This article was downloaded by: [Temple University Libraries]On: 12 November 2014, At: 20:38Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41Mortimer Street, London W1T 3JH, UK

The Journal of Educational ResearchPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/vjer20

Task Structures, Student Practice, and Skill in PhysicalEducationStephen Silverman a , Prithwi Raj Subramaniam a & Amelia Mays Woods ba University of Illinois at Urbana-Champaignb Indiana State UniversityPublished online: 01 Apr 2010.

To cite this article: Stephen Silverman , Prithwi Raj Subramaniam & Amelia Mays Woods (1998) Task Structures, Student Practice, andSkill in Physical Education, The Journal of Educational Research, 91:5, 298-307, DOI: 10.1080/00220679809597557

To link to this article: http://dx.doi.org/10.1080/00220679809597557

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in thepublications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations orwarranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsedby Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified withprimary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings,demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectlyin connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone isexpressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Task Structures, Student Practice, and Skill in Physical Education STEPHEN SILVERMAN PRITHWI RAJ SUBRAMANIAM University of Illinois at Urbana-Champaign

ABSTRACT Task structures and practice variables for students of differing skill levels were examined. Eight teach- ers and their middle school-aged students were videotaped during 2 class periods so that all instruction could subse- quently be coded. Teachers ranked their students on per- ceived skill level. In each class, 3 students at each of the 3 per- ceived skill levels (high, medium, and low) were selected (N = 72). Process data were collected on the tasks that teachers implemented and on the appropriate and inappropriate prac- tice trials executed by each of the selected students. Analyses occurred at different levels and indicated that task organiza- tion is associated with both the quantity and quality of stu- dent practice. Although the students of differing skill levels had similar numbers of practice trials, the relationships between trials and some organizational variables differed among skill levels.

esearch on teaching has yielded a great deal of infor- R mation on the dynamics of the teaching-learning envi- ronment. These gains, which have helped both practitioners and researchers understand the variables that relate to stu- dent learning, have been seen in a variety of classroom sub- jects, as well as in specialty areas. One specialty area that has been the focus of considerable research on teaching is physical education.

The most consistent and powerful finding in research on teaching physical education (RT-PE) is that appropriate (or successful) practice trials are related to student achievement (Ashy, Lee, & Landin, 1988; Buck, Harrison, & Bryce, I99 1 ; Dugas, 1984; PiCron, 1983; Silverman, 1985a, 1990). Although the relationship attenuates at high levels of prac- tice (Silverman, 1990), the more appropriate practice trials a student receives the greater the student’s motor-skill learn- ing. Conversely, inappropriate (or unsuccessful) practice trials are negatively related to achievement (Silverman, 1990). In addition, appropriate practice has been shown to be the variable that plays the greatest role in motor-skill learning in physical education classes (Silverman & Tyson, 1994).

Although student practice is a central variable in student learning in physical education, the relationships between practice and other teaching variables have not been inves- tigated directly. Some physical education research (Gra-

AMELIA MAYS WOODS Indiana State University

ham, 1987; Gusthart & Springings, 1989; Pellett & Harri- son, 1995a, 1995b; Rikard, 1991; Silverman & Tyson, 1994; Silverman, Tyson, & Krampitz, 1992; Werner & Rink, 1989) suggests that process variables are interrelat- ed and that examining these interrelationships will lead to a better understanding of teaching. Understanding how teachers’ structuring of the class environment is associat- ed with greater appropriate student practice clarifies these interrelationships. As was shown in the Beginning Teacher Evaluation Study (BTES; Fisher et al., 1981), what teach- ers do may be related to the quality and amount of student engagement.

One way in which teachers structure their classes is by assigning tasks and implementing accountability measures for these tasks. Siendentop and his students (Graham, 1987; Jones, 1992; Lund, 1991; Marks, 1989; Tousignant & Siedentop, 1983) have shown that a variety of task struc- tures and accountability mechanisms exist in physical edu- cation. Hastie and Saunders ( 1990, 199 I , 1992) investigat- ed accountability in secondary schools and found that teacher monitoring is related to student behavior (Hastie & Saunders, 1990) and that accountability plays an important role in physical education (Hastie & Saunders, 1992). Based on this research and on classroom research that sug- gested a relationship between student learning and the way teachers structure and present tasks and hold students accountable (Doyle, 1978, 1979, 1983; Doyle & Carter, 1984), Silverman, Kulinna, & Crull (1995) found similar results in physical education classes.

The way tasks are organized can be thought of in a num- ber of ways. In physical education, Silverman and his asso- ciates (Silverman, Tyson, & Morford, 1988; Silverman et al., 1995) examined the practice methods that the teacher assigned to students. These researchers found that the amount of time spent in practice situations in which stu- dents receive feedback was related to student achievement and that all types of practice were not equivalent. Rink and

Address correspondence and requests for a copv of the techni- cal manual on the data collected for this study to Stephen Silver- man. Universiw of Illinois at Urbana-Champaign, Louise Freer Hall, 906 South Goodwin Avenue, Urbana, IL 61801. E-mail: ssilverm @ uiuc.edu.

298

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

May/June 1998 [Vol. 91(No. 5)] 299

her associates (French et al., 1991; Rink, 1994, 1996; Rink, French, Werner, Lynn, & Mays, 1992) viewed tasks as con- tent development. Initially, teachers present the task; then, with the information gleaned from observing students exe- cute the task, the teachers modify subsequent tasks. These modifications can be thought of as extension (the teacher introduces a task that changes the complexity or difficulty of the task), refinement (the teacher refines the task because of a concern for student performance), and application (the teacher moves the students from how-to-do movement to how-to-use movement) tasks. Rink (1996) concluded that “appropriate progressions for learners is a critical function of effective teaching” (p. 190).

