures. Thompson (66) found significant differences in groups using graph check films involving an eight picture sequence of beginning golf subjects. The pictures were taken once a week and reviewed by the experimental group three times a week. Three golf tests were administered over the term. Significant improvement (.01 level) attributable to the use of immediate feedback as provided by the graph check pictures was obtained. In many of the studies using films, the review of the film did not occur until several lessons or even weeks 29 had passed. Therefore, the delay was not only considerable but also disruptive to learning. Between the performance and the review, learning may have occurred or performance may have changed. Thus the delay resulted in an inaccurate representation of skill performance. For example, Brown and Messersmith (12) filmed during lesson nine but showed the films during lesson thirteen. Lloyd (41) showed the films the week following the filming. Film processing took three days in Jones' study (35). Subjects in her study were view- ing the same films from December 4 through January 14. These films were inaccurate during the latter part of that period. Most of the previous studies which failed to demon- strate the effectiveness of visual feedback had three factors in common: feedback was delayed, beginners served as subjects, and comparisons were made with filmed experts. Therefore, the information feedback provided by these studies may not only have been inaccurate, but also disturbing. Generally, those studies in which significant differences were found in favor of performers who viewed films did not use films of experts for performance comparisons. These studies also reduced delay considerably, therefore providing a more accurate, useful picture of current skill behavior. Videotape replay provides immediate information feedback and therefore eliminates the factor of arrival time affecting results. Immediacy of knowledge of performance is important to the learner (Cratty, 21). The longer the delay between performance and feedback, the less effective the 30 information (Ammons, 3). The learner needs sufficient time to process information feedback, however, too great of a delay will result in the correction being forgotten or losing its relevance (Stallings, 64). In recent years the use of television replay has increased. The immediacy which this media provides eliminates many of the delay problems noted previously, therefore the error corrective nature of augmented feedback can be effective. Level of Learning The level of learning might affect the results of augmented information feedback. Lloyd (41) suggested that students in the middle third skill level improved signifi- cantly due to viewing during the middle stages of instruction. Priebe and Burton (54) concluded that films helped in the early stages of learning experiments. Jones (35) showed that the experimental group learned significantly more during the early part of the study. Hainfeld (31) and Thompson (66) found significant learning for beginners in two types of skills. Green (30) concluded that the effectiveness of television replay on learning was dependent upon the ability of the group. Only one author suggested that age may ha a factor reSponsible for the success in using videotape to teach motor skills to children. Trainable retarded children served as Smith's (61) subjects in a study comparing tradi- tional teaching to teaching with videotape feedback. The skills included balance tasks, obstacle course tasks, and 31 ball handling skills. Gain scores for the age groups were compared as provided by the evaluations of a team of experts. The teenage group was the only group revealing significant improvement due to the information feedback provided by videotape. The results may reflect the idea that certain cognitive skills may be necessary before the error informa- tion can be integrated and used meaningfully. There is evidence that beginners benefit from augmented, visual information feedback. Priebe and Burton (54) demonstrated that slow motion pictures were valuable early in learning but not as learning progressed. Hainfeld (31) and Thompson (16) found significant results in film studies involving beginners as did Morgan (62), Paulat (49), Patterson (48), and Del Rey (23) using videotape with beginners. Penman, Bartz, and Davis (50) concluded that videotape was more effective with poorly skilled (remedial) students. Contrary to these conclusions that poorly skilled (beginning) students benefit more from replay, there is considerable evidence to support the idea that information feedback is more effective after the performer has gained some proficiency in a skill. Lawther (39) suggests that moving pictures and 100p films help the learner in form analysis at higher skill levels. He recommends that exten- sive use of the aids with beginners is questionable. While studying gymnastics skills, Plese (52) revealed that the separation in learning between the control and experimental 32 groups started with the inclusion of more advanced gymnastic skills in the teaching sequence. Green (30) concluded that advanced beginners profit more from videotape replay than did beginners in learning swimming skills. Morgan's (45) study suggests that beginning swimmers may not take advant- age of videotape learning. Carre's (14) study of the discus throw supported the idea that videotape may not be effective for beginners by noting that early learning is trial and error and the beginner cannot assimilate large amounts of information. Investigations using delayed feedback also suggest that subjects beyond initial learning stages benefit most from replay. Jones (35) obtained significant gains in the experimental group between the first and third test. Nelson (46) concluded that slow motion 100p films seemed to favor the learning of golf skills in the later stages but not in the early stages. Berlin (8) noted that the begin- ner seemed to need uninterrupted practice, and after initial learning, subjects Seemed to be aided by varied teaching techniques including movies. Finally, Watkins (68) used varsity