Strategy and attributional learning

Strategy and attributional learning - Md 0! bearer-ond KM...

Info iconThis preview shows pages 1–9. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Md 0! bearer-ond KM Copyright I986 b} the ARI-nun hum-cu Ateoruutm, In: 1986. Vol 7!. No. 3. 20l-209 arm moms“) 73 Strategy Training and Attributional Feedback With Learning Disabled Students Dale H. Schunk and Paula D. Cox College of Education, University of Houston In this experiment, we investigated how verbalization of subtraction with tegrouping oper- ations influenced learning disabled students‘ self-efficacy and skillful performance and also explored how effort-attributional feedback affected these achievement behaviors. Students received training and solved problems over sessions. Students in the first condition verbalized aloud while solving problems (continuous verbalization), those in the second condition verbalized only during the fust half of training (discontinued verbalization), and those in 7 the third condition did not verbalize (no verbalization). All students were periodically mon- r‘”‘”**‘m‘m’“ r ’ . «imbaeletdun‘agfihgfimhalf of training, effort feedback during the second half of training, or no effort feedback. Continuous verbalization led to‘highér self-efficacy and skillful performance than did discontinued and no verbalization; providing effort feedback promoted these achievement behaviors more than not providing feedback did. Effort feedback during the first half of training enhanced effort attributions. According to Bandura (1982a, 1982b), psychological procedures change behavior in part by creating and strengthening self-eflicacy, or one’s perceived performance capabilities in a given activity. Self-efficacy is hypothesized to influence choice of activities, effort expended, persist- ence, and task accomplishments. Although self-efficacy originally was used to help explain coping behaviors in fearful situations, its use has beenextended to other con- texts, including cognitive-skill learning (Schunk. 1985). In the present study, we tested some predictions of the self-efficacy model with learning disabled students, who perform below their measured abilities but do not possess intellectual deficits. Especially when facing difficult tasks, they often are inattentive and display lackadaisical efforts (Licht, 1984; Torgesen & Licht, 1983). These behavioral deficits may occur in part because such students hold self- doubts about their capabilities to perform well (Boersma & Chapman, 1981). Interventions that promote students’ per- ceived capabilities (i.e., self-efficacy) might help to remedy behavioral dysfunctions (Schunk, 1985). Much classroom learning involves understanding how to apply task strategies. In mathematics, students who fail to acquire algorithmic knowledge through normal instructional procedures may benefit from explicit strategy training that includes verbalizing aloud the solution steps and their ap- This research was supported by Grant MH38147 from the Na- tional Institute of Mental Health. We wish to thank the participating students in the Alief Inde- pendent School District. Alief, Texas. We are especially indebted to middle school supervisors Pat Husbands and Larry Madison and to the following people for their assistance during various phases of the project: Kay Cunnane, Sharon Davies, Rachelle LaForrest, Tony Edmonds, Brenda Grenier. Lois Heitz. Mary Lynch, Kathy Migala, Lisa Montfort. Linda Peake. and Bettye Taylor. Correspondence concerning this article should be addressed [0 Dale H. Schunk, College of Education, University of Houston. Houston, Texas T7004. 20l plication to problems. Such overt verbalization is a form of private speech, which refers to the set of speech phenomena that has a behavioral self-regulatory function but is not so- cially communicative (Vygotsky, 1962; Zivin, 1979). Overt verbalization can facilitate learning because it directs stu- dents' attention to important task features and, as a type of rehearsal, assists strategy encoding and retention (Schunk, 1982). As a means of regulating one's task performance, verbalization also can convey to students a sense of personal control over learning, which promotes self-efficacy (Ban- dura, 1982a; Schunk, 1982). Verbalization seems most beneficial for students who typically perform in a deficient manner (Denney & Turner. 