Student skill level complicates the direct examination of the relationship between tasks and accountability and stu- dent practice in physical education. Student skill level has been shown to mediate many teaching effects in physical education (Graham, 1987; Grant, Ballard, & Glynn, 1989; Rikard, 1992; Silverman, 1985b, 1993; Silverman, Tyson, & Krampitz, 1993). For example, different practice- achievement relationships exist for high- and low-skilled students (Silverman, 1985b, 1993). Understanding the teaching-learning process for students of different skill lev- els provides much more information than merely examining all students as a group.

Our purpose in this study was to investigate the interrela- tionships between the ways teachers structure practice tasks and the amount of practice students receive. Specifically, we investigated the following research questions: (a) Are there differences in the number of students’ appropriate and inappropriate practice trials among teachers or skill levels? (b) Is total time in practice related to practice trials? (c) Are the total number of tasks related to practice trials‘? (d) Is there a difference in the number of practice trials among task type, explicitness, and skill-level category? and (e) Is there a difference in the number of practice trials among task organization, explicitness, and skill-level category?

We integrated these three levels of variables-task struc- ture, practice, and student skill level-into one study. Al- though much is known about each variable independently, the combination of variables in one study should provide an understanding of the intermediary effects of teacher action on student practice and how these effects vary for students of different skill levels. Student mediation of teacher inten- tions occurs in instructional settings (Doyle, 1978, 1979, 1983); understanding the factors that relate to higher levels of student practice is important to a complete theory of stu- dent learning in physical education.

Method

We created a database of eight physical education class- es for a multidimensional study of task structures and stu- dent variables in physical education. Two days of instruc- tion were videotaped. After instruction, the teachers rank ordered students on perceived skill levels. Data on task

structures and accountability and individual student practice were coded from the videotapes for this study. We per- formed analyses to examine relationships and differences in student practice among the tasks for students of differing skill level.

Participants

Eight physical education teachers and their students were the participants in this study. The teachers were recruited through local school districts; the school district, principal, and teachers consented to participate. The 8 teachers ranged from a 1 st-year teacher to a teacher with 2 1 years’ experi- ence. Information on each teacher, the grade level of the class, and the focus of instructional activity for this study are presented in Table I .

The student participants were the regularly assigned stu- dents. We used appropriate school district procedures to inform parents of the study and to obtain permission for the students to participate. As indicated in Table 1, the students were in the seventh, eighth, and ninth grades.

Instruction

Each teacher was asked to teach in a row two classes in which motor skill was the focus of instruction. We selected the first two classes in the unit for this study because the classes at the beginning of a unit typically focus on skill development. We selected 8 teachers as participants to increase the probability of seeing different tasks and accountability systems. In addition, we used two class ses- sions so that a number of tasks from each teacher would be observed. We sought representation of multiple skills to provide a broader base for understanding teaching and learning in physical education. The skills taught as a part of the study were badminton, basketball, frisbee, soccer, and volleyball. The teachers made all instructional decisions, with the goal of enhancing motor-skill learning.

Each class was approximately 30 min. Instruction was videotaped with a two-camera split-screen set-up so that the entire area of the gymnasium was recorded for subsequent coding. The teachers wore portable microphones, and the

Table l.--l)escription of Teachers, Grade, and Activities

Years Teacher Gender teaching Grade Activity

1 Female 8 8, 9 Volleyball 2 Male 10 8, 9 Volleyball 3 Male 16 8 Soccer 4 Female 21 8 Volleyball 5 Female 16 8 Badminton 6 Female 2 8 Basketball 7 Female 18 7 Ultimate frisbee 8 Male 1 8 Ultimate frisbee

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

300 The Journal of Educational Research

audio signal was recorded simultaneously with the video signal. In addition, to record elapsed time, we used an elec- tronic stopwatch that superimposed its image on the tape. During instruction, the students wore numbered and colored pinafores so they could be identified on videotape.

Skill Level Determination and Selection of Students

In order to use skill level as a variable in this study, on the day when instruction concluded and 2-4 days later, we had the teachers rank order every student on their perception of student skill level. Teachers’ ratings to determine skill level has been used previously by Martinek (1988) and Graham (1987) to identify students of varying skill. Using the Kendall coefficient of concordance, we compared the 2nd day’s rankings with the 1st day’s to determine reliability of the rankings. The reliability between the two measures was above .85 for all 8 teachers.

Based on the 1st day’s rankings, we categorized the stu- dent’s skill as high, medium, or low. Three students at each skill level in each class (n = 9 in each class) were random- ly selected for subsequent practice trial coding. This result- ed in 72 participants (n = 24 at each skill level). The distri- bution of students by skill level and gender is presented in Table 2.

