r ‘ EFFECT (TE-.OVEREEARNTRR i _ on RETENTION OF’A' GROSS - I MOTOR gsmu PERFORMED av T MENTALLY RETARDED ADOLESCENTS i Thesisfor the Degree of M. A. MICHIGAN STATE UNIVERSITY ,.~.-.- JOAN MARY HENN 1975 J flflfilfi H ' ‘ T. .2 .‘ ‘A r - m ‘5 . \-'02.1 Viv“ . . . L.) I . . l..-‘-" l _ 3‘1: #«bLjfZL [‘4‘ (,_ _..f \ - H9 3 1‘ _ iGIN‘m” )W M g m. «5 EW- "mfimu’c aw 3" HOAB FWNS' 800K BINDERV INC. UBR-1 "Y B'NDERS (0p? . - nth-12.,” . wclmm Willi/ll!!!fill/ll/l/l/lUl/IlllHI!!!Ill/IU/llllllllI/I/lll L 3 1293 10477 4850 ABSTRACT EFFECT OF OVERLEARNING ON RETENTION OF A GROSS MOTOR SKILL PERFORMED BY MENTALLY RETARDED ADOLESCENTS By Joan Mary Henn The purpose of the present investigation was to determine the effect of 100%, 150% and 200% overlearning on retention of a gross motor skill performed by mentally retarded adolescents. Subjects who were part of the secondary education program at Beekman Center, Lansing, Michigan were randomly assigned to one of three stations and were taught to throw quoits using a side arm throwing pattern. Each subject received learning trials on Thursday (Cycle I) or Friday (Cycle II). They continued to throw until they reached a learning criterion of three consecutive successful trials. A trial was suc- cessful when the entire quoit remained in contact with the 5 foot in diameter target which was located 18 feet from the front line of the restraining box. Those subjects who reached the learning criterion within the maximum limit of 60 trials, were randomly assigned to one of four treatment groups: 0% (control), 100%, 150% or 200% over- learning. All subjects assigned to one of the experimental groups began receiving overlearning trials on Monday (Cycle 1) or Tuesday Joan Mary Henn (Cycle II). Each subject received a maximum of 50% overlearning (11 trials) each day until they reached their designated level of over- learning. Thus, the 100% treatment groups received 22 trials in two days, the 150% groups received 33 trials in three days and the 200% overlearning groups obtained 44 extra practice trials in four days. Relearning trials were administered after approximately five weeks (plus four or minus three days) of no practice. All sub- jects were randomly assigned to one of two stations where they continued to throw quoits until they reached the learning criterion of three hits in a row. Retention scores were calculated by sub- tracting the number of trials needed to relearn from the number of trials taken to originally learn the task. The four factors of sex, chronological age, 1.0. and etiology were examined to determine if possible dependence existed between them and the retention scores. All factors were analyzed using contingency tables with two levels of performance (above and below the treatment group mean of the retention scores) and with the appropriate number of categories for each designated factor. The chi-square analysis indicated that the retention scores were highly dependent upon I.Q. The remaining three factors failed to show significance. Since the 1.0. scores were in nonparametric form, a two-way ANOVA was an inappropriate method of analysis. Instead, 2 separate one-way ANOVAs were computed. The first analysis, comparing the four treatment group means, failed to detect significance at the .05 level. The second analysis attempted to take into account the effect of I.Q. Joan Mary Henn on the retention scores. Each treatment group was divided in two, placing those subjects with 1.0. scores above the overall mean 1.0. into one group and those subjects with 1.0. scores below the overall mean I.Q. into a second group. A one-way ANOVA was calculated to determine if a significant difference existed between the eight treatment group means based on I.Q. level. The analysis failed to detect significance. EFFECT OF OVERLEARNING ON RETENTION OF A GROSS MOTOR SKILL PERFORMED BY MENTALLY RETARDED ADOLESCENTS By Joan Mary Henn A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Physical Education 1975 ACKNOWLEDGMENTS I would like to express my gratitude to Dr. Claudia Knowles and Dr. John Haubenstricker for their invaluable assistance during the development of the proposal and writing of the thesis. Special thanks is also extended to Dr. William Heusner for the many hours of discussion and assistance with the research design and statistical analyses of the study. In addition, I would like to express my sincere apprecia- tion to Dr. David Fuller and the faculty, staff and students at Beekman Center for their patience and tremendous cooperation during data collection. Finally, I would like to thank those MSU students who assisted me during data collection, especially Sally Kirleis, Linda Chopp and Barbara Marshall, for their time and dedication. Without their help, the study could never have been completed. ii TABLE OF CONTENTS Page LIST OF TABLES ........................ v CHAPTER I. THE PROBLEM ...................... 1 Need for the Study ................. 2 Statement of the Problem .............. 3 Research Hypotheses ................. 3 Scope of the Study ................. 4 Limitations of the Study .............. 5 Definition of Terms ................. 6 II. REVIEW OF RELATED LITERATURE ............. 8 Learning and Retention Studies Using Normal Subjects. 8 Level of Proficiency ............... 9 Degree of Overlearning .............. 10 Learning and Retention Studies Using Retarded Subjects ..................... 12 Verbal Learning and Retention ........... 12 Motor Proficiency ................. 15 Learning and Retention .............. 16 Overlearning and Retention ....... . . . . . 17 Summary ....................... 18 III. RESEARCH DESIGN .................... 20 Methods ....................... 20 Subjects ..................... 20 Equipment ..................... 22 Skill Technique .................. 23 Pilot-Training Session .............. 24 Learning Criterion ................ 24 Instructors and Recorders ............. 25 Learning Trials .................. 26 Overlearning Trials ................ 27 Retention Measurement ............... 29 CHAPTER IV. RESULTS AND DISCUSSION Method of Calculating Retention Scores ....... Relationship Between Instructors, Learning and Relearning . . . . Relationship Between Four Variables and Retention . . Sex . . ..... Chronological Age . I. Q. ....... Etiology ..... Justification For Using One- -Way ANOVA ...... . . Analysis of Retention Scores ............ Discussion of Procedural Errors ........... Discussion of Results ................ V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS ....... Summary ....... Conclusions ..... APPENDICES A. DESCRIPTION OF THE AVAILABLE POPULATION ........ B. RULES FOR THE GAME OF QUOITS ............. C. DESCRIPTION OF THE INSTRUCTIONAL PROCEDURES FOR THE SKILL DEMONSTRATION . D. RESULTS OF PILOT-TRAINING SESSION ......... E. SCORE SHEET ...... F. RAW DATA ....... iv 56 6O 62 63 65 66 67 LIST OF TABLES Table . Page 4. 1 One-Way ANOVA: Effect of Instructors on Initial Learning Conducted on Thursday . . . . . . .......... . 33 4. 2 One-Way ANOVA: Effect of Instructors on Initial Learning Conducted on Friday .................. 34 4. 3 One-Way ANOVA: Effect of Instructors on Relearning Trials ......................... 34 4. 4 Grouping of Retention Scores Based on Treatment Group Mean (TGM) to Eliminate Effect of Overlearning Trials . 35 4. 5 Chi-Square Analysis: Effect of Sex on Retention Scores . 36 4. 6 Chi-Square Analysis: Effect of Chronological Age (CA) on Retention Scores .................. 37 4. 7 Chi-Square Analysis: Effect of 1.0. on Retention Scores . 37 4. 8 Cramer's Chi-Square Contingency Coefficient: Effect of Etiology on Retention Scores ............ 39 4. 9 One—Way ANOVA: Cycle I-—Homogeneity of Treatment Groups' Initial Learning Trials ............ 41 4.10 One-Way ANOVA: Cycle II--Homogeneity of Treatment Groups' Initial Learning Trials ............ 42 4.11 One-Way ANOVA: Homogeneity of Treatment Groups' Initial Learning Trials . . .............. 42 4.12 One-Way ANOVA of Four Treatment Groups .......... 44 4.13 Treatment Group Means (TGMs) of Retention Scores for Two Levels of 1.0. . ................. 44 4.14 One-Way ANOVA of Eight TGMs (Four Treatments x Two 1.0. Levels) .................. , . . . 45 Table Page 4.15 Comparison of Number of Learning Trials to Criterion to TGMs and Negative Retention Scores (RSs) ........ 46 4.16 Comparison of Four Treatment Groups ........... 49 A-l Description of Subjects Who Completed Learning Trials . . 56 A-2 Description of Subjects Who Reached Criterion ...... 57 A-3 Description of Subjects Who Did Not Reach Criterion . . . 58 A-4 Description of Subjects Who Completed Study ....... 59 0-1 Learning Trials to Criterion ............... 64 F-l Raw Data ......................... 66 vi CHAPTER I THE PROBLEM Mental retardation is a condition whereby an individual possesses below average intelligence. That is, his score on a standard intelligence test is at least one standard deviation below the general population mean. This subaverage intelligence level "originates during the developmental period and is associated with impairment in adaptive behavior" (Heber, 1961). Mentally retarded individuals learn at a slower rate than their peers whose intelligence quotients fall within the normal range. Research evidence suggests that the learning process of the mentally retarded is hindered because of a short term memory deficit (Ellis, 1963). According to Ellis (1963), the duration and intensity of an environmental stimulus is decreased due to the low integrity of the central nervous system in retardates. Long term retention, however, appears to be equivalent to that of normals (Belmont, 1966). Once the retarded have learned a task through many repetitions, they generally retain it as well as normal individuals. Long term retention of motor skills in individuals of normal intelligence has been found to be enhanced by overlearning or extra practice. Few studies have been undertaken, however, to determine an optimal level of overlearning. Krueger (1929) found 50% overlearning to be the most economical level for the retention of nonsense syllables over intervals of 2 to 28 days. In a second study using a finger maze skill, he found 100% overlearning to be the most efficient (Krueger, 1930). Melnick (1971) found 200% overlearning to be the most benefi- cial in the retention of a balancing skill on a stabilometer. The results of this latter study, however, may be questioned due to the presence of a warm up decrement in the absolute recall scores and to inadequate retention measures for the per cent of savings scores. There is limited information available concerning the value of extra practice for retarded subjects when performing a gross motor task. Studies by Chasey (1971) and Chasey and Knowles (1973) indi- cated that the experimental group, which was required to attain a higher level of proficiency than the control group, received the best retention scores after a minimum of one month of no practice. The learning criterion was increased for the overlearning group so that no optimal per cent of overlearning was determined. Need for the Study Educators are aware of the fact that individuals retain skills best when they are overlearned. The mentally retarded can gen- erally retain a skill as well as the normal population if they receive adequate practice. However, no study appears to have been conducted using mentally retarded subjects to determine if an optimum level of overlearning exists that guarantees long term retention of a motor skill. Therefore, physical educators have no guidelines by which to determine how long practice should continue after a skill has been learned to insure maximum retention without overextending economical practice limits. More specifically the following question may be raised: is there an amount of practice which will guarantee long term retention but, if exceeded, will fail to significantly enhance long term retention to warrant the additional time and energy? The ulti- mate goal of most physical education programs for the mentally retarded is to develop skill in recreational activities so that during their abundant free time they may be able to enjoy leisure activities. This requires that the retardates retain the learned skill over a long period of time. The degree of extra practice which is most economical toward insuring long term retention can be worthwhile information for persons attempting to develop functional programs in physical educa- tion. Statement of the Problem The purpose of this study was (a) to determine whether 100%, 150% and/or 200% overlearning of a gross motor skill would signifi- cantly improve long term retention of that skill; and, (b) to determine the per cent of overlearning which was most economical for the reten- tion of a recreational motor skill after five weeks of no practice. Research Hypotheses Studies indicate that the influence of overlearning upon retention, when significant, is positive in nature. Therefore, the research hypotheses for this study were: l. The three experimental groups receiving either 100%, 150% or 200% overlearning will obtain significantly higher reten- tion scores than the control group. 