1979). Positive effects of verbalization on perforrnanoe have been obtained with children who do-not spontaneously re- hearse material to be learned (Asamow & Meichenbaum, 1979), impulsive subjects (Meichenbaum, 1977), and re- medial students (Schunk, I982). Verbalization has also helped mentally retarded and emotionally disturbed students ac- quire mathematical skills (Grimm, Bijou, & Parsons, 1973; Johnston, Whitman, & Johnson, 1980; Whitman &'John- ston, 1983). Learning disabled students, who often do not use efficient plans while learning, might benefit from ver- balization to the extent that it helps them work at tasks in a systematic manner (Hallahan, Kneedler, & Lloyd, 1983; Wilder, Draper, & Donnelly, 1984). One purpose of the present study was to determine h0w verbalization during cognitive-skill learninginfluenced stu- dents' self-efficacy and skills. Students received subtraction training over six sessions. One group of Students verbalized aloud while solving problems during all sessions, the sec— ond group verbalized aloud during the first half of the train— ing program (first three sessions) but not during the second half, and the third group did not verbalize. It was expected that the two verbalization conditions would develop higher self-efficacy and skills than the no-verbalization condition. but the present study tested the hypothesis that continuity of verbalization would be less important than verbalization 202 itself (i.e., the two verbalization conditions would not dif- fer). It was expected that overt verbalization during the first half of training would help students learn how to work subtraction problems in a strategic (algorithmic) fashion. To the extent that students could then shift this means of regulating their task performance to a covert level, we felt that continued verbalization would offer no benefits. Re— searchers have shown that once strategic task behaviors are instilled, overt verbalization may be discontinued with no performance decrement (Harris, 1982; Meichenbaum, 1977).1 The second purpose of this study was to investigate how the sequence of effort-attribution] feedback affected stu- dents' self—efficacy and skills. Attributional theories pos- v tulate that individuals form causalr causes) for the outcomes of their actions (Kelley’ydr’Mich ela, 1980). In achievement contexts, students often attribute ' their successes and failures to ability, effort, task difficulty, and luck (Weiner, 1979). Effort is presumably under vo- litional control and amenable to change. Researchers have shown that linking past failures with insufficient effort pro- motes effort attributions and persistence (Andrews & De- bus, 1978; Dweck, 1975) and that effort feedback for prior successes enhances children's motivation. self-efficacy, and skills (Schunk, 1985). In the present study, Students either periodically received attributional feedback linking their successful problem solv~ ing with effort during the first half of the training program, received effort feedback during the second half of training, or did not receive effort feedback. It was expected that effort-attributional feedback would promote students' self- efficacy and skills. Effort feedback might be especially ben- eficial with learning disabled students; who often do not place sufficient emphasis on effort as a cause of outcomes (Butkowsky & Willows, 1980; Licht, 1984; Pearl, Bryan, & Donahue, 1980). A condition in which students received effort feedback throughout training was not included because our central concern was to determine how the sequence, rather than the amount, of effort feedback affected achievement out~ comes. It was predicted that effort feedback during the first half of training would misc self—efficacy and skills better than later effort feedback. We expected that effort feedback for early successes would be viewed as credible by students, given that they lacked skills and needed to expend effort to perform well. As students improve their skills and perceive that they are becoming more competent. decreasing the sal- ience of effort as a cause of success by discontinuing effort feedback may better substantiate their perceptions of com- petence (Schunk, 1984). The belief that one can perform well with less effort builds self-efficacy more than greater effort being required does (Bandura, 1982b). Conversely, effort feedback for later successes could lead students to doubt their capabilities. They might wonder why they still have to work hard and whether they can sustain the level of effort needed for success (Schunk, 1984). Method Subjects DALE H. SCHUNK AND PAULA D. COX (Grades 6 through 8). Ages ranged from ll years 2 months to 16 years 2 months (M = 13 years 7 months); l2% of the subjects had repeated at least one grade. The 5l boys and 39 girls rep— resented different ethnic backgrounds as follows: 68% white, 15% black, 11% Hispanic. and 6% Asian. The socioeconomic status of the children, as gauged by school personnel. was 65% middle class, 28% lower middle class, and 7% lower class. These ethnic background. socioeconomic status. and gender percentages ap- proximated those of the school district’s middle-school. learning disabled population. All students had previously been classified by the school district as learning disabled in mathematics according to state guidelines (Texas Education Agency, 1983). The district followed a two- stage evaluation sequence. Initially, the student's physical con~ dition, typical