Videotape Coding

Trained coders coded videotapes to collect data for each class on tasks and accountability and practice trials for each of the 72 students in this study. Additional data were col- lected on teacher feedback to individual students; those feedback data are part of a larger study and not reported here.

nKo of the authors collected all data. Training occurred before data collection. The instruments initially were based on previous research (French et al., 1991; Rink, 1994; Rink et al., 1992; Silverman, 1990, 1993; Silverman et al., 1992, 1993, 1995). We developed a technical report and pilot test- ed the instrument to determine if reliability could be obtained on the database. During training, modifications were made to each instrument; along with specific deci- sions, the modifications were recorded in a revised copy of the technical report. Training and instrument refinement

Table 2 .Student Skill Level, by Gender

Total for Female Male skill level

Skill level n % n % n %

High 6 8.3 18 25.0 24 33.3 Medium 10 13.9 14 19.4 24 33.3 L O W 20 21.8 4 5.6 24 33.3

All skill levels 36 50.0 36 50.0 72 100.0

took place over hundreds of hours and 5 months. Data col- lection occurred after three successive occasions of interob- server agreement on all parts of the instrument at levels of .90 or above. Specifics on how agreement was calculated are presented with each instrument.

Tusks and accountability coding. The task instrument required the coder to collect data on task type (e.g., exten- sion, refinement, application); task organization (e.g., indi- vidual, reciprocal, small-group practice and lead-up game); task explicitness (combinations of outcome, situation, crite- rion-product, and criterion form); primary and secondary accountability (e.g., monitoring with individual skill-relat- ed feedback); task presentation time, and total task time. Categories and subcategories of the task instrument are pre- sented in the Appendix.

As coders viewed the videotape, they observed each task presentation and recorded the time when the presentation began. The coders focused on the instructions that students received on the task they were to complete and how they were to complete it. Therefore, extended explanations and demonstrations were not included in task presentation. Dur- ing the presentation, the coder recorded the task type and organization intended by the teacher and the extent to which the teacher was explicit in presenting the task. The coder recorded the time when the presentation ended and when the task began. During the actual practice task, the coder deter- mined the primary and secondary accountability mecha- nisms used by the teacher. If only one type of accountabili- ty was used, i t was recorded as primary. If multiple accountability mechanisms were used, the coder noted the more predominant device as primary and the next important as secondary. Finally, the time was recorded when the task ended. The coders were free to rewind the tape and reexam- ine any part of the task before recording data.

At the beginning and throughout the data collection, the coders checked reliability on 7 of the 16 class sessions. The second coder coded the entire class; reliability was calcu- lated by determining the intraclass correlation coefficient for presentation time and the amount of time spent in each unique combination of task variables. For the task data, all reliability checks were 3 5 or higher.

Practice trial coding. Within each task, the 9 randomly selected students were coded for the number of practice tri- als each student executed. Each practice trial was coded as an appropriate practice trial (successful trial) or an inappro- priate practice trial (unsuccessful trial) based on the diffi- culty the student had in completing the trial. The coder watched each task again, following 1 student each time, and repeated this procedure until all student data were recorded. The coders identified the students based on the numbered pinafores and notes that were taken by one of the authors during instruction. The notes contained practice formations and the students’ initial placement for each task. In addition, student names were a part of the note packet that was designed to help the coders find and follow the student. The coders rewound the videotape as needed to collect data.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

MayIJune 1998 [Vol. 91(No. S)] 301

From the seven class sessions that were coded for relia- bility, we determined practice-trial coding reliability by using intraclass correlation coefficients for the number of appropriate and inappropriate practice trials in each unique combination of task variables. All reliability checks were .90 or higher for the trial data.

Data Andysis

Task and accountability data. Data were summed across the two class sessions for each class. We calculated both the number of tasks and the time spent in each task classifica- tion. In addition, we calculated combinations of task explic- itness (i.e., when the teacher used more than one explicit- ness variable in presenting the task--e.g., the teacher presented the task by telling students the outcome, situa- tion, and criterion-product) and accountability (a combina- tion of the primary and secondary accountability cate- gories). These calculations resulted in aggregate categories that could be used in the statistical analysis and were based on previous research on tasks and accountability (Silverman et al., 1995).

Practice trial data. We examined the trials within each task combination in the study; therefore, we coded the prac- tice trials in the data set so they were tied to specific tasks. Within each task, additional variables (total trials; percent- age of appropriate trials; and appropriate, inappropriate, and total trials per minute) were calculated for each student.

Statistical anulysis. To examine the relationships between tasks and practice trials for students of varying skill level, we summarized data from each task at the student, skill, and class levels. This permitted the appropriate unit of analysis to be used in each statistical analysis. We used individual students and the mean for each skill level within each class for analyses, where appropriate. Because tasks varied in length based on the teachers’ structuring of the class, we used the number of (appropriate, inappropriate, or total) tri- als per minute as the dependent variable.

We performed the first level of analysis with students as the unit of analysis. We examined differences in the number of appropriate, inappropriate, and total trials and the per- centage of appropriate trials among classes and skill levels. An analysis of variance (ANOVA) (Teacher x Skill Level) was computed for percentage of appropriate trials. A simi- lar multivariate analysis of variance (MANOVA) was per- formed for appropriate and inappropriate trials. A signifi- cant MANOVA was followed by discriminant analysis, ANOVA, analysis of covariance (ANCOVA), and the Stu- dent-Newman-Kuels test to determine differences. Pearson product-moment correlation coefficients were calculated between total time spent on practice tasks and the practice- trial variables. In addition, class mean values for each prac- tice variable were correlated with the total number of tasks. We performed additional correlations at each skill level and correlated the class skill-level means for practice variables with the number of tasks.