2. Each of the three experimental groups will differ signifi- cantly from each other in the following manner: a) 200% overlearning will have the highest retention scores. b) 150% overlearning will have greater retention scores than the 100% overlearning. Scope of the Study A modification of the recreational game called quoits was the gross motor skill presented to 68 trainable adolescents attending Beekman Center for the Mentally Impaired, Lansing, Michigan. Each subject was randomly assigned to one of two days used to initially learn the skill. Two cycles were run concurrently since all subjects could not receive the learning trials during the time span of one school day. Learning began on a Thursday for Cycle I and on a Friday for Cycle II. Each subject was randomly assigned to one of the three testing stations where he received instruction on a one-to-one basis. He stood within a 3 x 3 foot box drawn on the floor 18 feet from the midpoint of a target 5 feet in diameter. Each subject continued to throw quoits at the target until he reached the learning criterion of three consecutive successful trials or reached the limit of 60 throws. A trial was successful when the entire quoit remained in contact with the target. Each subject reaching the learning criterion within the 60 trial limit was randomly assigned to one of four treatment groups. The three experimental groups were defined as 100%, 150% and 200% overlearning, the control group was designated as 0% overlearning. Extra practice trials began on a Monday and continued through Thursday for Cycle I, while overlearning began on a Tuesday and continued through Friday for Cycle II. Each subject in the experimental groups reported each day for extra practice trials, receiving 50% over- learning per day, until he reached his designated level of overlearn- ing. Therefore, the 100% overlearning groups ended practice after two consecutive days, the 150% overlearning groups stopped after three consecutive days and the 200% overlearning groups discontinued practice after four consecutive days. Retention was measured using gain scores five weeks from the first day overlearning trials began. The data was analyzed using both nonparametric and para- metric statistics. To determine if dependence existed between the retention scores and either sex, chronological age, 1.0. or etiology, chi-square analyses were completed. The effect of overlearning on the retention scores was analyzed through a one-way ANOVA. Limitations of the Study The results of the investigation were limited in respect to the following: 1. The subjects were part of an available population of train- able retarded adolescents attending Beekman Center, Lansing, Michigan. They were exposed to a well constructed physical education program. Therefore, generalizations must be limited to those populations displaying the same or similar characteristics and having similar experiences as the subjects participating in the present study. 2. Generalizations regarding the per cent of overlearning which is most economical for long term retention must be limited to the recreational skill of quoits. Definition of Terms Three terms are defined so that it is clear as to their use in this study. They are: Per cent of overlearning.--The amount of extra practice taken after the learning criterion has been reached. It is performed during the initial learning stages in order to reinforce the learning that has occurred. Retention.--The degree to which a subject can perform to the original learning criterion after a designated period of no practice. Long term retention implies a minimum rest interval of one month. Retention score.--A gain score computed by subtracting the number of trials needed to relearn the task to criterion level (B) from the number of trials needed to initially reach the learning criterion (A), (RS=A-B). Four categories of etiology, as identified by Fuller (1971), were used to define the probable causes of retardation within the population used for this study. They are defined as follows: Genetic.--Mental retardation is present in other siblings and/or in one or both parents. Brain Injured.--Medical evidence indicates that the child has incurred brain damage. Mongolism.--Medical evidence indicates that the child has acquired the genetic defect where the chromosome count is 47 rather than 46. Unknown.--No known cause for retardation can be clearly diagnosed. CHAPTER II REVIEW OF RELATED LITERATURE The purpose of this study was to determine if an optimal level of overlearning exists which will enhance the retention of a recreational skill. After all subjects reached the learning criter- ion, they were randomly assigned to a control group or to one of three experimental groups. Those assigned to an overlearning group, received either 100%, 150% or 200% additional practice. Retention was measured after a five week rest interval. Little research is available involving the learning and retention of gross motor skills by mentally retarded subjects. Reported research involving the degree of overlearning which will enhance retention for the retarded is apparently nonexistent. How- ever, certain indicators of expected outcomes can be drawn from the literature. The intent of this chapter is to review research which deals with learning and retention of verbal and motor tasks, and to compare the abilities of retarded and nonretarded individuals. Learning and Retention Studies Using Normal Subjects Research clearly indicates that the level of proficiency attained in learning by individuals possessing normal intelligence has the greatest influence on retention. Studies using motor skills and dealing with the variable of proficiency level will be reviewed so that comparisons may be made which are pertinent to the retarded population. Level of Proficiency Research generally reveals that when total learning trials for subjects provide extra practice, the ability of the participants to retain the motor skill even after long periods of no practice is high. Studies using fine motor tasks found extremely high retention after intervals of l to 28 days (Jahnke and Duncan, 1956; Oxendine, 1965; Ryan, 1962). Bell (1950) also found excellent retention after one year of no practice on a pursuit rotor task. When high school girls were retested on a Bachman ladder climbing skill after lay-offs of either 10 minutes, 1 day, 1 week, 4 weeks or 13 weeks, it was found that retention was extremely high for all groups with no signifi- cant differences existing between any of the rest intervals (Meyers, 1967). Fleishman and Parker (1962) also obtained high retention scores after 24 months of no practice on a complex tracking task. The investigators concluded that the most important factor influencing retention is the level of proficiency attained during the learning phase. In two other studies involving motor tasks, the investigators varied the amount of original practice to determine the effect of level of proficiency on retention. Ammons, et. a1., (1958) conducted two experiments where college males were trained to perform a sequential task manipulating a series of controls. The first experiment involved 10 two degrees of learning consisting of 5 and 30 practice trials, respectively. The second experiment involved two degrees of training, 1 hour and 8 hours, respectively. Subjects relearned the task after rest intervals of 1 day, 1 month, 6 months, 1 year or 2 years. The results in both experiments indicated that the groups receiving less practice and longer no practice intervals required more trials to relearn the task than those subjects receiving greater practice and shorter no practice intervals. Jahnke (1958) found that the per- formance of high school students on a pursuit rotor task after 10 minutes, 1 day or 1 week of rest improved as the amount of practice increased. Practice schedules involved a lS-second work, 45-second rest cycle of 1 minute, 2.5 minutes, 5 minutes or 10 minutes duration. The investigator concluded that initial postrest performance is related to the amount of initial practice. Degree of Overlearning Several studies using normal subjects attempted to determine the degree of overlearning that would be most beneficial toward reten- tion after all subjects reached a learning criterion. All of the studies determined the number of overlearning trials by multiplying the proportion of extra practice an individual was to receive by the number of trials needed for him to originally reach the learning criterion. In a study involving the memorization of lists of mono- syllabic nouns, the investigator found that 50% overlearning was the most economical level when measuring retention after no practice intervals of 2, 4, 7, l4 and 28 days (Krueger, 1929). The results 11 differed from an earlier study conducted by Luh (1922) where 50% overlearning did not enhance retention. A second study conducted by Krueger (1930) required that the subjects learn a finger maze skill. He found that 100% overlearning was most economical even though 200% overlearning produced the highest average retention scores. Melnick (1971) investigated the effect of three levels of overlearning on the retention of stabilometer performance. All of the college male students practiced until they reached a learning criterion. The experimental groups then received an additional 50%, 100% or 200% overlearning trials. Retention was measured after one week and one month. The first retention trial was used to measure absolute recall and it was found that each of the overlearning groups had significantly higher retention scores than the control group with the 200% over- learning group proving to be significantly better than any of the other experimental groups. However, when per cent of savings scores were used to measure relative retention, only the 200% group was superior to the control group. The one week and one month retention groups did not differ significantly from each other on absolute recall or on per cent of savings scores. The investigator stated, however, that the effectiveness of the degree of overlearning may have been masked since the learning criterion of 28 seconds was set very close to the maximum possible score of 30 seconds. Therefore, the possible range of scores was somewhat restricted. Although the amount of initial practice was a positive influence on long term retention, studies conducted to determine the proportion of overlearning which is most economical for long term 12 retention are inconclusive. There does appear, however, to be a level of extra practice which, if exceeded, does not significantly enhance retention scores. Thus, practice beyond the optimal level is of little value when retention is considered. Learning and Retention Studies Using Retarded Subjects Extensive research has been conducted to compare retardates with the normal population in their ability to learn and retain. From these studies we have an indication of the modifications in the instructional program which must be made to accommodate those indi- viduals possessing below average intelligence. Although research is available concerning the level of motor proficiency of retardates as compared to normals, little information is available pertaining to the learning and retention of motor skills. Therefore, research involving the comparison of low 1.0. and normal subjects in their abilities to learn and retain verbal tasks will also be included in this review. Verbal Learning and Retention Valuable research has been conducted to discern what differ- ences exist between mentally retarded and normal individuals in the ability to learn verbal paired-associate tasks. When comparing the initial learning of retarded and nonretarded children, three studies found no significant differences when the task was simple and involved familiar objects (Cantor and Ryan, 1962; Eisman, 1958; Vergason, 1964). Two of the studies (Eisman, 1958; Vergason, 1964) used chronological 13 age as the