players and found significant differences in correc- ting batting faults. The middle stage of learning may be more affected by visual augmented information feedback than later or earlier stages. Taylor (65) compared the groups according to high, middle, and low skill level in learning the whip kick in swimming. The factors he studied included videotape replay, 33 verbal replay, and combinations of each, as well as initial skill level. He concluded: (a) treatments were equally effective for subjects at the high skill level, (b) video- tape replay plus verbal feedback were more effective than no feedback for subjects at the middle skill level, and (c) videotape replay plus verbal feedback was superior to verbal feedback or no feedback for subjects at the low skill level. These conclusions are contrary to those of Penman, Bartz, and Davis (50) but similar to those suggested by Lloyd (41). From these studies, it appears that the evidence is not conclusive that level of skill alone affects learning as it relates to information feedback. Most of the studies cited were done with beginners. Both positive and negative results were obtained. Studies in which periodic testing revealed the amount of learning that took place in given time, generally indicated that beginners benefit more after they understand the skill. There is less evidence indicating that highly skilled players are helped by such feedback. Watkins (68) and Penman, Bartz, and Davis (50) indicated that varsity players or the best performers benefited from replay. Taylor (65), Priebe and Burton (54), and Lloyd (41) revealed contrary findings. In general, there seems to be indication that subjects in the middle levels of skill do significantly benefit from immediate or delayed information feedback. There does seem to be agreement that feedback is most effec- tive during the practice phases of learning a skill (Robb, 58). Exactly how early in practice has not been established. 34 Nature of the Task Feedback is task specific (Robb, 56). It is the teacher's task to augment information and to call the lear- ner's attention to it (Stallings, 64). The types of tasks studied by various authors can be categorized as open or closed (Poulton, 53). Such skills as swimming, dancing, and tumbling are closed skills. The most effective type of feed- back for closed skills (in which stereotyped form is important) is terminal augmented feedback as provided by performance replay (Robb, 59). Such skills as baseball batting and basketball guarding are open skills. A.major source of feedback for these skills is the environment (Robb, 59). Skills can also be classified as tasks involving object projection or body manipulation. Object projection skills provide knowledge of results in which the speed, dis- tance or accuracy of the object provide the learner with immediate, extrinsic feedback. Many of these skills are classified as Open skills. Body manipulation skills, on the other hand, provide little extrinsic feedback and depend primarily on pr0prioceptive information. These skills are closed skills requiring augmented information feedback. The performer cannot see the performance, and therefore is more likely to benefit from supplementary information which demonstrates performance. However, for this information to be valuable, immediate feedback would be most effective. 35 The results of studies employing delayed information feedback did not appear to be dependent on the type of task involved. For example, studies failing to show significant results involved both body manipulation and object projec- tion. Priebe and Burton (54), and Brown and Messersmith (12) used body manipulation skills, i.e., high jumping and tumbling. Jones (35), Wyness (71), and Lloyd (41) used object projection skills, i.e., badminton serves, shot put, and tennis ground strokes. Studies in which positive results were obtained also employed various types of skills. Hain- feld (31) used a swimming, a body manipulation skill. Watkins (68) and Thompson (66) used object projection skills, i.e., baseball batting and the golf swing.‘ Augmented information feedback as supplied by video- tape replay is especially critical in skills with little intrinsic feedback such as body manipulation skills. Where there is sufficient knowledge of results inherent in a task, additional knowledge of results does not affect further acquisition or retention of a skill at the beginning level (Bell, 6). The value of this type of information feedback in skills with little intrinsic feedback has been established reguardless of the means used to measure the dependent variables--i.e., despite whether the process or product was measured. For example, Carre (14) used form ratings Of the discus throw as well as distance measurements. He found that there is a significant correlationship between perfor- mance measures of form and distance. Yet, he failed to 36 demonstrate the effectiveness of videotape replay for object projection tasks. On the other hand, Taylor (65) using judges' ratings and Morgan (45) using a timed test both obtained significant results with videotape in learning swimming skills. Studies using Object Manipulation Most skills involving object manipulation provide inherent knowledge of results by means of the performer's ability to observe the accuracy or speed of the projectile. With the availability of such information, there is reason to question whether additional feedback is necessary or helpful. Most of the studies involving videotape of such skills failed to obtain significance. For example, Carre (14) obtained no significance using the discus throw, as did Gasson (28) with badminton volleys, Caine (13) with bowling skills, and Sanders (60) using volleyball wall volleys. Two authors using the object projection skills found significant differences between learning with video- tape replay information feedback and learning without it. Paulat (49) used the tennis forehand drive. His study compared videotape replay vs. no replay, loOp film models vs. no filmed models, and males vs. females. Using Hewitt's forehand drive test and ratings of two tennis pros, he obtained significance (p