We used similar procedures when examining differences in practice trials for various task variables. In those analy- ses, we did not use accountability because teachers showed little variability. For the task-organization analyses, we eliminated tasks in which the organization was a lead-up game or a game was dropped because it was used only once across all classes. The task served as the unit of analysis, and the means for each skill level were computed for each task in each class. Because of similarities, we completed separate analyses for task types and task organizations. We completed separate three-way MANOVAs (Task v p e x Explicitness x Skill Level) and (Task Organization x Explicitness x Skill Level) with appropriate and inappropri- ate trials per minute as the dependent variables. Because of empty cells in the three-way analysis and large differences in cell sizes, we followed only significant MANOVA main effects as indicated above. We computed separate and sim- ilar ANOVAs for percentage of appropriate trials and total practice trials per minute.

Results

A variety of results were found at each level of analysis. When the student was the unit of analysis, there was a sig- nificant skill level by class interaction effect, F(14,48) = 2.77, p < .005), for the percentage of appropriate trials. All three skill levels (low, medium, and high) in Class 6 had a greater percentage of appropriate trials than did Class I-low-skilled, Class 2-low-skilled, Class 5-medium- skilled, and Class 5-low-skilled students. In addition, Class 5-low-skilled students scored significantly lower than all class skill-level groups except Class I-low-skilled, Class 2-low-skilled, and Class 5-medium-skilled students. Means and standard deviations for all practice variables are presented in Table 3.

The MANOVA for differences among classes and skill levels for appropriate and inappropriate trials revealed both a significant class (Wilks’s lambda = .003, F(14, 94) = 124.20, p c .001) and skill-level (Wilks’s lambda = .703, F(4, 94) = 4.53, p < ,005) effect. Follow-up analysis indi- cated that Class 6 had significantly more appropriate trials than all other classes and that Class 4 had significantly more appropriate trials than all other classes except Class 6. In addition, Class 1 had significantly more appropriate trials than Classes 8, 3, 7, and 2, respectively. Class 4 had more inapproproiate trials than all other classes, and Classes 5 and 1 had significantly more appropriate trials than all classes other than Class 4. For skill level, however, follow- up analyses did not indicate a difference.

The total number of appropriate, r(70) = .66, p < .OO I , inappropriate, r(70) = .21, p < .05 and total, r(70) = .67, p < .001, trials were related to the total time spent in prac- tice. In addition, when we performed the analyses by skill level, we found high positive correlations for all skill levels for appropriate trials. For inappropriate trials, this relation- ship was significant only for high-skilled students. All cor-

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

302 The Journal of Educational Research

’hble 3.-Means (Standard Deviations) of Practice Variables, by Class and Skill Level

Low skilled Medium skilled High skilled All students Class Appro. Inappro. % appro. Appro. Inappro. % appro. Appro. Inappro. % appro. Appro. Inappro. % appro.

I

2

3

4

5

6

7

8

(

39.6 42.7 .so (21.4) (21.1) (.25)

9.3 7.0 .49 (4.9) (2.6) (.14)

24.3 2.33 3 1 ( 1 1 . 1 ) (2.5) (.22)

182.0 41.0 .78 (37.0) (1.0) (.04)

12.7 38.3 .24 (3.1) (4.5) (.03)

755.7 4.7 .98 (78.2) (1.5) (.02)

35.0 10.7 .74 (3.5) (2.1) (.06)

37.3 11.0 .69 (7.6) (2.6) (.lo)

137.0 19.7 .65 :246.1) (18.1) (.25)

73.0 21.7 .76 138.0 28.0 .82 (39.0) (3.2) (.09) (35.5) (18.3) (.12)

16.7 5.0 .60 13.7 5.0 .69 (6.8) (1.7) (.28) (3.1) (1.0) (.06)

35.3 4.3 .84 53.7 11.0 .82 (8.1) (3.2) (.05) (17.7) (1.73) (.W)

191.7 46.0 .79 216.3 40.7 3 2 (32.7) (24.4) (.12) (26.3) (15.5) (.09)

57.7 40.7 .46 98.0 19.7 .84 (34.1) (5.1) (.20) (38.0) (8.7) (.06)

708.7 12.7 .96 791.7 18.7 .93 (49.8) (10.7) (.01) (42.3) (11.2) (.03)

39.7 11.0 .75 27.0 16.7 .66 (9.6) (1.0) (.05) (2.6) (4.6) (.04)

39.0 14.0 .77 45.0 12.0 .79 (9.6) (10.5) ( . I I ) (10.1) (3.5) (.09)

145.2 19.4 .74 172.9 18.0 .80 (224.9) (17.4) (.19) (19.0) (13.6) ( . I I )

83.6 30.8 .70 (51.8) (16.9) (.20)

13.2 5.7 .59 (5 .5 ) (1.9) (.IS)

37.8 5.9 .82 (17.0) (4.5) (.12)

196.7 42.6 .80 (31.9) (14.7) (.08)

56.1 32.9 .5 1 (44.6) (11.4) (.28)

752.0 12.0 .96 (62.4) (9.9) (.03)

33.9 12.8 .72 (7.4) (3.9) (.06)

40.4 12.3 .75 (8.7) (5.9) (.lo)

151.7 19.4 .73 (237.1) (16.2) (.20)

Nofr. N = 72; n = 3/class skill level.

relation coefficients for each variable by skill level are shown in Table 4.

When the total number of tasks used by a teacher was correlated with practice trial variables, there were no sig- nificant relationships. When the correlations were complet- ed separately for each skill level, appropriate practice trials, r(6) = .77, p < .05, and total trials, 46) = .65, p < .05, were significantly related to the number of tasks for low-skilled students but not for medium- or high-skilled students.