behavior, intelligence, and emotional stability were WEWiflfa-reacher‘mfemwmnt form, behavior rating, hearing and vision tests. and the Wechsler (1974) Intelligence Scale for Children: Form R. During the second stage, the student‘s academic achievement was assessed (Woodcock & Johnson, 1977). A student was classified as learning disabled in mathematics when his or her mathematical achievement score was more than one standard deviation (at least 16 points) lower than his or her intel- ligence score. The intelligence scores of students in this sample ranged from 80 to 115 (M = 93); mathematical achievement scores ranged from 65 to 90 (M = 75). All subjects received daily special education services in mathematics: 48% of the subjects also re- ceived reading" instruction in resource rooms. Students‘ resource-room mathematics teachers initially identi- fied 100 students who had encountered difficulties learning sub- -traction with regrouping skills. This selection procedure was followed because this study focused on processes whereby self- efficacy and skills could be developed when they were low. Five students were excluded from this sample due to absences, and five others were randomly excluded to equalize the cell sizes. Materials Attribution. The attribution measure consisted of four scales on a sheet of paper (Schunk, 1984). Each scale ranged in lO-unit intervals from not at all (0), through intemiediate values (40-60), to a whole (or (100). The four scales were labeled good at it (i.e., ability). worked hard (effort). easy problems (task), and lucky (luck). Label order was counterbalanced on four different forms. This attributional assessment is an example of a su'uctured un- idimensional scale (Elig & Frieze. 1979). Such scales assume independence of ratings and allow attributions to be assessed sep- arately. A structured scale was chosen because children seem to understand it more readily than an unstructured assessment (Di- IWe decided not to include a condition in which students ver- balized only during the second half of the training program. Al- though this condition would have created a more balanced experimental design, we felt that the best way to determine the effects of continuity of verbalization was to compare these effects with those due to discontinued verbalization. Theory and research on verbalization suggest that it may be beneficial as a means of instilling strategic behaviors and that once these behaviors have been acquired, students can regulate their performances covertly (Meichenbaum. 1977; Zivin. 1979). We also felt that asking stu- dents to begin verbalizing after they had been silently solving problems for three sessions might prove confusing and actually disrupt their performances. From an applied perspective. knowing the effects of discontinued verbalization is important; an entire _ r ,7 . . glassycrbalizing aloud would undoubtedly prove distracting to The sample included 90 students drawn from six middle schools some students. STRATEGY TRAINING 203 ruyama, 1982). ln previous research with students younger than those in the present study (Schunk. 1984), students readily understood the meaning of the scales and experienced no difficulties completing the instrument. Prior to the Schunk (1984) study, a separate re- liability assessment was conducted with 15 students who did not participate in that study. The test-retest reliability coefficient was .80. Self-(meaty. The self-efficacy test assessed students‘ perceived capabilities for correctly solving different types of subtraction problems. For this assessment, 25 scales were portrayed on five sheets of paper (5 scales per page). Each scale ranged in lO-unit angwalsm-from-Mt sureafilflfigthrough intermediate values(50-_60)w, _ ‘ to really sure (100). 'The'"’§i'imulus materials comprised 25 sample pairs of subtraction problems; each pair of problems was shown on a separate index card. The two problems constituting each pair were similar in form and operations required and corresponded to one problem on the ensuing skill test, although they involved different numbers. Reliability was assessed in conjunction with previous research by using 17 children who did not participate in that study (Bandura & Schunk, 1981). The test-retest reliability coefficient was .82. Subtraction skill. The skill test comprised 25 problems ranging from two to six columns. The problems tapped various regrouping operations. ordered from least to most difficult as follows: re- grouping once, regrouping caused by a zero, regrouping twice. regrouping from a one, and regrouping across zeros (Friend & Burton. 