basis for defining comparable experimental groups while the third study (Cantor and Ryan, 1962) used mental age. Ring and Palermo (1961) conducted a study in which two experimental groups of retarded subjects were matched to normals by either chronological age or mental age. Mentally retarded children of the same chronological age as normal subjects were significantly inferior in their ability to learn a paired-associate task. However, no significant difference existed between normal subjects and retardates who were of the same mental age. Other studies equating retarded and nonretarded experimental groups by chronological age also found that the normal children required significantly fewer trials to initially learn the assignment (Blue, 1963; Lance, 1965; Stinnett and Prehm, 1970). Research gener- ally indicates that retarded individuals learn at a slower pace than normals of similar chronological ages. This may be due to a deficit in short term memory demonstrated by retarded subjects. However, other factors, particularly the degree of familiarity and meaningful— ness of the task to the individuals, may play an important role in the number of trials needed to reach a learning criterion (Ellis, 1963). The results of comparative studies of retention vary. When mental age was used to define the retarded and nonretarded experimental groups, no significant difference was found in per cent of savings scores after one week and one month (Cantor and Ryan, 1962). When chronological age was used to match experimental groups, Lance (1965) also found no significant difference in per cent of savings scores. However, Vergason (1964) found that the normal subjects were superior to retardates when the paired-associate task was learned to a criterion T4 of one correct response, but no significant difference existed after one month between the two groups when the participants were required to reach a learning criterion of five correct responses. Retention scores are influenced by the amount of practice the subjects are allowed to initially learn the task. Stinnett and Prehm (1970) established six treatment groups. Retarded and non- retarded children of equal chronological ages were assigned to one of three levels of learning. All subjects in the Equal Amount Learned group (EAL) were required to complete three totally correct runs through a list of six stimulus—response items. The Modified Method of Adjusted Learning group (MMAL) had to make three consecutive cor- rect responses on each paired—associate before it was removed from the packet, while the Method of Adjusted Learning group (MAL) needed only to perform three correct responses before the item was removed from the stack. When 24-hour retention was measured through relearning trials, the nonretarded and retarded EAL groups were not significantly different nor were the nonretarded EAL and MMAL groups. All other groups were significant at the .05 level. The authors attributed the nonsignificance to overlearning. Both of the EAL groups and the non- retarded MMAL group received enough practice so that a ceiling effect was imposed by the learning method. The differences in results found between the six groups clearly indicated that retention is dependent upon the amount of initial practice. Although mentally retarded individuals appear to demonstrate a short term memory deficit, long term memory appears to be equal to that of the normal population. Belmont (1966) conducted an extensive 15 review of retention studies comparing the abilities of retarded and nonretarded subjects. All of the studies varied considerably in their design. According to Belmont, only the study conducted by Klausmeier, Feldhusen and Check (1959) equalized learning for all subjects so that the results could be considered accurate. The investigators had children of high, average and low 1.0. learn four tasks in arithmetic. The problems were graded to each child's ability level. When retention was measured after a six week rest interval, the results indicated that the retention scores were not significantly different between the three groups. Most research conducted on long term retention involving retarded subjects does indicate that once a task is learned, the individuals will retain it as well as normal subjects (Ellis, 1963). Motor Proficiency Several researchers have investigated the motor performance of mentally retarded individuals (DiStefano, Ellis and Sloan, 1958; Francis and Rarick, 1959; Howe, 1959; Kral, 1972; Malpass, 1964; Rarick, 1973; Sloan, 1951). Three generalizations are consistently upheld when retardates are compared to normal children of the same chronological age and sex. First, retarded children usually show inferior motor proficiency. Generally they are two to four years behind the national norms of normal children on tests of motor ability. Second, the retarded follow a similar, but slower, pattern of develop- ment to that of intellectually normal children. Third, there appears to be a low positive correlation between the retarded's 1.0. and 16 motor proficiency. This finding is similar to that found within the normal population. Several factors may cause the discrepency which exists between the normal and retarded populations in motor proficiency. Some retarded children may experience poor motor performance due to a delay in general physical growth and maturation. Dawson and Edwards (1965) found that boys of low 1.0. (50-84) were not signifi- cantly different in grip strength from boys of high intelligence (85-149) when they were matched according to race, chronological age, height and weight. However, significance was found in favor of the high intelligence group when height and weight were removed as matching variables. Another factor which may cause inferior motor ability in the retarded is the lack of opportunity to learn and practice motor activ- ities. They may also receive less appropriate reinforcers for learning than do children of normal intelligence. Ellis and Craig (1969) demon- strated that when a novel motor task is used in a study and the level of cognitive understanding is minimal, retarded subjects do not differ in their ability to perform when compared to normal children on the basis of either chronological or mental age. Learning and Retention Information concerning the learning and retention of motor tasks by the mentally retarded is limited. However, one study com- paring learning and retention of retarded and nonretarded children on a motor task was reported by Ellis, Pryer and Barnett (1960). The 17 investigators trained subjects of similar chronological ages to perform a pursuit rotor task. The normal children were superior to the retarded when comparing initial performance levels. However, reten- tion scores were equivalent when differences in the amount learned were taken into consideration. Overlearning and Retention Few studies have been conducted to determine the effect of overlearning on the retention of motor skills performed by mentally retarded subjects. Chasey (1971) had institutionalized retarded subjects learn a diagonal jump pattern to one of two levels of learn- ing, one correct trial or three consecutive correct trials. When retention was measured after a four week rest interval, the subjects receiving extra practice needed fewer trials to complete the learning task three times without error than the learning group needed to complete the task once without error. Chasey and Knowles (1973) used institutionalized males to measure retention after five weeks of no practice of an underhand throw for accuracy performed at a distance of 8 and 16 feet. Four points were scored if the center circle was hit by a bean bag; 3, 2 and 1 points from inside to out- side, respectively, were awarded for hitting the remaining three circles. Participants were randomly assigned to a learning or over- learning treatment group. All subjects reached the learning criterion which was a minimum of 10 points in three consecutive trials at 8 feet, and a minimum of 6 points in three consecutive trials at 16 feet. The overlearning group continued practice until the task at both distances 18 was completed three consecutive times. A significant difference in the number of trials to relearn the task favored the overlearning group at 16 feet only. Summar Retarded and nonretarded individuals demonstrate similarities when learning and retention are studied. Although the retarded learn at a slower pace than normals, once they have overlearned a task, long term retention does not appear to differ between the two popula- tions. The importance of the relationship between retention and initial learning is also supported when analyzing motor studies involving fast and slow learners of normal intelligence. Two studies allowed all subjects to receive the same amount of initial practice. When retention was measured between individuals reaching high and low performance levels, the high ability group was generally superior to the low ability group (Carron and Martenuik, 1970; Purdy and Lockhart, 1962). However, when fast and slow learners were given an equal opportunity to reach a learning criterion, no difference was found in retention scores (Baer, 1940; Fleishman and Parker, 1962; Melnick, Lersten and Lockhart, 1972; Ryan, 1965). If all subjects are allowed to reach the same level of proficiency, there appears to be no rela- tionship between speed of learning and retention within the normal population. Level of proficiency appears to be the variable which has the greatest influence on retention. Distribution of practice is one factor that may affect the speed at which one reaches a designated 19 proficiency level. Studies involving normal subjects practicing a motor skill consistently indicated that distributed practice groups obtained higher initial learning scores than the massed practice groups (Cook and Hilgard, 1949; Lewis and Lowe, 1956; Reynolds and Bilodeau, 1953; Singer, 1965). Bassin and Webster (1973) found similar results with retarded subjects. It appears that distributed practice sessions would be the most economical method to use during initial learning sessions. Both the retarded and nonretarded populations benefit when a task is overlearned. Retention is high even after long periods of no practice. However, research attempting to define the most eco- nomical level of overlearning is scant and inconclusive. CHAPTER III RESEARCH DESIGN The purpose of this study was to determine if 100%, 150% and/or 200% overlearning would significantly improve long term reten- tion of a recreational skill after a five week rest interval. If a significant main effect was found between the control and the three experimental groups, the degree of overlearning which was most eco- nomical, when compared to the retention scores, was to be determined. Methods The experimental study was intended to implement one feasible method of teaching the beginning skills needed to perform and retain the recreational activity of quoits. The heterogeneity of the subjects, the distribution of practice during learning and the method of individualized instruction was intended to exemplify the type of structure which may be found within a regular school setting. Subjects A total of 92 trainable mentally retarded students attending the secondary education program at Beekman Center, Lansing, Michigan were the potential subjects for the study. Of that group, nine 20 21 individuals were eliminated from random selection due to physical handicaps, personal injuries limiting physical activity or restric- tions preventing participation in research studies. Also, eight subjects were randomly selected to participate in a pilot-training session used to test the appropriateness of the learning criterion and target size, and to train the three instructors (see page 24). The remaining 75 subjects of the available population were selected to participate in the study. During the two days used to initially teach the subjects the gross motor skill, seven individuals were absent and four others were eliminated due to test administra- tion errors. 