Perceptualand Motor Skills, 1992, 75, 555-562. O Perceptual and Motor Skills 1992
EFFECTS O F CONTEXTUAL INTERFERENCE O N LEARNING TECHNICAL SPORTS SKILLS LAURA BORTOLI, CLAUD10 ROBAZZA, VALTER DURIGON, CARLO CARRA Physical Education Institute of Padova Summary.-The effects of contextual interference on learning s k i h of volleyball (volley, bump, serve) are influenced by the scheduling of actual practice sessions: the activities can be proposed in a repetitive practice schedule (blocked practice) by continuously repeating the same task (low interference) or in random practice schedules by performing more tasks or variations of one same activity (high interference). High contextual interference, even though causing immediate limited performance, leads to superior performance on retention and transfer tests. Four experimental groups (13 students each) were placed in conditions of random, blocked, serial, and serial with high interference practice for 8 meetings (2 tests and 6 practice). Analysis yielded s i g d i cant differences among the groups on a transfer test (long transfer) for the serve, so results in this instructional setting are partially in line with those generally found in laboratory experiments.
The concept of contextual interference was introduced by Battig (1966, 1979) to show the aspects of interference that occur when one practises a certain task within a specific context: high interference is observed if in the same session different tasks are performed, even if they are similar or related and low interference when one practises, instead, just one task. From studies of verbal material learning, Battig noticed that high contextual interference situations, generally considered a cause of forgetfulness, led to better performance on retention and transfer tests than a situation of low interference. Contextual interference was considered an instrument that permitted overcoming some of the limits of mnestic performance: according to the principle of specificity encoding, for example, to improve memory performance it is necessary to reinstate the original context in which the learning took place. The diversity of situations in learning would render, on the other hand, the subject less dependent on the context and would stimulate him in multiple and variable elaborative activities to be able to produce better retrieval of information and a greater adaptation to the different contextual situations. Shea and Morgan (1979) were the first to verify Battig's predictions for motor skills learning, showing the superiority of random practice schedules over blocked practice schedules in facihtation of retention and transfer. Many other studies have led to similar results for a variety of laboratory tasks (Del Rey, 1982, 1989; Del Rey, Whitehurst, & Wood, 1983; Gabriele, Hall, & 'This research was financed by CONI, Divisione CoUegamento Ricerche e Sperirnentazione, on $e basis of the agreement protocol CONI-MPI. Direct correspondence to Laura Bortoli, Istituto Superiore Educazione Fisica, via dei Colli, 35 100 Padova, Italia.
556
L. BORTOLI, ETAL
Lee, 1989; Lee & Magill, 1983; see Magdl & Hall, 1990, for a review). Few, however, are the experiments aimed at generalizing the advantageous effects of contextual interference in instructional settings (Goode & Magill, 1986; French, Rink, & Werner, 1990; Pigott & Shapiro, 1984; Wrisberg, 1991; Wrisberg & Liu, 1991). Goode and Magill (1986), for example, using three types of badminton serves noticed, as in laboratory studies, better performances in retention and transfer phases by a random practice group; blocked practice, instead, led to high performance in acquisition, but low in retention and transfer ~ h a s e s .Wrisberg and Liu (1991) also, using badminton tasks, conducted an experiment in a standard physical education instructional setting to determine the effect of contextual variety on learning and transfer of the long and short serve. Although little difference was observed in the performance among the blocked and the varied groups during practice, the varied group showed higher retention of the short serve and higher transfer of both serves. French, et al. (1990) obtained no significant differences in acquisition and retention for three groups of high school children, assigned to one of three practice conditions (random, random-blocked, or blocked practice) wherein the students during their physical education classes practiced the volleyball forearm pass, set, and overhead serve. The aim of this paper, in certain aspects similar to that of French, et al. (1990), was to examine the possibihty of generalizing effects of contextual interference on retention and transfer, verified by laboratory tasks, to the activities carried out during physical education classes. In this way, important observations reported in the literature can be checked for application in an instructional context. The skills here considered are the technical skills of volleyball (volley, bump, and serve), placed in a learning context that takes into account the actual school situations of high school students. The learning objective was not so much the correct execution referred to an ideal technical model but the accuracy of hitting a target. If effects of contextual interference are shown, a random organization should lead to a better score on retention and transfer tests than blocked practice. Next to the random and blocked conditions that represent, respectively, situations of high and low interference, two other modalities were considered in the present work, serial organization (mixed interference; see Goode & Magill, 1786) and serial organization with very high interference. The serial group practiced, in a prearranged sequence, each skill six times before passing to the next one while the serial group with high interference executed, in the same prearranged order, each skill only twice before passing to the next one.