There were only a few significant differences for the task-type analyses. We found significant task type by explicitness interaction effect, Wilks’s lambda = .67, F(22, 226) = 2.27, p < .005. Because of large variability in cell size, the follow-up analyses did not indicate differences. We found the same result for the significant task-type effect Wilks’s lambda = .867, F(8, 226) = 2.07, p < .05. The explicitness data follow.

For the task organization analysis, there were significant differences, F(3, 81) = 76.56, p < .001 in total trials per minute among the task organizations used by the teachers. Individual instruction (M = 51.5 trialdmin, SD = 26.4) resulted in more trials than reciprocal practice (M = 8.8 tri- als/min, SD = 11.3), small-group practice (M = 2.9 tri- als/min, SD = 2. l), and large-group practice (M = 1.2 tri- als/min, SD = .67); reciprocal practice resulted in more practice trials than group practice did. The percentage of appropriate practice trials was significantly greater, F = 7.26, p < .OO 1, for individual practice (M = 96.2% appro- priate) than for group (M = 76.8% appropriate), small-

~~~

I Table 4.-Relationships of Task Time With Practice Mals, by Skill Level

Low Medium High All skilled skilled skilled students

Trial (n = 24) (n = 24) (n = 24) (N = 72)

Appropriate .64** .64** .70** .66** Inappropriate .I4 .I4 .39* .21*

Total trials .66** .65** .71** .67**

* p < .05. **p < ,001.

group (M = 76.6% appropriate), and reciprocal (M = 69.4% appropriate) practice. Percentage of appropriate trials per minute for task organizations and explicitness categories are reported in Table 5 .

The MANOVA for appropriate and inappropriate trials indicated a significant difference among task organizations, Wilks’slambda=.151,F(6, 160)=41.9,p<.001.Aswould be expected from the total-trials and percentage of appro- priate-trials results above, individual practice resulted in more appropriate trials per minute than any of the other organizations did. Reciprocal practice resulted in more inappropriate trials than other organizations did, and small group also had more inappropriate trials per minute than group instruction did. There also was a significant explicit- ness effect, Wilks’s lambda = .624, F( 12, 160) = 3.55, p < .001. Tasks in which situation, outcome, and criterion-

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

May/June 1998 [Vol. 91(No. 511 303

Table 5.-Means (Standard Deviations) of Percentage-Appropriate Mals per Minute for "ask Organization and Explicitness

Task organization Individual Reciprocal Small group Group All tasks

Situation

Situation+utcome

Situation-criterion form

Situation-criterion product

Situation4utcome- criterion form

Situation-outcome- criterion product

Situation-criterion form-criterion product

I .o (n = 3)

.92 (.lo)

(n = 6 )

I .o (NIA) (n = 3 )

.96 (.OW

(n = 12)

.71 (.28)

( n = 15)

.72 (.22)

(n = 15)

.81 (.22)

( n = 9)

.39 (.35) (n = 5 )

.55 (.21)

(n = 3)

.77 (.05)

(n = 3)

.69 (.26)

( n = 50)

.84

( n = 13)

.53 (.28)

(n = 6)

.73

( n = 12)

3 4 (.17)

( n = 9)

.76 (.18)

(n = 9)

.85

( n = 3 )

~ 1 9 )

~ 1 7 )

~ 0 9 )

.76

( n = 52) (.21)

.89 (.20)

(n = 6)

.77 (.lo)

( n = 6 )

.60 (.29)

( n = 6)

.86

( n = 3)

.77 (.22)

(n = 21)

.79 (.24)

(n = 34)

.72 .26)

( n = 15)

.74 (.22)

(n = 39)

.85 (.191

(n = 21)

.67

( n = 17)

.70

( n = 6)

.77 (.05)

( n = 3 )

.76 (.23)

( n = 135)

~ 2 9 )

(.22)

product were used resulted in more inappropriate trials per minute than all other combinations except tasks i n which we used situation, criterion-form, and criterion-product. Means and standard deviations for the combinations of appropriate and inappropriate trials per minute are present- ed in Table 6.

Discussion

The data from this study provide interesting insights into the teaching of physical education. The class by skill level interaction for percentage of appropriate trials indicated great differences among classes and skill groups. The stu- dents in Class 6 were very high in comparison with many other subgroups, and fewer appropriate trials were observed in low-skilled students and in Class 5 . The interaction sug- gests that it is not necessary for low-skilled students to always have a high proportion of inappropriate trials. Previ- ous research (Silverman, 1993) has shown that differences in student practice among skill levels can be either high or low; this suggests that the teacher may be important in determining practice patterns.

Various relationships were found when practice variables were correlated with time spent in practice for each skill level. Although the trends were similar for all skill levels for appropriate and total trials, it is noteworthy that inappropri-

ate trials were positively correlated with total time only for high-skilled students. High-skilled students had, on aver- age, more trials; the trend indicated that they had the high- est percentage of appropriate trials. Perhaps high-skilled students use the information from both positive and nega- tive trials and maximize the appropriate-inappropriate ratio. It is also possible that high-skilled students use knowledge of results better while practicing and incorporate this knowledge into subsequent practice trials (Magill, 1994).