0f the 64 subjects who completed the training session, 20 were unable to reach the learning criterion. During the days used for overlearning and re-test trials, nine subjects were absent and one was eliminated due to testing errors. Thus, a total of 34 subjects completed the study. The population which completed the initial learning session was heterogeneous according to sex, age and 1.0. (see Appendix A). The males totaled approximately 53% of the population. The mean age of the subjects as of October, 1974 was 17 years. Their 1.0., as measured by the Stanford-Binet Intelligence test (n=39) or the Wechler Intelligence Scale for Children (n=24), had a range of 25 to 73 with a mean of 47.23. Only 53% of the subjects who participated in the initial learning session completed the study. Of that group approximately 65% were males and 35% were females. The age range of 13-0 to 22-7 was similar to that obtained from the total population completing the 22 initial learning session thus yielding an identical mean age of 17 years. The categories of etiology, as identified by Fuller (1971), indicated that 17.6% of those who completed the study were retarded as a result of genetic causes, 23.5% were classified as brain injured, 23.5% were diagnosed as mongoloid and 35.3% of the subjects had unknown causes. A separate range and mean 1.0. score for the 34 subjects completing the study was not computed since it was believed to be nearly identical to those subjects who completed the initial learning session. Equipment The recreational game of quoits was the gross motor activity used in this study. A quoit is a hard rubber ring, 6 inches in diam- eter with a center hole 3 inches in diameter. Its weight is 16 ounces and it is made with one side flat so that it settles flush to the ground. The top is rounded so that it is thicker around the inside edge than the outside. A green, blue and white patterned oil cloth material served as the target. It measured 5 feet in diameter with a solid red circle 6 inches in diameter marking the exact center of the target. It was taped securely to the floor so that the midpoint of the target was 18 feet from the front line of the restraining box (see page 23). Although a stake is used as the target in the regulation game (see Appendix B for rules and scoring), it was not used in this study. The 5 foot target, as determined in the pilot-training session (see page 24), was believed to be the appropriate size to indicate to the subjects 23 the importance of accuracy in performance while allowing the investigator to obtain a measure of their level of proficiency. Skill Technique The skill technique resembled the movement pattern which is generally used when throwing a frisbee. Each subject stood within a 3 x 3 foot box marked on the floor so that the dominant hand side of the individual's body was facing the target. The quoit was held in the dominant hand so that the fingers were clamped around the flat side of the quoit while the thumb was in opposition wrapped around the top side. Starting the quoit near the hip fartherest from the target, the subject swung the arm across the front of the body and extended the arm and hand toward the target. When released correctly, the quoit travelled so that the flat side was nearly parallel to the floor. Upon landing, it would slide a short distance on the tiled floor but would stop soon after it came in contact with the rough target face, provided it was not thrown too hard. The movement pattern used in the recreational game of quoits was believed to be a novel task for the available population used in this study. The game of quoits was not a popular activity for the residents of Michigan. Also, the physical education program at Beekman Center did not offer specific instruction to the students in an activity which used the movement pattern described above. 24 Pilot-Training Session Two weeks prior to the start of the study, a pilot and training session was held to test the feasibility of the target size and the learning criterion, and to train the three instructors. Eight subjects were randomly selected from the available population to participate in the session. A set of instructions for performance was given to each of the instructors so that a standard procedure would be followed (see Appendix C). After the investigator demon- strated the teaching technique using two subjects, each of the three instructors practiced with two subjects. During the pilot-training session, information was collected to determine what the diameter of the target should be in relation to a feasible learning criterion (see below). The target was identi— cal to the ones used in the actual study. It measured 5 feet in diameter and also contained two concentric circles drawn inside at 3 and 4 feet, respectively. Results showed that in order to use a learning criterion of three hits in a row, a 5 foot in diameter target was necessary (see Appendix D). LearningiCriterion The purpose of the learning criterion was to establish a minimum level of proficiency by which the majority of potential subjects would demonstrate consistency in the throwing pattern and accuracy in performance. The criterion was set at three consecutive successful trials. A trial was considered successful if the entire quoit settled on the target face. If only part of the quoit remained 25 in contact with the target, the trial was recorded as a miss. The number of trials needed to reach the stated criterion was recorded as the learning trials. No warm up throws were given. Thus, each throw executed by a subject was recorded and counted in the learning trials. Instructors and Recorders Three female graduate students attending Michigan State University volunteered their services to act as instructors during the learning, overlearning and relearning trials. The three instruc- tors were enrolled in a course dealing with physical education for the handicapped at the time the study was being conducted. They also had previous teaching experience. Seven female students at Michigan State University volun- teered as assistants to record results during the learning, over- learning and relearning sessions. Each recorder was shown a score sheet and given an explanation of the information needed to be recorded (see Appendix E). The recorder had a score sheet for each subject assigned to the station for that day. When a subject entered the classroom to begin instruction, the recorder informed the instruc- tor of the dominant hand of that subject and recorded the starting time. The recorder positioned herself behind the partition dividing two of the stations and opposite the target. When a quoit was thrown, the recorder retrieved the ring to clear the area for the next throw. She also recorded the result of that throw. If the entire quoit remained in contact with the target, a "/" was placed in the 26 appropriate box, otherwise an "X" was recorded. After all available quoits were thrown, the recorder returned the rings to the instructor so that very little time elapsed between each trial. Learning Trials A large classroom at Beekman Center was the site where the study was conducted. The room was partitioned into three sections by dividers so that three testing stations were in use simultaneously. Each station was approximately 8 feet wide and 30 feet long, allowing ample space. Each station contained a restraining box 3 x 3 feet marked on the floor with the front restraining line 18 feet from the midpoint of the target. There were four quoits placed at each station. One instructor and one recorder was randomly assigned to each of the three testing stations. Two cycles had to be conducted since all subjects could not receive learning trials in one school day due to a lack of time. Thus, all subjects were randomly assigned to one of two days to receive learning trials. Cycle I started on Thursday, October 24th and Cycle II began on Friday, October 25th. 0n the day of the learning trials, subjects were randomly assigned to a testing station at 15 minute intervals beginning at 9:30 a.m. and finishing by 2:00 p.m. each day. Each instructor worked with each subject individually until the subject reached the learning criterion. The instructional pro- cedures were the same as those used in the pilot-training session. Each subject continued to throw quoits until he reached the learning 27 criterion of three hits in a row or he reached 60 trials. A total of 20 subjects were unable to reach the learning criterion within the trial limit and were dropped from the study. The maximum limit of 60 trials was imposed to prevent the possibility of boredom and fatigue, and to allow the study to be conducted within the time schedule imposed by the school day. Overlearningrlrials All subjects who reached the learning criterion were randomly assigned to one of four treatment groups. The 0% overlearning groups or the control groups (n=5 Cycle I, n=6 Cycle II) did not receive any additional practice trials. The subjects assigned to the three experi- mental groups, 100% (n=5 Cycle I, n=6 Cycle II), 150% (n=6 Cycle I, n=5 Cycle II) and 200% (n=5 Cycle I, n=6 Cycle II) overlearning, con- tinued receiving instruction and practice trials until they reached their designated level of overlearning. The number of extra practice trials performed by each of the overlearning groups was calculated by using the mean number of trials needed for all subjects to reach the learning criterion. The mean learning score rounded to the nearest even number was 22. This number was multiplied by the per cent of overlearning. Thus, the subjects in the 100% overlearning groups each received 22 extra prac- tice trials, and the subjects in the 150% and 200% overlearning groups each received a total of 33 and 44 overlearning trials, respectively. There were two reasons why the number of overlearning trials was computed using the overall mean. First, as individuals continue 28 to practice a skill, the amount of intra-individual variability tends to decrease and scores obtained by the participants tend to regress toward the mean. Secondly, research involving motor skills of fast and slow learners of normal intelligence indicates that there is no significant difference in retention scores between those subjects needing few trials to reach a specified learning criterion compared to those individuals requiring considerably greater number of prac- tice trials, when all individuals are allowed to reach the same level of proficiency (Baer, 1940; Fleishman and Parker, 1962; Melnick, Lersten and Lockhart, 1972; Ryan, 1965). The overlearning trials began after a three day rest period, since school was not in session over the weekend, and were spaced over a four day period. Those subjects in Cycle 1, who received learning trials on Thursday, began overlearning on Monday, October 28th. Those subjects in Cycle II, who received learning trials on Friday, began overlearning on Tuesday, October 29th. Each day, the subjects received 50% overlearning or a total of 11 trials. Therefore, the 100% over- learning groups discontinued practice sessions after the subjects reached 22 extra practice trials in two days, the 150% overlearning groups stopped after three days and the 200% overlearning groups finished extra practice trials after four days. Basically, the same organizational procedures were followed as those used during the initial learning session. All subjects, regardless of the experimental group to which they were designated, were randomly assigned to a training station each time they reported for extra practice trials. Two procedures were modified slightly 29 from the previous testing situation. There were only two stations used during the overlearning sessions due to the reduced number of subjects. Also, subjects were randomly assigned to test order at 5 minute intervals and all extra practice sessions were completed by 11:00 a.m. each day. Retention Measurement The length of the rest interval between the last day of practice and the relearning trials was not identical for all of the treatment groups. Only one day for each cycle was used to retest the subjects after approximately five weeks (plus four or minus three days) of no practice. It was felt that certain variables, especially climate and special events conducted during the school day, may have created dependence due to the day of testing. When analyzing reten- tion curves of verbal and motor tasks of normal subjects, it was noted that the rate of forgetting was greatest during the first few days of no practice and the speed at which the curve descended decreased so that only minimal differences in forgetting could be seen during the fourth week of rest (Hovland, 1951; Luh, 1922). Therefore, it was believed that the variability in the length of the no practice interval would not have a significant effect on the retention scores. The relearning trials were scheduled for Monday, December 2nd for those subjects in Cycle I and for Tuesday, December 3rd for those in Cycle II. A snow storm which caused hazardous driving conditions