METHOD Subjects
Four groups of students i n Grade 7 (91 in all) participated in the first
CONTEXTUAL INTERFERENCE IN SPORTS
557
phase of administration of the tests. Each group was then randomly assigned to one of four treatments (blocked, random, serial, and serial with very high interference). Of the initial 91 students 52 were then chosen (13 per group; M age = 11.6 yr., SD = .7) to take part. The initial test gave homogeneous scores (F= 1.86, p>0.10).
Procedure Eight lessons were proposed to each experimental group, one lesson a week, during two months. The first session was dedicated to the initial tests on the fundamentals of volleyball; in the following six meetings the students practiced the technical elements in a way specific and differentiated for each group. The final retention and transfer tests were administered in the last session. Each subject from a specific position on the court had, using the bump, the two-hand volley, and the under-hand serve, to hit the center of a target drawn on the floor on the other side of the volleyball net (height of 2.43 meters). The target to be centered was square, with sides measuring 2.97 meters (the center of the square measured .33 meters from the sides, and each consecutive target area increased by .66 meters). These areas were assigned decreasing points from the center (5 points) to the perimeter (one point; no point if the target was missed). The points, therefore, could vary from a maximum of five to a minimum of zero points. The blocked practice group carried out in each session just one type of technical skill. The first and the fourth lessons were dedicated to the volley, the second and the fifth to the bump, the third and the sixth to the serve. The random practice group carried out in a predetermined random order the three fundamentals in each session, repeating each technical skill six times before passing to the next one. The serial practice group practiced, each session, repeating the prearranged sequence of volley, bump, and serve (each technical skill for six times before passing to the next one). The very high interference serial group, like the serial group, repeated the sequence of volley, bump, and serve, but each technical skill was only carried out twice before practice on the next one. With this procedure each subject carried out at each session 36 repetitions, and in all 72 repetitions of each technical skill in the six learning meetings (72 volleys, 72 bumps, and 72 serves) for a total of 216 trials. From the organizational point of view, three groups of students were formed, each of which occupied just one area of the volleyball court (right, left, or central); this arrangement was then maintained at each meeting, at both assessment and work. Each group, in turn, was made up of three students each of whom, in turn, practiced; the companions not directly involved in the task assisted in the recovery of the volleyballs. The signal to carry out the technical skiU was given by the teacher, at a constant rate of 5 sec. between one attempt and the next.
558
L. BORTOLI, E T A L
The practice areas, initial test, and retention test were for the bump at three meters from the net, for the volley at four meters, and for the serve at five meters. In the transfer test, at the last meeting, each fundamental was carried out one meter forward with respect to the test and practice areas (short transfer) and one meter behind (long transfer). O n the tests (initial, final, short transfer, and long transfer) the subjects carried out six trials for each technical slull in a predetermined order (bump, volley, serve). For each fundamental skill, therefore, each student could achieve a minimum of zero points and a maximum of 30 (90 considering together the three fundamentals) . The data were analyzed using the SPSS/PC + package. The initial tests were analyzed by means of a one-way variance analysis to verify the homogeneity of the data. The same procedure on the initial data was executed at the end using only the data on the subjects who participated in all eight meetings. The data for each test were analyzed, with respect to each fundamental and to the total, by a multivariate analysis of variance 4 (groups) x 2 (test, retest) factorial procedure, with repeated measures on the second factor. The analysis, a 4 (groups) x 3 (retention, short transfer, and long transfer tests) procedure, was used to observe possible differences among the groups and among the three rests of retention, short transfer, and longer transfer. The comparison of the groups was then conducted with the one-way procedure, while the comparison of the final retention, short transfer, and long transfer tests was carried out with the t test for paired data.