1981). Of these 25 problems, 12 were similar to some of the problems that subjects solved during the ensuing training ses- sions, whereas the other 13 were more complex. For example. during training, students solved problems that required regrouping twice; some problems on the skill test required regrouping three times. There were two forms of the skill test (pretest and posttest) to eliminate possible effects due to problem familiarity. These p - allel forms were developed in previous research (Bandura & Schunk, l98l); the two forms correlated highly (r = .87) in a reliability assessment conducted in conjunction with that study. Training materials. Six sets of instructional material were used. Each set incorporated one subtraction with regrouping operation ordered from least to most difficult as follows: regrouping once in two—column problems. regrouping once in three-column prob- lems. regrouping caused by a zero. regrouping twice, regrouping from a one, and regrouping across zeros (Friend & Burton. l98l). The format of each instructional set was identical. The first page of each set contained a full explanation of the relevant re- grouping operation and two examples illustrating the application of the solution strategy. The following six pages each contained several similar problems to be solved using the designated strat- egy. The problems portrayed on these six pages did not become progressively more complex but rather required that students use the solution strategy exemplified on the first (explanatory) page. Students worked on one set of material during each training ses- sion (e.g.. during Session 1, students solved problems requiring tegrouping once in two—column problems). Each set included suf- ficient problems so that students could not finish it during the session.2 Procedure Preterl. Children were administered the pretest individually by One of six- femaleadulttestersdrawn from outside the school. In admrmstenng the pretest. tests followd a Script to ensure stan- dardization across subjects. For the attributional assessment, the tester showed the paper to the student and explained that it showed four things that can help students work problems. The tester pointed out the numerical and verbal designators on each scale and ex~ plained that the higher the number a student marked on a scale. the more important he or she felt that factor was in helping him or her solve problems. The tester also provided two examples of how hypothetical students might mark the scales (e.g., a student marked 90 for worked hard because he thought that was very important. 70 for lucky because he thought that was pretty im- portant, 40 for easy problem because he felt that was somewhat important. and l0 for good at it because he did not think that was important). The tester then said to students, “I'd like you to think about the work you do in math. For example. suppose you did really well in math; that is, you worked a lot of problems correctly or you got a high score on a test. Why do you suppose that might happen?" The tester explained that marks did not have to add to a certain number (e.g., lOO). Students privately recorded their ratings. Subjects understood these directions and did not experi~ ence difficulties completing their four judgments. Students next received the self-efficacy assessment. They ini- tially received practice with the scale by judging their certainty of successfully jumping progressively longer distances. in this concrete fashion, students learned the meaning of the scale's di- rection and the different numerical values. Following this practice. students were briefly shown the 25 sample pairs of subtraction problems for about 2 s each. This brief duration allowed assessment of problem difficulty but not actual solutions; thus. students judged their capability to solve different types of problems rather than whether they could solve any par- ticular problem. The tester advised students to be honest and mark the efficacy value that corresponded to their level of certainty for being able to correctly solve the type of problem depicted. After privately making each judgment. students covered it with a blank sheet of paper to preclude observation of prior efficacy ratings from affecting subsequent judgments. The 25 judgments were summed and averaged. ficacy assessment. Each of the 25 problems was portrayed on a separate sheet of paper. The tester presented the problem to stu- dents one at a time and verbally instructed students to examine each problem and to place the page on a completed stack when they finished solving the problem or chose not to work on it any longer. Students were given no performance feedback on the ac— curacy of their solutions. The measure of skill was the number of problems solved correctly. Training sessions. Following the pretest. we randomly assigned students within gender and school to one of nine experimental conditions. All students received the program of subtraction train~ ing during 45-min sessions on 6 consecutive school days. Training sessions were conducted by one of six adult female proctor: drawn from outside the school. For any given child, the ‘ same proctor administered all six training sessions. The child's training proctor had not administered the pretest