kept 50% of the Cycle I subjects home from school on Monday. There- fore, they were tested the following day. 30 A time span of five weeks was used as the rest interval because the investigator was interested in measuring long term reten- tion, where a large degree of forgetting has already occurred. Therefore, results would indicate the degree to which a skill would be retained for continued use. According to Hovland (1951) and Luh (1922), retention curves show that most forgetting has occurred by the fourth week. A five week interval was used to prevent possible conflicts with special school functions during the two days used for the subjects to perform the relearning trials. Procedures were nearly identical to those followed during the overlearning trials. Two testing stations were used and each subject was randomly assigned to a station at 10 minute intervals. All relearning trials were completed by 11:00 a.m. each day. As in the initial learning session, all subjects had to continue throwing quoits until they reached the learning criterion of three hits in a row. The instructional procedures were the same as those used during learning, except no physical manipulation or specific verbal cues were given after the demonstration. Instead the instructors emphasized the importance of accuracy by encouraging each subject to aim at the target. CHAPTER IV RESULTS AND DISCUSSION The study was conducted to determine if 100%, 150% or 200% overlearning would significantly enhance long term retention. After receiving learning trials, all subjects were assigned to one of the four treatment groups. Each treatment group received either 0%, 100%, 150% or 200% extra practice trials and all subjects were retested five weeks later to determine the number of trials needed to relearn the gross motor task of quoits. Method of Calculating Retention Scores Originally, per cent of savings scores were to be used as the measure of retention. The method of calculating per cent of savings scores is shown in the following formula: .5: RS - A x 100 where RS (retention score) represents the per cent of savings score, A is the number of trials taken to originally reach the learning criterion and B indicates the number of trials needed to relearn the skill. The difference obtained between the learning and relearning trials (A-B) is divided by the number of learning trials (A) to take into account the speed at which the subjects initially learned the skill. However, since 15 of the 34 subjects who completed the study 31 32 took more trials to relearn than to learn the task, per cent of savings scores were inappropriate. When the difference between A and B yields a negative score, a disproportionate bias is created favoring those subjects who demonstrated poorer performance on the relearning than the learning trials. Therefore, gain scores were used in place of per cent of savings scores. Gain scores are calculated using the fol- lowing formula: RS = A-B Expressed in words, this formula demonstrates that the retention score (RS) is the difference between the number of trials needed to originally reach the learning criterion (A) and the number of trials taken to relearn the task to criterion level (B). Although gain scores do not create a bias, they still possess two major limitations. First, negative retention scores are obtained for those subjects who took fewer trials to learn than relearn the gross motor skill. Secondly, gain scores provide a less accurate measure of retention than per cent of savings scores since they do not take into account the speed at which the subjects initially learned the task. For example, a subject who took only 6 trials to initially reach the learning criterion of three hits in a row has less chance of demon- strating a significant gain when his relearning score is subtracted from his learning trials than a subject who took 26 trials to reach the learning criterion. Therefore, the validity of the retention scores in this study is probably low and thus reduces considerably the accuracy with which the results may be interpreted. 33 Relationship Between Instructors, Learning and Relearnigg The effect of three different instructors on the subjects' abilities to originally learn the gross motor task was checked for dependence by calculating a one-way ANOVA. Since two cycles were run within the study, the effect of different instructors on the learning scores was analyzed separately for each of the two days used to com- plete the learning trials. The results of the analysis for Thursday, learning day one, (Table 4.1) failed to detect significance between the learning scores of the subjects and the three instructors (F=.642, P>.05). Table 4.l--One-Way ANOVA: Effect of Instructors on Initial Learning Conducted on Thursday. Source SS df MS F P A 356.01 2 178.00 .642 >.05 Error 5546.86 20 277.34 Total 5902.87 22 Although the F statistic was higher for the second learning day, Friday (F=2.35), it was not significant at the .05 level (Table 4.2). Thus, it was assumed that there was no dependence between the three different instructors and the subjects' abilities to initially learn the skill. Since all subjects were randomly assigned to stations and instructors each day, it was assumed that no dependence developed during the overlearning and relearning trials. To re-affirm this 34 Table 4.2--One-Way ANOVA: Effect of Instructors on Initial Learning Conducted on Friday. Source SS df MS F P A 1095.24 2 547.62 2.35 >.05 Error 4200.57 18 233.36 Total 5295.81 20 assumption, the possible dependence of the three instructors and the relearning trials was also checked by calculating a one—way ANOVA. Only one analysis was computed combining both the Monday and Tuesday retest; days since 50% of the population was absent on Monday, the first day of relearning (Table 4.3). Results failed to show dependence (F=.637, P>.05). Table 4.3--One-Way ANOVA: Effect of Instructors on Relearning Trials. Source SS df MS F P A 410.01 2 205.00 .637 >.05 Error 10292.68 32 321.65 Total 10702.69 34 Relationship Between Four Variables and Retention The four factors of sex, chronological age, 1.0. and etiology were examined to determine if they significantly affected the retention scores. Possible dependence between retention scores and sex, chron- ological age and 1.0. was checked by using the NUCROS program 35 (chi-square contingency tables) of the CISSR routine (nonparametric statistical analysis) on the CDC 6500 computer at Michigan State University. However, a chi-square analysis was inappropriate to test for possible dependence between retention scores afid etiology since the expected n per cell was not equal to or greater than five. Therefore, a chi-square coefficient was calculated and conclusions regarding retention scores and etiology were drawn on a subjective basis. The use of a nonparametric method of analysis to test for dependence required that retention scores be grouped to eliminate the possible bias of different levels of overlearning. This was accom- plished by calculating a treatment group mean (TGM) for each of the four treatment groups. The subjects were then placed into one of two levels of performance, those whose performance was above the TGM or those whose performance was below the TGM (Table 4.4). Each of the four factors was then tested for dependence using a contingency table containing the two levels of performance and the appropriate number of categories for the designated factor. Table 4.4--Grouping of Retention Scores Based on Treatment Group Mean (TGM) to Eliminate Effect of Overlearning Trials. Treat- No. of $5 No. of Ss ment TGM Above TGM Below TGM Total 0% +5.8 4 6 10 100% +2.3 5 5 10 150% +3.7 3 4 7 200% +11.7 4 3 7 T0ta1: 16 18 34 36 §_e_x A total of 12 females and 22 males completed the study. The results of the chi-square analysis of the 2 x 2 contingency table (x2=.216) failed to show significance at the .05 level (Table 4.5). Therefore, dependence due to sex was not indicated. Table 4.5--Chi-Square Analysis: Effect of Sex on Retention Scores. Retention Sex 2 Scores Females Males Total X df P Above TGM 5 ll 16 .216 l >.05 Below TGM 7 ll 18 Total 12 22 34 Chronological Age The chronological age for each subject was calculated from day of birth to October, 1974 in years and months. When analyzing the frequency distribution of ages, it was noted that natural groupings occurred. Thus, chronological age (CA) was divided into three cate- gories: 13-0 to 16-9, l7-2 to 18-9 and 19-2 to 22-7. The chi-square statistic obtained in the analysis of the 2 x 3 contingency table (x2=.405), was not significant at the .05 level (Table 4.6). It was concluded that no dependence existed between CA and the retention scores. 37 Table 4.6--Chi-Square Analysis: Effect of Chronological Age (CA) on Retention Scores. CA Retention 13-0 to 17-2 to 19-2 to 2 Scores 16-9 18-9 22-7 Total X df P Above TGM 7 5 4 16 .405 2 >.05 Below TGM 6 7 5 18 Total 13 12 - 9 34 I.g. A mean 1.0. of 47.40 was obtained from those subjects who originally reached the learning criterion. The range included 1.0. scores from 25 to 70. All subjects were placed into one of two sub- groups depending on the relationship of their individual I.Q. score to the overall group mean. Subjects were classified as either above or below the mean 1.0. The statistical analysis of the 2 x 2 contingency table yielded a chi-square of 9.72. The results disclosed that the reten- tion scores were highly dependent upon I.Q. (P<.Ol). The data are found in Table 4.7. Table 4.7--Chi-Square Analysis: Effect of 1.0. on Retention Scores. Retention I'Q' 2 Scores Above Mean Below Mean Total x df P Above TGM l3 3 16 9.722 1 <.Ol Below TGM 5 13 18 Total 18 16 34 38 Although the relationship between 1.0. and motor performance yields a low but positive correlation for subjects within the normal intelligence range, the relationship tends to strengthen as the degree of mental retardation increases (Cratty, 1969; Kahn and Burdett, 1967). That is, level of intelligence does affect a retardate's ability to learn a novel motor task. However, little is known about the rela- tionship between intelligence and a retardate's ability to retain a motor skill. Chasey and Knowles (1973) found no significant inter- action effect between level of retardation (mild, moderate and severe) and post-test scores (retention) in an underhand throw for accuracy. Although the present study found a strong relationship between 1.0. and retention scores, results must be questioned due to the probable low validity of the retention scores and the inappropriateness of the learning criterion. Etiology There were four categories of etiology defined in the present study. They were brain injured, genetic, mongoloid and unknown. When the 34 subjects were grouped according to etiology and retention scores, it was noted that only four of the eight cells had expected cell frequencies equal to or greater than five (Table 4.8). Therefore, determination of a chi-square statistic and significance level was inappropriate. However, Cramer's contingency coefficient was calculated (C=.124). If the contingency coefficient is near or at zero, little or no dependence is indicated. The coefficient is not considered highly reliable, however, when the number of rows and 39 columns is small. Although the contingency coefficient in this case is close to zero, it is difficult to interpret since the contingency table is composed of only two rows and four columns. However, when comparing the obtained cell frequencies to the expected cell fre- quencies, there does not seem to be any definite indication of dependence. Also, since the four categories of etiology are dis- tributed randomly across the four treatment groups, it is believed that if any dependence did exist, it would have little, if any, effect on the final analysis. Table 4.8--Cramer's Chi-Square Contingency Coefficient: Effect of Etiology on Retention Scores. Etiology Retention Scores Brain Injured Genetic Mongoloid Unknown Total Observed Cell Frequencies Above TGM 6 3 2 5 16 Below TGM 2 3 6 7 18 Total 8 6 8 12 34 Expected Cell Frequencies Above TGM 3.8 2.8 3.8 5.6 16 Below TGM 4.2 3.2 4.2 6.4 18 Total 8 O 6 O 8.0 12 0 34 Cramer's Contingency Coefficient (C)=.124 40 Although there was no definite indication of dependency of etiology on the retention scores, it was interesting to note the variability of performance between each of the four categories of etiology. Both the genetic and unknown classifications had nearly 50% of their respective group above the treatment group mean (TGM). Those subjects defined as brain injured had 75% of its population above the TGM while the mongoloids only had 25% of its group above the TGM. Although there is evidence that reveals a relationship between etiology and motor proficiency, no research could be found that reported the effect of etiology on a retardate's ability to retain a motor skill. The literature generally indicates that mongoloids demonstrate poorer motor abilities than their peers of different etiological classifications (Cratty, 1967; Fuller, 1971; Knights, Hyman and Wozny, 1966). Research also has shown that brain injured retardates usually perform better than mongoloids, and that genetic (familial) groups generally perform better than brain injured retardates (Auxter, 1967; Fuller, 1971). It is not known whether a similar relationship or, indeed, if any relationship could be expected between etiology and motor skill retention. Since the size of the population in the current study was small and the validity of the retention scores must be considered low, no conclusions may be drawn concerning the relationship between etiology and retention. 