RESULTS The statistical indexes of the retention and transfer tests for each technical fundamental skill are reported in Table 1. From the multivariate analysis of variance for 4 (groups) x 2 (test, retest) factors of the total points for the three fundamental skills the only difference was between the initial and final tests. There was an improvement in performance from the first to the last session (F,,,, = 68.26, p 0.05). Similar results were found on the analysis of each of the three fundamental skills (bump, volley, and serve). The multivariate analysis of variance of 4 (groups) x 3 (retention, short transfer, and long transfer tests) factors did not produce differences among groups (F< 1.00) or in the interaction of the main effects (F< l.OO), while significant differences emerged between the tests (F,,9, = 28.45, p < 0.001). In particular the short transfer was the easiest test, followed by the retention test and then by the long transfer test which was more difficult than the other two. With regards to the transfer tests, analyzed with the one-way analysis,
TABLE 1 MEANA N D STANDARD DEVIATION OF POINTSEARNED ON RETENTION AND TRANSFER TESTSM RANDOM,BLOCKED, SEW, A N D SERIALWITH HIGHINTERFERENCE GROUPS(n = 13 EACHGROUP)
G coup
Bump Pretest
M Random Blocked Serial Serial, High Interference
Volley Retention
Pretest
Random Blocked Serial Serial, High Interference
Retention
SD
M
SD
M
SD
M
SD
9.62 8.85 9.77
3.01 3.56 3.68
10.38 13.31 12.00
3.52 4.39 4.16
14.62 13.77 11.62
4.54 3.22 2.43
17.46 16.69 15.69
9.85
4.10
12.46
6.67
10.00
4.74
17.92
Bump
2 4
Serve Pretest
Retention
-5? m
M
SD
M
SD
4.27 4.46 4.15
8.62 8.54 6.46
3.20 3.36 4.58
11.38 11.54 11.85
3.10 3.18 4.60
4.44
7.69
4.66
11.85
5.34
ztn
Long Transfer
2
Volley
L
2
rj
m
Serve
Z
Short Transfer M SD
Long Transfer
M
SD
M
SD
M
SD
M
SD
2
13.46 16.69 13.46
2.85 3.33 5.47
12.31 13.54 12.92
4.33 4.12 5.53
20.85 18.92 18.31
3.21 4.73 5.09
16.54 15.38 12.54
4.16 5.12 3.31
13.54 12.23 13.62
5.04 5.34 5.59
11.85 8.31 11.77
4.26 4.17 3.81
v,
14.77
4.32
12.54
4.10
21.08
2.93
14.77
3.17
13.15
4.36
9.23
2.52
Short Transfer
Long Transfer M SD
Short Transfer
5 ? Ln
560
L. BOKTOLI, ETAL.
differences emerged among groups in relation to the test for the serve in long transfer (F,,,, = 2.97, p
EFFECTS O F CONTEXTUAL INTERFERENCE O N LEARNING TECHNICAL SPORTS SKILLS LAURA BORTOLI, CLAUD10 ROBAZZA, VALTER DURIGON, CARLO CARRA Physical Education Institute of Padova Summary.-The effects of contextual interference on learning s k i h of volleyball (volley, bump, serve) are influenced by the scheduling of actual practice sessions: the activities can be proposed in a repetitive practice schedule (blocked practice) by continuously repeating the same task (low interference) or in random practice schedules by performing more tasks or variations of one same activity (high interference). High contextual interference, even though causing immediate limited performance, leads to superior performance on retention and transfer tests. Four experimental groups (13 students each) were placed in conditions of random, blocked, serial, and serial with high interference practice for 8 meetings (2 tests and 6 practice). Analysis yielded s i g d i cant differences among the groups on a transfer test (long transfer) for the serve, so results in this instructional setting are partially in line with those generally found in laboratory experiments.
The concept of contextual interference was introduced by Battig (1966, 1979) to show the aspects of interference that occur when one practises a certain task within a specific context: high interference is observed if in the same session different tasks are performed, even if they are similar or related and low interference when one practises, instead, just one task. From studies of verbal material learning, Battig noticed that high contextual interference situations, generally considered a cause of forgetfulness, led to better performance on retention and transfer tests than a situation of low interference. Contextual interference was considered an instrument that permitted overcoming some of the limits of mnestic performance: according to the principle of specificity encoding, for example, to improve memory performance it is necessary to reinstate the original context in which the learning took place. The diversity of situations in learning would render, on the other hand, the subject less dependent on the context and would stimulate him in multiple and variable elaborative activities to be able to produce better retrieval of information and a greater adaptation to the different contextual situations. Shea and Morgan (1979) were the first to verify Battig's predictions for motor skills learning, showing the superiority of random practice schedules over blocked practice schedules in facihtation of retention and transfer. Many other studies have led to similar results for a variety of laboratory tasks (Del Rey, 1982, 1989; Del Rey, Whitehurst, & Wood, 1983; Gabriele, Hall, & 'This research was financed by CONI, Divisione CoUegamento Ricerche e Sperirnentazione, on $e basis of the agreement protocol CONI-MPI. Direct correspondence to Laura Bortoli, Istituto Superiore Educazione Fisica, via dei Colli, 35 100 Padova, Italia.