The number of tasks correlated with appropriate and total trials for low-skilled students, but not for other students. That both correlated is not surprising because, overall, 76% of the trials were appropriate. It is surprising that more practice tasks may help low-skilled students have more appropriate practice trials. More tasks certainly can mean more total practice time during class. This finding may, however, also reflect that teachers changed tasks more fre- quently and had better content development in the classes with more tasks. As Rink (1996) has indicated, content development is important to learning, and performing one task for a long period of time may not be beneficial for low- skilled students.

Meaningful differences in practice were not found for the task-type analyses. Because Rink (French et al., 1991; Rink, 1994, 1996; Rink et al., 1992) found that the use of

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

304 The Journal of Educational Research

Table B.-Means (Standard Deviations) of Appropriate and Inappropriate Trials per Minute for Task Organization and Explicitness

Individual Reciprocal Small group Group All tasks Task organization Appro. Inappro. Appro. Inappro. Appro. Inappro. Appro. Inappro. Appro. Inappro.

Situation

Situation-outcome

Situation-criterion form

Situation-criterion product

Situation-outcome- criterion form

Situation+utcome- criterion product

Situation-criterion forn-criterion product

73.8 (5.4)

(n = 3)

32.9 (26.7) (n = 6)

65.1 (11.4) (n = 3)

51.2 (27.0) (n = 12)

8.8 (.99)

(n = 15)

5.6 (6.7)

(n = 15)

16.4 (17.8) (n = 9)

1.9 (2.2)

(n = 6)

2. I (1.6) (n = 3)

3.9 ( 1 . 1 ) (n = 3)

7.7 (11.6) (n = 51)

1.5 (1.6) (n = 15)

I .7 (1.4) (n = 6)

1.8 (.75)

(n = 12)

3.5 (1.6) (n = 9)

2.6 (2.1) (n = 9)

6.4

(n = 3) (2.4)

2.4 ( 1.9) (n = 54)

.28 (33)

.98 (.33)

.54 (.43)

.55 ( .56)

.53 (.43)

I .4 ( 1 . 1 )

.57 (5%)

.84

(n = 6)

I .22 (.83)

(n = 6)

.83 (.66)

(n = 6)

~ 4 0 )

.92 (.23)

(n = 3)

.96 (.60)

(n = 21)

.05 4.4 (.lo) (9.3)

(n = 36)

.34 15.9 (.26) (30.0)

(n = 15)

.27 7.9 (.05) (15.2)

(n = 39)

17.8 (23.9)

(n = 21)

.I5 2. I (.18) (2.0)

(n = 18)

4.3 (3.0)

(n = 6)

3.9 ( 1 . 1 )

(n = 3)

.2 1 8.3 (. 19) (17.1)

(n = 138)

refinement and extension tasks aid in content development, we did not expect this result. Research design differences are probably responsible for these inconsistencies. In the studies cited above, the researchers used experimental treat- ments; we used naturally occurring teaching tasks as planned by the teacher. The training and strong treatment effect from Rink’s work was important; some effects may require training to have teachers implement them at high levels.

The task-organization data did show differences, and, in most instances, during individual practice situations the stu- dents had more practice and more appropriate practice. In the classes with the most trials (particularly Classes 6 and 4), there was a reliance on individual and reciprocal prac- tice. Almost all students (or student pairs) had their own equipment (i.e., balls in these classes) and did not have to wait in line to practice. The availability of equipment and student-paced practice clearly was associated with more- and more overall appropriate-practice. The late Muska Mosston (Mosston & Ashworth, 1994) and other authors of teaching methods textbooks have suggested this for some time; we provided empirical verification of that suggestion.

Although Silverman et al. (1995) found that explicitness is related to achievement, in this study we found only a few relationships and they were not clear. The explicitness vari- ables of situation-outcome and situation+xiterion product

had the greatest numbers of appropriate trials (see Table 6). However, the standard deviations were large for all explic- itness groups, and the number of tasks in each category (3-36) varied greatly. The data for explicitness suggest that this variable may not be as powerful as the organization variables in influencing appropriate practice.

As indicated earlier, we found differences in practice variables among the three skill levels only a few times. However, we did find differential relationships with the practice variables. This may indicate that although students have about the same amount of practice, the way practice is structured differentially affects students of differing skill levels. The structure of the class may be important in pro- viding low-skilled students an equal opportunity to learn.

We incorporated multiple skills in the design of this study to obtain a diverse perspective on the variables being stud- ied. We were concerned that the activity may play a role in practice situations. This was not the case. It seems much more likely that the organizational strategies selected by the teacher were most important in influencing practice. For instance, the students in Class 6 had many practice trials in comparison with other classes; these trials were largely appropriate trials. This occurred with a team sport, basket- ball, where students often have little practice. The teacher of Class 6 relied on individual and reciprocal tasks, and there was little time spent waiting.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

May/June 1998 [Vol. 91(No. 5)J 3005

The effect of the teacher is even more apparent when one examines Class 4 and Class 2. Both teachers taught similar volleyball skills, but the students in Class 4 had, on average, 15 times the number of appropriate trials that Class 2 had. The organization selected by the teacher seems to be instru- mental in determining practice. The skill probably is not nearly as important as the way in which practice is structured.