41 Justification For Using One-Way ANOVA Several assumptions were checked to determine if a one-way ANOVA could be used to analyze the retention scores. These included treatment group homogeneity, variance equivalency, normality and independence. After the subjects received learning trials, they were randomly assigned to one of the four treatment groups so that all four groups were believed homogeneous according to the number of trials needed to initially reach the learning criterion. Since two cycles were run concurrently, a one-way ANOVA was calculated to check for homogeneity of the treatment groups for each cycle (Tables 4.9 and 4.10). The F values of .188 for Cycle I and .137 for Cycle II did not detect significance at the .05 level. Table 4.9-~0ne-Way ANOVA: Cycle I--Homogeneity of Treatment Groups' Initial Learning Trials. Source SS df MS . F P A 169.91 3 56.64 .188 >.05 Error 5732.96 19 301.73 Total 5902.87 22 Since 10 subjects were lost during the overlearning and relearning trials, a one-way ANOVA was calculated after completion of the study to determine if the four treatment groups could still be considered homogeneous according to the number of learning trials to 42 Table 4.lO--One-Way ANOVA: Cycle II--Homogeneity of Treatment Groups' Initial Learning Trials. Source SS df MS F P A 125.28 3 41.76 .137 >.05 Error 5170.53 17 304.15 Total 5295.81 20 criterion (Table 4.11). The two cycles were combined due to the reduced n and because no differences in performance were found between the two cycles for either the learning, overlearning or relearning trials. There was no significant difference found at the .05 level, thus, the four treatment groups were considered homogeneous according to original learning (F=.29l). Therefore, if a significant difference was found between the retention scores of the four treatment groups, it could be attributed to the overlearning trials. Table 4.ll--One-Way ANOVA: Homogeneity of Treatment Groups' Initial Learning Trials. Source SS df MS F P A 279.16 3 93.05 .291 >.05 Error 9599.07 30 319.97 Total 9878.24 33 The assumptions of normality and equality of variance were also examined due to the extreme variability of retention scores found in the treatment groups. The Kolmogorov-Smirnov Test of Normality was 43 calculated for the 150% overlearning group since it had the greatest variability (S.D.=27.34) and the two extreme retention scores (-26 and +54) (See Appendix F for Raw Data). The statistical analysis yielded a K of .1969 which was not significant at the .05 level. Since the 150% overlearning group was considered normally distributed, the same assumption was justified for the remaining three treatment groups. An Fmax Test was used to test for the equality of variances of the four treatment groups. The analysis yielded an Fmax of 6.49 (4, 6 df) which did not indicate significance at the .05 level. There- fore, it was assumed that the treatment group variances were homo- geneous and thus, a one-way ANOVA was justified. The assumption of independence was the only assumption not met. As previously stated, retention scores were found to be dependent upon 1.0. scores. Even though a one-way ANOVA was not an entirely acceptable method of analyzing the retention scores, no other method of analysis was available. A two-way ANOVA was inappropriate because the 1.0. scores were in nonparametric form, above and below the overall 1.0. mean. Exact 1.0. scores could not be obtained due to the con- fidentiality of the information. Analysis of Retention Scores The STAT program of the CDC 6500 computer at Michigan State University was used to analyze the data. Two separate UNEQl programs (one-way ANOVA with unequal treatment group frequencies) were used. Since the two levels of 1.0. were random throughout the treatment 44 groups, the first program compared the four treatment groups ignoring 1.0., the dependent factor. The results failed to show significance at the .05 level. Power was found to be .56, which was too low to detect significant differences between the treatment groups. Table 4.12--0ne-Way ANOVA of Four Treatment Groups. Source SS df MS F P A 395.82 3 131.94 .441 .725 Error 8968.56 30 298.95 Total 9364.38 33 The second program divided each treatment group in two, placing those subjects whose 1.0. was above the overall mean 1.0. into one group and those below the overall mean 1.0. into a second group. The treatment group means (TGMs) with the two levels of 1.0. taken into account may be found in Table 4.13. Table 4.13--Treatment Group Means (TGMs) of Retention Scores for Two Levels of 1.0. 55 Above Mean 1.0. 55 Below Mean 1.0. Treatment Group n TGM n TGM 0% 5 + 5.2 5 + 6.4 100% 5 + 9.2 5 - 4.6 150% 4 +18.0 3 -15.3 200% 4 +18.8 3 + 2.3 45 The results of the one-way analysis of the eight TGMs may be found in Table 4.14. The F value (1.96) failed to indicate sig- nificance (P=.099). Power was calculated to be less than .30 which was too low to assume that no treatment effect was present. Table 4.14-~One-Way ANOVA of Eight TGMs (Four Treatments x Two 1.0. Levels). Source SS df MS F P A 3242.30 7 463.19 1.967 .099 Error 6122.08 26 235.46 Total 9364.38 33 Discussion of Procedural Errors Several problems reduced the validity of the present study placing definite restrictions on the interpretation of the results. Most of the problems stemmed directly or indirectly from procedural errors. The major problem was concerned with the learning criterion. As previously stated, the purpose of the learning criterion of three hits in a row was to establish a level of learning where the majority of subjects demonstrated a minimum level of proficiency in performance. The learning criterion did not meet the stated purpose. A score of three hits in a row was too difficult for a majority of the subjects in this study to consistently attain. Also, the criterion did not take into account the variability in performance of the retarded popu- lation. Even though a subject reached the learning criterion once, it did not mean he could do it a second time. A total of 11 subjects 46 were able to initially reach the learning criterion in seven or less trials. They were, however, unable to repeat the outstanding per- formance during the relearning trials. A comparison was made between the number of trials taken to initially learn the task and the retention scores of each of the subjects (Table 4.15). Results indicated that those subjects requiring between 20 to 60 trials to initially reach the learning criterion tended to obtain retention scores above their TGM. Those subjects who received fewer than 14 learning trials generally received reten- tion scores below their respective TGM. It is also interesting to note that of the 15 subjects who received negative retention scores (RSs), 12 of them had fewer than fourteen learning trials. Results clearly indicated that the learning criterion was inappropriate. Table 4.15--Comparison of Number of Learning Trials to Criterion to TGMs and Negative Retention Scores (RSs). No. of Trials to No. of $5 No. of $5 No. of $5 With Reach Learning Criterion Above TGM Below TGM Negative RS Total 1-13 (fast learners) 2 14 12 16 20-26 (average learners) 5 l O 6 32-60 (slow learners) 9 3 3 12 Total l6 18 15 34 A more appropriate criterion would have been to consider 5 throws as one trial. A subject would obtain a successful trial when 2 out of the 5 throws remain on the target. The learning criterion might require that the subject obtain four successful trials before 47 learning is completed. Thus, the learning criterion would require a minimum of 20 throws while allowing variability in performance during each trial. A second problem stemmed from the equipment used to test the skill performance of the subjects. When the quoits landed flat on the tiled floor, they would generally slide. Depending on the amount of force imparted, they would slide for a few inches or several feet. Also, the quoits have a hole in the middle which allowed the subjects to wrap their fingers tightly around the rubber ring. This caused some subjects to release the quoit so that it had a wobbly flight pattern. If the quoit did not land flat, it would roll on its outside edge. Since the quoits did not remain where they initially made contact with the floor, they tended to increase variability in performance. Third, the small number of subjects who completed the study reduced its effectiveness. A total of 20 subjects were eliminated since they could not reach the learning criterion. Absences also reduced the size of the n considerably. Thus, conclusions were limited to 34 of the 75 subjects who were part of the available population. The fourth major problem was related to the method of sta- tistical analysis. First of all, gain scores rather than per cent of savings scores had to be used as the measure of retention because 15 of the 34 subjects did poorer on the relearning than the learning trials. Since gain scores do not indicate the speed at which the subjects originally learned the task, the validity of the retention measure is questionable, particularly since the range of scores was 48 high (3 to 60). Secondly, no adequate method of analysis was available to test for differences between the four treatment groups. Due to the procedural problems, it was highly unlikely that significant differ- ences existed between the treatment groups. If there was significance, it could not be found because the dependent factor, 1.0., could not be accounted for adequately through the method of analysis. Discussion of Results Although the statistical analyses failed to detect signifi- cance between the four treatment groups, it is believed that the errors committed in the procedures of the study tended to suppress the true results. Research consistently reveals that the retarded population can retain a skill when adequate practice is provided during the initial learning stages. Since a majority of the subjects who received between 1 and 13 learning trials (LTs) also had negative retention scores (RSs), it seemed worthwhile to observe the number of subjects in each treatment group who initially reached the learning criterion within 13 trials but performed worse on the relearning trials (Table 4.16). If a rela- tionship could be found between the number of subjects who received less than 14 LTs and negative RSs, and each treatment group's mean (TGM) and standard deviation (5.0.), then a subjective conclusion could be made as to the effect of the three levels of overlearning. It would be expected that as the number of subjects who had 13 or less LTs and negative RSs increased within each treatment group, the TGM would tend to decrease. 0n the other hand, if the amount of 49 overlearning had a direct positive effect on retention, it would be expected that the TGM would increase and the group variability would decrease as the number of extra practice trials increased. Table 4.16--Comparison of Four Treatment Groups. No. of $5 No. of $5 No. of 55 With Treat- With With l-13 LTs & ment n TGM S.D. Negative RS l-l3 LTs a Negative RS 0% 10 5.6 13.58 5 6 5 100% 10 2.3 15.39 5 5 3 150% 7 3.7 27.34 4 3 3 200% 7 11.7 10.73 1 2 1 The number of subjects who received no more than 13 LTs in each treatment group was not identical. Thus, the retention scores of the four treatment groups were not affected equally by the subjects designated as fast learners. Nor could any direct relationship be found between the TGMs and the number of subjects who received fewer than 13 LTs and had negative RSs. The 0% overlearning group had the largest per cent of its subjects with less than 14 LTs and negative RSs (50%), yet the 0% group had the second largest TGM (5.6). The 200% overlearning group had the largest TGM (11.7) and the smallest per cent of subjects with less than 14 LTs and negative RSs (14%). On the other hand, the 100% treatment group had the smallest TGM (2.3) but the second smallest per cent of subjects with less than 14 LTs and negative RSs (30%). Finally, the 150% overlearning group had the second smallest TGM (3.7) and the second largest per cent of its sub- jects with 13 or less LTs and negative RSs (43%). 