556
L. BORTOLI, ETAL
Lee, 1989; Lee & Magill, 1983; see Magdl & Hall, 1990, for a review). Few, however, are the experiments aimed at generalizing the advantageous effects of contextual interference in instructional settings (Goode & Magill, 1986; French, Rink, & Werner, 1990; Pigott & Shapiro, 1984; Wrisberg, 1991; Wrisberg & Liu, 1991). Goode and Magill (1986), for example, using three types of badminton serves noticed, as in laboratory studies, better performances in retention and transfer phases by a random practice group; blocked practice, instead, led to high performance in acquisition, but low in retention and transfer ~ h a s e s .Wrisberg and Liu (1991) also, using badminton tasks, conducted an experiment in a standard physical education instructional setting to determine the effect of contextual variety on learning and transfer of the long and short serve. Although little difference was observed in the performance among the blocked and the varied groups during practice, the varied group showed higher retention of the short serve and higher transfer of both serves. French, et al. (1990) obtained no significant differences in acquisition and retention for three groups of high school children, assigned to one of three practice conditions (random, random-blocked, or blocked practice) wherein the students during their physical education classes practiced the volleyball forearm pass, set, and overhead serve. The aim of this paper, in certain aspects similar to that of French, et al. (1990), was to examine the possibihty of generalizing effects of contextual interference on retention and transfer, verified by laboratory tasks, to the activities carried out during physical education classes. In this way, important observations reported in the literature can be checked for application in an instructional context. The skills here considered are the technical skills of volleyball (volley, bump, and serve), placed in a learning context that takes into account the actual school situations of high school students. The learning objective was not so much the correct execution referred to an ideal technical model but the accuracy of hitting a target. If effects of contextual interference are shown, a random organization should lead to a better score on retention and transfer tests than blocked practice. Next to the random and blocked conditions that represent, respectively, situations of high and low interference, two other modalities were considered in the present work, serial organization (mixed interference; see Goode & Magill, 1786) and serial organization with very high interference. The serial group practiced, in a prearranged sequence, each skill six times before passing to the next one while the serial group with high interference executed, in the same prearranged order, each skill only twice before passing to the next one.
METHOD Subjects
Four groups of students i n Grade 7 (91 in all) participated in the first
CONTEXTUAL INTERFERENCE IN SPORTS
557
phase of administration of the tests. Each group was then randomly assigned to one of four treatments (blocked, random, serial, and serial with very high interference). Of the initial 91 students 52 were then chosen (13 per group; M age = 11.6 yr., SD = .7) to take part. The initial test gave homogeneous scores (F= 1.86, p>0.10).
Procedure Eight lessons were proposed to each experimental group, one lesson a week, during two months. The first session was dedicated to the initial tests on the fundamentals of volleyball; in the following six meetings the students practiced the technical elements in a way specific and differentiated for each group. The final retention and transfer tests were administered in the last session. Each subject from a specific position on the court had, using the bump, the two-hand volley, and the under-hand serve, to hit the center of a target drawn on the floor on the other side of the volleyball net (height of 2.43 meters). The target to be centered was square, with sides measuring 2.97 meters (the center of the square measured .33 meters from the sides, and each consecutive target area increased by .66 meters). These areas were assigned decreasing points from the center (5 points) to the perimeter (one point; no point if the target was missed). The points, therefore, could vary from a maximum of five to a minimum of zero points. The blocked practice group carried out in each session just one type of technical skill. The first and the fourth lessons were dedicated to the volley, the second and the fifth to the bump, the third and the sixth to the serve. The random practice group carried out in a predetermined random order the three fundamentals in each session, repeating each technical skill six times before passing to the next one. The serial practice group practiced, each session, repeating the prearranged sequence of volley, bump, and serve (each technical skill for six times before passing to the next one). The very high interference serial group, like the serial group, repeated the sequence of volley, bump, and serve, but each technical skill was only carried out twice before practice on the next one. With this procedure each subject carried out at each session 36 repetitions, and in all 72 repetitions of each technical skill in the six learning meetings (72 volleys, 72 bumps, and 72 serves) for a total of 216 trials. From the organizational point of view, three groups of students were formed, each of which occupied just one area of the volleyball court (right, left, or central); this arrangement was then maintained at each meeting, at both assessment and work. Each group, in turn, was made up of three students each of whom, in turn, practiced; the companions not directly involved in the task assisted in the recovery of the volleyballs. The signal to carry out the technical skiU was given by the teacher, at a constant rate of 5 sec. between one attempt and the next.