In this study we have shown that skill level, task structure, and practice are related. In addition, the results clearly rein- force the view that teachers’ decision making about the tasks their students will complete are related to the amount and type of practice students receive. If, as has been suggested (Silverman, Devillier, tk Ramtriz, 199 I ) , appropriate prac- tice is a good proxy for student achievement, then teachers can design tasks that maximize appropriate practice and help all students learn motor skills.

NOTES

Funding for this study was provided by a grant to the first author from

The authors thank Ronald Skonie for his help in videotaping instruction the University of Illinois at Urbana-Champaign Research Board.

for this study.

REFERENCES

Ashy, M. H., Lee, A. M., & Landin, D. K. (1988). Relationships of prac- tice using correct technique to achievement in a motor skill. Journal of Teaching in Physical Education. 7, I 15- 120.

Buck, M.. Harrison, J. M., & Bryce, G. R. (1991). An analysis of learning trials and their relationship to achievement in volleyball. Journal of Teaching in Physical Education, 10, 134-152.

Doyle, W. (1978). Classroom tasks and students’ abilities. In P. L. Peterson & H. J. Walberg (Eds.). Research on teaching: Concepts. jndings. and implications (pp. 189-209). Berkeley, CA: McCuthan.

Doyle. W. (1979). Classroom effects. Theory Into Practice, 18(3).

Doyle, W. (1983). Academic work. Review .f Educational Research, S3, 159- 199.

Doyle, W., & Carter, K. (1984). Academic tasks in classrooms. Curriculum Inquiry, 14, 129-149.

Dugas, D. M. (1984). Relationships among process and product variables in an experimental teaching unit. Dissertation Abstracts International. 44, 2709A. (University Microfilms No. 84-00, 193)

Fisher, C. W., Berliner, D. C., Filby, N. N., Marliane, R., Cahen. L. S., & Dishaw, M. M. (1981). Teaching behaviors. academic learning time, and student achievement: An overview. Journal of Classroom Interaction.

French, K., Rink, J., Rikard, L., Mays, A,. Lynn, S., & Werner, P. (1991). The effects of practice progressions on learning two volleyball skills. Journal of Teaching in Phvsical Education. 10. 261-275.

Graham, K. C. (1987). A description of academic work and student per- formance during a middle school volleyball unit. Journal of Teaching in Physical Education. 7. 22-37.

Grant, B. C., Ballard, K. D., & Glynn, T. L. (1989). Student behavior in physical education lessons: A comparison among student achievement groups. The Journal of Educational Reseurch, X2, 2 16-226.

Gusthart, J. L., & Springings, E. J. (1989). Student learning as a measure of teacher effectiveness in physical education. Journal of Teaching in Phpsical Education. 8. 298-3 I 1.

Hastie. P. A., & Saunders, J. E. ( I 990). A study of monitoring in secondary school physical education classes. Journal of Classroom Intermtion, 25, 47-54.

Hastie, P. A., & Saunders, J. E. (1991). Accountability in secondary school physical education. Teaching and Teacher Education. 7. 373-382.

Hastie. P. A., & Saunders, J. E. (1992). A study of task systems and accountability in an elite junior sports setting. Journul of Teaching in Physical Education. 1 1 . 376-388.

138-144.

17(1), 2-15.

Jones, D. L. (1992). Analysis of task structures in elementary physical edu- cation classes. Journal of Teaching in Physical Education. 1 1 . 41 1-425.

Lund, J. L. (1991). Student performance and accountability conditions in physical education. Dissertation Abstracts International. 51, 3358A- 3359A. (University Microfilms No. 91-05.161)

Magill, R. A. (1994). The influence of augmented feedback on skill learn- ing depends on characteristics of the skill and the learner. Quest, 46. 314-327.

Marks, M. C. (1989). Development of a system for the observation of task structures in physical education. Dissertation Abstracts International, 49, 2579A. (University Microfilms No. 88-24, 565)

Martinek, T. J. (1988). Confirmation of a teacher expectancy model: Stu- dent perceptions and causal attributions of teaching behaviors. Research Qwrterly,for EJ-ercise and Sport. 5Y. I 18-126.

Mosston, M., & Ashworth, S. (1994). Teaching phvsical education (4th ed.). New York: Macmillan.

Pellett, T. L., & Harrison, J. M. (199%). The influence or refinement on female junior high school students’ volleyball practice success and achievement. Journal of Teaching in Physical Education. 1.5, 4 1-52.

Pellett, T. L., & Harrison, J . M. (199%). The influence of a teachers’ spe- cific, congruent, and corrective feedback on female junior high students’ immediate volleyball practice success. Journal of Teaching in Physical Education. I S . 53-63.

Pibron, M. (1983). Teacher and pupil behavior and the interaction process in p.e, classes. In R. Telama, V. Varstala. J. Tiainen, L. Laakso, & T. Haa- janen (Eds.), Research in school phvsical education (pp. 193-202). Jyvaskyla, Finland: Foundation for the Promotion of Physical Culture and Health.

Rikard, G. L. (1991). The short term relationship of teacher feedback and student practice. Journal of Teaching in Physical Education, 10, 175-185.

Rikard, G. L. (1992). The relationship of teachers’ task refinement and feedback to students’ practice success. Journal of Teaching in Physical Education. I I , 349-357.

Rink, J. E. (1994). Task presentation in pedagogy. Quest, 46, 270-280. Rink, J. E. (1996). Effective instruction in physical education. In S. Sil-

verman & C. Ennis (Eds.), Student learning in physical education: Applying research to enhunce instruction (pp. I7 1-197). Champaign, IL: Human Kinetics.