50 When attempting to determine the effect of the three levels of overlearning on the TGMs and the 5.0.5, again no direct relation- ship could be observed. Both the 100% and 150% treatment groups had TGMs below and 5.0.5 above the control group (Table 4.16). Since the per cent of subjects with less than 14 LTs and negative RSs for both the 100% and 150% groups was less than that found in the 0% group, the results may indicate that the two levels of extra practice (100% and 150%) were not sufficient to positively effect the retention scores. However, the 200% group had the lowest 5.0. (10.73) and the highest TGM (11.7) of all of the treatment groups. It would appear that the level of extra practice (200%) tended to positively influence the treatment group's RSs. However, no concrete conclusion may be drawn since the 200% group also had the smallest per cent of its sub- jects with fewer than 14 LTs and negative RSs. CHAPTER V SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary The purpose of the present investigation was to determine the effect of 100%, 150% and 200% overlearning on retention of a gross motor skill performed by mentally retarded adolescents. Subjects who were part of the secondary education program at Beekman Center, Lansing, Michigan were randomly assigned to one of three stations and were taught to throw quoits using a side arm throwing pattern. Each subject received learning trials on Thursday (Cycle 1) or Friday (Cycle 11). They continued to throw until they reached a learning criterion of three consecutive successful trials. A trial was suc- cessful when the entire quoit remained in contact with the 5 foot in diameter target which was located 18 feet from the front line of the restraining box. Those subjects who reached the learning criterion within the maximum limit of 60 trials, were randomly assigned to one of four treatment groups: 0% (control), 100%, 150% or 200% over- learning. All subjects assigned to one of the experimental groups began receiving overlearning trials on Monday (Cycle 1) or Tuesday (Cycle 11). Each subject received a maximum of 50% overlearning (11 trials) each day until they reached their designated level of overlearning. Thus, 51 52 the 100% treatment groups received 22 trials in two days, the 150% groups received 33 trials in three days and the 200% overlearning groups obtained 44 extra practice trials in four days. Relearning trials were administered after approximately five weeks (plus four or minus three days) of no practice. All sub- jects were randomly assigned to one of two stations where they continued to throw quoits until they reached the learning criterion of three hits in a row. Retention scores were calculated by sub- tracting the number of trials needed to relearn from the number of trials taken to originally learn the task. The four factors of sex, chronological age, 1.0. and etiology were examined to determine if possible dependence existed between them and the retention scores. All factors were analyzed using contingency tables with two levels of performance (above and below the treatment group mean of the retention scores) and with the appropriate number of categories for each designated factor. The chi-square analysis indicated that the retention scores were highly dependent upon 1.0. The remaining three factors failed to show significance. Since the 1.0. scores were in nonparametric form, a two-way ANOVA was an inappropriate method of analysis. Instead, 2 separate one-way ANOVAs were computed. The first analysis, comparing the four treatment group means, failed to detect significance at the .05 level. The second analysis attempted to take into account the effect of 1.0. on the retention scores. Each treatment group was divided in two, placing those subjects with 1.0. scores above the overall mean 1.0. into one group and those subjects with 1.0. scores below the overall 53 mean 1.0. into a second group. A one-way ANOVA was calculated to determine if a significant difference existed between the eight treat- ment group means based on 1.0. level. The analysis failed to detect significance. Conclusions Several factors, particularly the inappropriateness of the learning criterion, the use of gain scores rather than per cent of savings scores as the measure of retention and the inability to adequately control for the dependent factor, 1.0., in the analysis of the retention scores, place definite restrictions on the value of the conclusions. However, within the limitations of the investiga- tion, the following conclusions were drawn: 1. Sex, chronological age and etiology failed to show sig- nificant influences on retention scores of the gross motor skill of quoits. 2. 1.0. was found to have a highly significant effect (P<.Ol) on the retention scores. 3. A one-way ANOVA calculated to determine the effect of four levels of overlearning on retention scores ignoring the dependency of 1.0. failed to show significance at the .05 level (P=.725). 4. When the effect of 1.0. was taken into account by running a one-way ANOVA comparing eight treatment group means (four treatments divided by above and below overall mean 1.0.), there failed to be any significance obtained (P=.099). 54 Recommendations Before attempting to reconstruct a similar study to deter- mine the effect of overlearning on retention of a gross motor skill, several problems must be solved. Some suggestions for further research include: 1. A study comparing the retention scores of mentally retarded subjects who take few trials to learn a task (fast learners) and of subjects who take considerably more trials to originally learn the task (slow learners) should be inves- tigated. Although this has been studied using subjects of normal intelligence, no study could be found which compared mental retardates. The method of calculating overlearning trials needs to be studied. Should overlearning be individualized, i.e., should each subject's overlearning bebased on the number of trials it took him to originally learn the task, or should it be equalized across all treatment groups using the mean number of trials it took all subjects to originally reach criterion? This question needs to be investigated for both the normal and the mentally retarded populations. The effect of 1.0., sex, chronological age and etiology on long term retention of a gross motor skill also needs con- tinued investigation. The number of subjects who completed the present study was small, thus, no definite conclusions could be drawn. 55 The effect of different levels of overlearning on the retention of a gross motor task needs further investiga- tion. A meaningful novel task and an appropriate learning criterion must be established in order to obtain meaningful results. APPENDICES APPENDIX A DESCRIPTION OF THE AVAILABLE POPULATION .emu: Aumsz cmgu_wco Low mqum mucmmw——mch LmFmgumz Anv Lo .mmuc ummh mucmmwppmpcm pmcmmuugoecmpm Amy xn nmczmmme mm .o.Ht¥ .eemp .eeeoeeo to we emg mm< cowum—aaoa .mmech mcwccemg umumpasou on: mgumwnzm mo cowpawcumwonlpi< wFQMH 56 57 .emu: Aomsz :mgupwsu Lee meum wocmmPPFmch cmFmgumz Any co .mch ummp mocmmwFpmch umcwmlucomcepm Aev An umgzmmme mm .c.Ht« .eemp .eeeoeeo to we emoa< cum: wanna 1:: nomcoz :chm -mcmw cam: magma a c R c .q a ««.o.H Acv Amopowuu *Amgpcosn.mgzv wo< op mmmwewmwou .aoa .cowgmpwgu vasommm on: muumnazm mo cowuaweommonuml< anMH .ewu: Aomsz :wgupwcu Low mpwum mucmmwppmucH prmguoz any co .mmh: ummh mocwmwppmch nocwmnugoecmpm Amy 55 cmgzmmoe mm .o.Hte .e~a_ .eeeoeeo to we emoa< new: macmm -c: tomcoz cwecm -mcmo cam: magma & c x : te.o.H Acv Amopowum tfimnucosu.mcxv mm< .aom —muop .qoa op comwcmnEou .aezem eeee_aeeo on: meeenezm co eoeeaeeumea--e-< ePeee APPENDIX B RULES FOR THE GAME OF QUOITS RULES FOR THE GAME OF QUOITS] For indoor play, the stakes, also called hobs, can be placed any required distance apart depending upon the length of the space available. For outdoors, the usual distance is 30 feet. When two are playing, each player has two of the matched quoits. The first player takes his position at one of the pins and pitches his quoits to the other pin. His opponent follows. The player who scores, has the first turn to pitch back to the other stake. Ringers . l. A pitched quoit which encircles the hob, called a ringer, shall count 3 points. 2. A double ringer shall count 6 points. 3. A ringer topped by an opponent's quoit shall count 6 points for the person pitching the last ringer. 4. A triple ringer scores 9 points for the person pitching the last ringer. 5. A ringer topped by two hobbers (both quoits must touch the hob) counts 7 points for the person pitching both of the hobbers. 6. A pitcher who has a ringer and a hobber but is then topped with a hobber by an opponent, scores three points for his ringer. Hobbers l. A pitched quoit resting on or touching the hob, called a hobber, shall count 2 points. 2. A hobber topped by another quoit shall count 2 points for the quoit nearest or resting on the hob. 3. In case A's quoit and B's quoit both rest on the hob, they shall both be removed and the nearest of the remaining two quoits counts as one point. 4. Two hobbers shall count 4 points providing both quoits touch the hob. Combination of Ringers and Hobbers -l O A ringer and a hobber shall count 5 points. 2. A ringer topped by a hobber pitched by an opponent shall count 3 points for the person pitching the ringer. 1Rules taken from the game manufactured by Hohwieler Rubber Co., Morrisville, Pennsylvania. 60 61 In case of a double ringer being topped by a hobber, 6 points shall be scored for the person scoring the second ringer. In case a ringer is topped by a ringer and then a hobber by an opponent, 8 points shall be scored for the person pitching the last ringer and hobber. Single Points When neither a ringer nor a hobber is made, the nearest quoit to the hob shall count 1 point. If two of A's or two of 8's quoits are nearest the hob, 2 points shall be scored for that player. In case of a tie between A's and B's quoits, they shall be removed and the remaining two measured with one point being awarded to the nearest. Should three quoits be in contact with the hob, two shall be con- sidered as ties, and the remaining quoit shall score 1 point for its owner. APPENDIX C DESCRIPTION OF THE INSTRUCTIONAL PROCEDURES FOR THE SKILL DEMONSTRATION DESCRIPTION OF THE INSTRUCTIONAL PROCEDURES FOR THE SKILL DEMONSTRATION Instructions to the Subjects —-l 0 "We are going to learn a new game today called ring toss. You stand here and watch me as I do it". (Place yourself inside the 3 x 3 foot box so that your side is to the target. Make sure you toss the ring with whichever hand is dominant for the sub ect who is watching you. Have the subject stand so he is facing you and can clearly see the throwing hand and arm). "You are going to stand in the box like this". (Give the subject a chance to notice the position of your feet so he can realize that you are standing within the box). "You are going to hold the ring like this. See how my fingers are under and my thumb is on top? See that target? You are going to toss the ring all the way to the target. Now watch me toss the ring all the way to the target". (Throw the ring. If it does not remain on the target, throw another one until it does remain on the target). "See how the ring landed on the target? Now you stand in the box". (Make sure the subject's side is to the target and his feet are within the box. Place yourself outside the box so you are facing him and he is facing you). (Hand him a ring and check his hand position making sure the fingers are on the bottom and the thumb is on top). "Now toss the ring all the way to the target". (Give positive comments--if he misses, say "good try, but see if we can hit the target this time"; really praise him if the quoit remains on the target). Suggestions 1. 