558
L. BORTOLI, E T A L
The practice areas, initial test, and retention test were for the bump at three meters from the net, for the volley at four meters, and for the serve at five meters. In the transfer test, at the last meeting, each fundamental was carried out one meter forward with respect to the test and practice areas (short transfer) and one meter behind (long transfer). O n the tests (initial, final, short transfer, and long transfer) the subjects carried out six trials for each technical slull in a predetermined order (bump, volley, serve). For each fundamental skill, therefore, each student could achieve a minimum of zero points and a maximum of 30 (90 considering together the three fundamentals) . The data were analyzed using the SPSS/PC + package. The initial tests were analyzed by means of a one-way variance analysis to verify the homogeneity of the data. The same procedure on the initial data was executed at the end using only the data on the subjects who participated in all eight meetings. The data for each test were analyzed, with respect to each fundamental and to the total, by a multivariate analysis of variance 4 (groups) x 2 (test, retest) factorial procedure, with repeated measures on the second factor. The analysis, a 4 (groups) x 3 (retention, short transfer, and long transfer tests) procedure, was used to observe possible differences among the groups and among the three rests of retention, short transfer, and longer transfer. The comparison of the groups was then conducted with the one-way procedure, while the comparison of the final retention, short transfer, and long transfer tests was carried out with the t test for paired data.
RESULTS The statistical indexes of the retention and transfer tests for each technical fundamental skill are reported in Table 1. From the multivariate analysis of variance for 4 (groups) x 2 (test, retest) factors of the total points for the three fundamental skills the only difference was between the initial and final tests. There was an improvement in performance from the first to the last session (F,,,, = 68.26, p 0.05). Similar results were found on the analysis of each of the three fundamental skills (bump, volley, and serve). The multivariate analysis of variance of 4 (groups) x 3 (retention, short transfer, and long transfer tests) factors did not produce differences among groups (F< 1.00) or in the interaction of the main effects (F< l.OO), while significant differences emerged between the tests (F,,9, = 28.45, p < 0.001). In particular the short transfer was the easiest test, followed by the retention test and then by the long transfer test which was more difficult than the other two. With regards to the transfer tests, analyzed with the one-way analysis,
TABLE 1 MEANA N D STANDARD DEVIATION OF POINTSEARNED ON RETENTION AND TRANSFER TESTSM RANDOM,BLOCKED, SEW, A N D SERIALWITH HIGHINTERFERENCE GROUPS(n = 13 EACHGROUP)
G coup
Bump Pretest
M Random Blocked Serial Serial, High Interference
Volley Retention
Pretest
Random Blocked Serial Serial, High Interference
Retention
SD
M
SD
M
SD
M
SD
9.62 8.85 9.77
3.01 3.56 3.68
10.38 13.31 12.00
3.52 4.39 4.16
14.62 13.77 11.62
4.54 3.22 2.43
17.46 16.69 15.69
9.85
4.10
12.46
6.67
10.00
4.74
17.92
Bump
2 4
Serve Pretest
Retention
-5? m
M
SD
M
SD
4.27 4.46 4.15
8.62 8.54 6.46
3.20 3.36 4.58
11.38 11.54 11.85
3.10 3.18 4.60
4.44
7.69
4.66
11.85
5.34
ztn
Long Transfer
2
Volley
L
2
rj
m
Serve
Z
Short Transfer M SD
Long Transfer
M
SD
M
SD
M
SD
M
SD
2
13.46 16.69 13.46
2.85 3.33 5.47
12.31 13.54 12.92
4.33 4.12 5.53
20.85 18.92 18.31
3.21 4.73 5.09
16.54 15.38 12.54
4.16 5.12 3.31
13.54 12.23 13.62
5.04 5.34 5.59
11.85 8.31 11.77
4.26 4.17 3.81
v,
14.77
4.32
12.54
4.10
21.08
2.93
14.77
3.17
13.15
4.36
9.23
2.52
Short Transfer
Long Transfer M SD
Short Transfer
5 ? Ln
560
L. BOKTOLI, ETAL.
differences emerged among groups in relation to the test for the serve in long transfer (F,,,, = 2.97, p