Rink, J., French, K., Werner, P., Lynn, S.. & Mays, A. (1992). The inf lu- ence of content development on the effectiveness of instruction. Journal of Teaching in Phvsical Education. 11. 139-149.

Silverman, S. (l985a). Relationship of engagement and practice trials to student achievement. Journal of Teaching in Physical Educarion, 5, 13-21.

Silverman, S. ( 1985h). Student characteristics mediating engagement-out- come relationships in physical education. Research Quarterlyfor Exer- cise and Sport, 56, 66-72.

Silverman. S. (1990). Linear and curvilinear relationships between student practice and achievement in physical education. Teaching and Teacher Education. 6. 305-314.

Silverman, S. (1993). Student characteristics, practice and achievement in physical education. The Journal of Educational Research, 87, 54-61,

Silverman, S.. Devillier, R.. & Rarnirez, T. (1991). The validity of acade- mic learning time-physical education (ALT-PE) as a process measure of student achievement. Research Quarterlv for Exercise and Sport, 62.

Silverman, S., Kulinna. P., & C ~ l l , G. (1995). Skill-related task structures, explicitness, and accountability: Relationships with student achieve- ment. Research Quarterly for Exercise and Sport, 66, 3 2 4 0 .

Silverman, S.. & Tyson. L. (1994. April). Modeling the teaching-learning process in physicul education. Paper presented at the annual meeting of the American Educational Research Association, New Orleans. LA.

Silverman, S., Qson, L., & Krampitz, J. (1992). Teacher feedback and achievement in physical education: Interaction with student practice. Teaching and Teacher Education. 8. 333-344.

Silverman, S.. Tyson. L. A., & Krampitz, J. (1993). Teacher feedback and achievement: Mediating effects of initial skill level and sex. Journal of H u m n Movement Studies, 24, 97-1 18.

Silverman, S., Tyson. L., & Morford, L. M. (1988). Relationships of orga- nization, time, and student achievement in physical education. Teaching and Teacher Education, 4, 247-257.

Tousignant, M., & Siedentop, D. (1983). A qualitative analysis of task structures in required secondary physical education classes. Journal of

3 19-325.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

306 The Journal of Educational Research

Teaching in Physical Education, 3( I ) , 41-51. Werner. P., & Rink, J. (1989). Case studies of teacher effectiveness in sec-

ond grade physical education. Journal of Teaching in Physical Educa- tion. 8, 280-297.

APPENDIX Categories and Subcategories of the ‘ h k Instrument

l b k 5 P e Informing task (I) Extension task (E) Refinement task (R) Application task (A) Repeat task (RE)

h k Organization Individual practice (I) Reciprocal practice (R) Paired group practice (PG) Small group practice (SG) Group practice (G) Lead-up game (LU) Scrimmage ( S ) Game (GA)

Tpsk Explicitness Based on:

Outcome (0) Situation (S) Criteria-product (CP) Criteria-form (CF)

Accountability Monitoring

Off-task (OT) Classlnon-skill-related

feedback (CN)

Class/skill-related feed-

Individuallnon-skill-

IndividuaVskill-related

IndividuaIhkill-related

back (CS)

related feedback (IN)

feedback (IS)

feedback with follow- UP (IF)

Aversive Exercise (AE) Practice (AP) Reprimand (AR) Point IossExtra work

(AL)

Public recognirion Public posting (RP) Accountability check

Quit when miss (RQ)

Bonus points (GB) Formal grades (GF) Pretest (GP) Practice grade (GG) Teacher records (GR)

(RA)

Grades

Send for the latest edition of the free Consumer Information Catalog. 0 Thc Catalog lists over 200 selected fedcral publications of consumer interest on subjects like health, nutrition, fedcral benefits, money management. 0 The Catalog is free and so are many of the booklets. 0 Just send your name and ad- dress, no strings attached. Write today:

Consumer Information Center Depart men t DF Pueblo, Colorado 81009

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014

Dr. ROBERT H. SEIDMAN The Journal of Educational Computing

MANAGING EDITOR Research publishes original peer- Mr. ED ZIELINSKI refereed articles on important empirical

research, conceptual and theoretical analyses, design and development REVIEW EDITORS studies, and critical reviews. Included Dr. DAVID PENNER

Dr. SHARON CARVER are: outcome effects; design and G ~ S / C O ~ ~ C T S development of innovative hardware and LISTING EDITOR software; interpretations and Dr. RAM N. SINGH implications of research in education

computing and related fields; theoretical EDITORIAL BOARD and historical foundations. The Journal Dr. Ronald E. Anderson is dedicated to advancing knowledge and Dr. Alfred Bork practice in the field. In addition, the Dr. John Seely Brown Journal serves as a forum for

professional dialogue by publishing Dr. Robert B Davis Dr. Edward J. Fuentes Dr. J.A.M. Howe correspondence, grants/contracts listing, Dr. Mark R. Lepper research notes, and reviews. Dr. Thomas T. Liao Dr. Richard E. Mayer Dr. Harry F. O’Neil, J r Dr. Seymour Papert Dr. Nancy Roberts Dr. Gavriel Salomon SUBSCRIPTION INFORMATION Dr. Michael Scnven Dr. Thomas J. Shuell Dr. Elliot Soloway Dr. John Vinsonhaler Dr. Herbert J. Walberg Dr. Joseph Weizenbaum

Dr. Karl L. Zinn

Complimentary sample issue avatlable upon request

Price per 2 volume set- 8 issues yearly Institutional Rate: $209.00 Individual Rate: $98.00 Postage & handling: $13.00 in US. & Canada; $23.50 elsewhere

Dr. William Winn ISBN 0735-6331

Baywood Publishing Company, Inc. 26 Austin Avenue, P.O. Box 337, Amityville, NY 11701

call (516) 691-1270 fax (516) 691-1770 toll-free orderline (800) 638-7819

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

20:

38 1

2 N

ovem

ber

2014