2. Do not hand the subject a second ring until the recorder has cleared the target area. If the subject is having difficulty: a. demonstrate again and give the one cue that hopefully will cor- rect the error. b. you may manipulate so he gets a feeling for the correct arm motion. If a subject is not getting the ring to the target because he is not throwing it high enough or hard enough, you may just say, "throw it harder (or higher)"; or you may take a ring, toss it yourself to the target overemphasizing the action you want. Then have the subject perform it. You may demonstrate standing outside the box while facing the subject (mirror demonstration so you will be using the opposite hand from the subject), then have the subject perform it. 62 APPENDIX 0 RESULTS OF PILOT-TRAINING SESSION RESULTS OF PILOT—TRAINING SESSION Data was collected on the eight subjects participating in the pilot-training session to determine the appropriate target size in relation to the learning criterion which had been set at three con- secutive successful trials. A 5 foot in diameter target with two concentric circles drawn at 3 and 4 feet, respectively, was used to determine the appropriate target size. If the entire quoit landed within the 3 foot in diameter circle, a "3" was recorded on the sub- ject's score sheet. A "4" was recorded if any part of the quoit contacted the area greater than the 3 foot in diameter circle, but less than the 5 foot in diameter circle. When any part of the quoit landed outside the 4 foot in diameter circle but was still within the 5 foot in diameter circle, a "5" was recorded. All subjects were given 50 learning trials regardless of performance ability. The group was heterogeneous according to sex and age, and was believed representative of the available population. By observing the results of the throws of each subject over the 50 trials, it was decided that the learning criterion of three consecu- tive successful trials was feasible provided the 5 foot in diameter target was used. Since all the subjects completed the 50 trials within a time span of 12 minutes, it was also decided that the maximum number of learning trials allowed each subject would be increased to 60. 63 64 The number of trials needed for each subject to reach the learning criterion of three consecutive successful trials is shown below. Table D-l--Learning Trials to Criterion. Subject No. Sex Age Number of Trials to Criterion 1 F 16-1 did not reach criterion; had 5 successful trials. 2 F 20-7 15 3 M 16-7 47 4 F 15-0 34 5 M 16-0 6 M 14-3 5 7 M 17-3 14 8 M ]4_0 did not reach criterion; had 7 successful trials. APPENDIX E SCORE SHEET SCORE SHEET NAME: DOM. HAND: SEX; AGE: 1.0.: ETIOLOGY: L R M F ************************************************* LEARNING INSTRUCTOR: STATION #: TEST ORDER #: STARTING TIME: (7Ehit; X=miss) DATE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ***************'1"k******************************** OVERLEARNING: % (JEhit; X=miss) 0A1; INSTRUCTOR STATION# ORDER# TI 1 2 3 4 5 5 7 8 9 1011 ************************************************* RELEARNING INSTRUCTOR: STATION #: TEST ORDER #: STARTING TIME: (7%hit; X=miss) DATE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25' 65 APPENDIX F RAW DATA Table F-l--Raw Data. RAW DATA ‘Treat- Learn Relearn S No. Sex Age 1.0. Etiology ment Ts Ts RS 1 M 17-10 Above Mean Unknown 0% 24 14 +10 2 M 18-8 Above Mean Genetic 0% 6 5 + l 3 M 13-1 Below Mean Unknown 0% 35 24 + 9 4 M 16-0 Above Mean Mongoloid 0% 32 11 +21 5 M 17-2 Below Mean Genetic 0% 5 8 - 3 6 F 20-1 Above Mean Brain Injured 0% 4 9 - 5 7 F 22-7 Above Mean Mongoloid 0% 5 6 - l 8 M 16-9 Below Mean Unknown 0% 12 20 - 8 9 M 14-9 Below Mean Mongoloid 0% 3 6 - 3 10 M 18-4 Below Mean Genetic 0% 58 23 +35 11 F 17-10 Above Mean Brain Injured 100% 7 26 -19 12 F 22-1 Below Mean Mongoloid 100% 59 7O -11 13 F 13-0 Above Mean Unknown 100% ll 5 + 6 14 M 14-9 Above Mean Brain Injured 100% 23 5 +18 15 M 19-2 Above Mean Brain Injured 100% 37 8 +29 16 F 15-3 Above Mean Genetic 100% 34 22 +12 17 F 19-2 Below Mean Genetic 100% 12 4 + 8 18 M 16-7 Below Mean Unknown 100% 13 15 - 2 19 F 18-1 Below Mean Unknown 100% 45 46 - l 20 M 21-8 Below Mean Unknown 100% 6 23 -17 21 M 18-9 Below Mean Unknown 150% 12 38 -26 22 M 18-2 Below Mean Mongoloid 150% 3 8 - 5 23 F 18-2 Above Mean Brain Injured 150% 60 6 +54 24 F 13-9 Above Mean Mongoloid 150% 34 48 -14 25 M 15-9 Below Mean Genetic 150% 4 l9 -15 26 F 18-8 Above Mean Brain Injured 150% 40 28 +12 27 M 21-2 Above Mean Unknown 150% 25 5 +20 28 M 15-7 Above Mean Unknown 200% 36 18 +18 29 M 16-5 Below Mean Unknown 200% 5 10 - 5 30 F 17-4 Below Mean Unknown 200% 20 9 +11 31 M 18-0 Above Mean Brain Injured 200% 25 9 +16 32 M 20-2 Above Mean Brain Injured 200% 26 12 +14 33 M 22-5 Below Mean Mongoloid 200% 6 5 + l 34 M 15-8 Above Mean Mongoloid 200% 35 8 +27 66 LIST OF REFERENCES LIST OF REFERENCES Ammons, R. B., Farr, R. G., Bloch, E., Neumann, E., Dey, M., Marion, R., & Ammons, C. H. Long-term retention of perceptual- motor skills. Journal of Experimental Psychology, 1958, 55, 318-328. Auxter, 0. Strength and flexibility of differentially diagnosed educable mentally retarded boys. Research Quarterly, 1967, 37, 455-461. Baer, R. A. The relationship between rate of learning and retention of several motor activities. Unpublished doctoral disserta- tion. Johns Hopkins University, 1940. Bassin, S. J. & Webster, M. Distribution of practice as a factor in mental retardates acquisition of a motor skill. Abstracts of Research Papers 1973 AAHPER Convention, 35. (Abstract) Bell, H.1L Retention oprursuit rotor skill after one year. Journal of Experimental Psychology, 1950, 40, 648-649. Belmont, J. M. ,Long-term memory in mental retardation. In N. R. Ellis (Ed.), International review of research in mental retardation. New York: Academic Press, 1966, pp. 219-255. Blue, C. M. Performance of normal and retarded subjects on a modified paired-associate task. American Journal of Mental Deficiency, 1963, 60, 228-234. Cantor, G. N. & Ryan, T. J. Retention of verbal paired-associates in normals and retardates. American Journal of Mental Deficiency, 1962, 66, 861-865. Carron, A. V. & Marteniuk, R. G. Retention of a balance skill as a function of initial ability level. Research Quarterly, 1970, 41, 478-483. I Chasey, W. C. Overlearning as a variable in the retention of gross motor skills by the mentally retarded. Research Quarterly, 1971, 42, 145-149. Chasey, W. C. & Knowles, C. J. Effects of overlearning on retention and relearning of gross-motor skill by mentally retarded males. Perceptual and Motor Skills, 1973, 36, 503-509. 67 68 Cook, B. S. & Hilgard, E. R. Distributed practice in motor learning: progressively increasing and decreasing rests. Journal of Experimental Psychology, 1949, 39, 169-172. Cratty, B. J. Developmental sequences of perceptual-motor tasks. Freeport, New York: Educational Activities, Inc.;_l967. Dawson, S. W. & Edwards, R. W. Motor development of retarded children. Perceptual and Motor Skills, 1965, 21, 223-226. DiStefano, M. K., Ellis, N. R. & Sloan, W. Motor proficiency in mental defectives. Perceptual and Motor Skills, 1958, 8, 231-234. Eisman, B. S. Paired associate learning, generalization and reten- tion as a function of intelligence. American Journal of Mental Deficiency, 1958, 63, 481-489. Ellis, M. J. & Craig, T. T. A note on the inferiority of retardates' motor performance. Journal of Motor Behavior, 1969, 1, 341-346. Ellis, N. R. The stimulus trace and behavioral inadequacy. In N. R. Ellis (Ed.), Handbook of mental deficiency. New York: McGraw-Hill, 1963, pp. 134-158. Ellis, N. R., Pryer, M. W. & Barnett, C. 0. Motor learning and retention in normals and defectives. Perceptual and Motor Skills, 1960, 10, 83-91. Fleishman, E. A. & Parker, J. F. Factors in the retention and relearning of perceptual motor skill. Journal of Experi- mental Psychology, 1962, 64, 215-226. Francis, R. J. & Rarick, G. L. Motor characteristics of the mentally retarded. American Journal of Mental Deficiency, 1959, 63, 792-811. Fuller, 0. A. Evaluating motor performance of trainable mentally handicapped boys and girls, ages eight through twenty-one. Unpublished doctoral dissertation. Michigan State Univer- sity, 1971. Heber, R. Modifications in the manual on terminology and classifica- tion in mental retardation. American Journal of Mental Deficiency, 1961, 65, 499-501. Hovland, C. L. Human learning and retention. In S. S. Stevens (Ed.), Handbook of experimental psychology. New York: J. Wiley & Sons, Inc., 1951, pp. 613-689. ‘J 69 Howe, C. E. A comparison of motor skills of mentally retarded and normal children. Exceptional Children, 1959, 25, 352-354. Jahnke, J. C. & Duncan, C. P. Reminiscence and forgetting in motor learning after extended rest intervals. Journal of Experi- mental Psychology, 1956, 52, 273-282. Jahnke, J. C. Retention in motor learning as a function of amount of practice and rest. Journal of Experimental Psychology, 1958, 55, 270-275. Kahn, H. & Burdett, A. Interaction of practice and rewards on motor performance of adolescent mental retardates. American Journal of Mental Deficiency, 1967, 72, 422-427. Klausmeier, H. J., Feldhusen, J. & Check, J. An analysis of learning efficiencygin arithmetic of mentally retarded children in comparison with children of average and high intelligence. Madison: University of Wisconsin Press, 1959. Knights, R., Hyman, J. & Wozny, M. Psychomotor abilities of familial, brain-injured and mongoloid retarded children. American Journal of Mental Deficiency, 1966, 70, 454-457. Kral, P. Motor characteristics and development of retarded children: success experience. Education and Training of the Mentally Retarded, 1972, 7, 14-19. Krueger, W. C. The effect of overlearning on retention. Journal of Experimental Psychology, 1929, 12, 71-78. Krueger, W. C. Further studies in overlearning. Journal of Experi- mental Psychology, 1930, 13, 152-163. Lance, W. 0. Effects of meaningfulness and overlearning on retention in normal and retarded adolescents. American Journal of Mental Deficiengy, 1965, 70, 270-275. Lewis, 0. & Lowe, W. F. Retention of skill on the SAM complex coordinator. Preceedings of the Iowa Academy of Science, 1956, 63, 591-599. Luh, C. W. The conditions of retention. Psychological Monographs, 1922, 31, (3, Whole No. 142). Malpass, L. F. Motor skills in mental deficiency. In N. R. Ellis (Ed.), Handbook of mental deficiency. New York: McGraw- Hill, 1963, pp. 602-631. Melnick, M. J. Effects of overlearning on the retention of a gross motor skill. Research Quarterly, 1971, 42, 60-69. 70 Melnick, M. J., Lersten, K. C., & Lockhart, A. S. Retention of fast and slow learners following overlearning of gross motor skill. Journal of Motor Behavior, 1972, 4, 187-193. Meyers, J. Retention of balance coordination learning as influenced by extended lay-offs. Research Quarterly, 1967, 38, 72-78. Oxendine, J. 8. Effects of progressively changing practice schedules on learning of a motor skill. Research Quarterly, 1965, 36, 307-315. Purdy, B. J. & Lockhart, A. S. Retention and relearning of gross motor skills after long periods of no practice. Research Quarterly, 1962, 33, 265-272. Rarick, G. L. Motor performance of mentally retarded children. In G. L. Rarick (Ed.), Physical activity: humanygrowth and development. New York: Academic Press, 1973, pp. 225-256. Reynolds, B. & Bilodeau, I. M. Acquisition and retention of three psycho-motor tests as a function of distribution of practice during acquisition. Journal of Experimental Psychology, 1953, 44, 19-26. Ring, E. M. & Palermo, E. S. Paired associate learning of retarded and normal children. American Journal of Mental Deficiency, 1961, 66, 100-107. Ryan, E.[L Retention of stabilometer and pursuit rotor skills. Research Quarterly, 1962, 33, 592-598. Ryan, E.[L Retention of stabilometer performance over extended periods of time. Research Quarterly, 1965, 36, 46-51. Singer, R. N. Massed and distributed practice effects on the acquisition and retention of a novel basketball skill. Research Quarterly, 1965, 36, 68-77. Sloan, W. Motor proficiency and intelligence. American Journal of Mental Deficiency, 1951, 55, 394-406. Stinnett, R. D. & Prehm, H. J. Retention in retarded and non- retarded children as a function of learning method. American Journal of Mental Deficiency, 1970, 75, 39-46. Vergason, G. A. Retention in retarded and normal subjects as a function of amount of original training. American Journal of Mental Deficiency, 1964, 68, 623-629. Typed and Printed in the U.S.A. Professional Thesis Preparation Cliff and Paula Haughey 144 Maplewood Drive ,: ,- ."’ East Lansing, Michigan 48823 - 3" Telephone (517) 337-1527 "1111111111'1111'11111 I