to the child and was unaware of the child’s pretest performance. At the start of each session, students met in groups of four to five with their proctor. Each proctor administered the different treatments to pre- clude confounding proctor: with treatments. Proctors followed a 1Copies of all test instruments and training materials are avail- able from the first author. ‘ Tags. seated subjects at ‘ ‘ 204 script to ensure standardized implementation of treatments across subjects. Except as noted here. the format of each training session was identical. The proctor initially reviewed the explanatory page by verbalizing aloud the solution steps and their application to the sample problems. Following this instructional phase (about 5 min). the proctor gave the appropriate verbalization instructions. All students in each small group received the same verbalization in- structions; had students in the same small group been assigned to different verbalization conditions, they might have wondered why some were not inStructed to verbalize. Students then solved the practice problems while the proctor obserVed (about 5 min); stu- dents assigned to the verbalization treatments verbaliaed aloud while solving these problems. The proctor then stressed the im- portance of performing the steps as shown on the explanatory one another, and moved out of sight. Students" solved pro ems alone during the remainder of the session (about 35 min). If they were baffled on how to solve a problem, they could consult the proctor. who reviewed the troublesome operation.3 Treatment conditions. The experimental design was a 3 x 3 (Verbalization: continuous, discontinued. or none X Effort Feed- back: first half. second half, or none) crossed factorial (N = 10 in each of the nine experimental conditions). At the start of the first training session,_the proctor told students assigned to the continuous-verbalizarion treatment the following: I’m really interested in knowing what students think about as they solve problems. So as you're working problems I'd like you to think out loud; that is, say out loud what you’re thinking about. just like I did while I was solving problems. You'll probably be drinking about what to do next, what numbers to use, how much is one number minus another, and so on. Remember. say out loud what you're thinking about. just like I did. Students were not instructed to verbalize any specific words because we did not want to constrain the nature of» their verbali- zations (Schunk, 1982). Rather, the instructions were designed to convey that students should freely verbalize while solving prob- lems. The proctor asked students to verbalize aloud while solving the practice problems to ensure that they understood these instruc- tions. At the start of each of the five subsequent training sessions, the proctor reminded students to verbalize aloud while solving problems. Students’ verbalizations were not continuously monitored dut: ing the sessions (e.g., tape recorded). We felt that such monitoring could prove distracting and thereby alter the nature of the verbal- izations. Two sources of evidence indicated that subjects verbal- ized aloud while solving problems and that their verbalizations focused on the application of regrouping steps to the problems they solved. One source was the periodic proctor monitoring to deliver the attributional feedback. A second source was brief ques- tioning by the proctor at the end of each training session (e.g., “What kinds of things did you say out loud while solving prob- lems?"). ' Students assigned to the discontinued—verbalization condition received the same instructions and postsession questioning as those in the continuous-verbalization condition during the first three training sessions. At the start of the fourth session, the proctor asked these subjects to discontinue overt verbalization as follows: You've been talking out loud while solving problems for qmtc a While. and I‘ve appreciated it because it’s helped me DALE H. SCHUNK AND PAULA D. COX loud. l'm sure that you‘ll be thinking and working just like before, but now please don't talk out loud as you solve prob— lems. At the start of the next two training sessions. the proctor re- minded subjects not to verbalize aloud. The proctor continued to emphasize that while solving problems. students should foll0w the solution steps portrayed on the explanatory page so that stu- dents would not interpret the nonverbalization instructions to mean that they were to abandon the solution strategy. At the end of the- second three training sessions, the proctor questioned subjects about their work (e.g.. “What kinds of things did you think about while solving problems?"). Students assigned to the nooverbalt'zatr'on treatment received the same training procedures as the previous two groups but were ‘3 r v. r l-r-I I In? . . ace-This treatment was comparable to students' regular resource-room mathematics instruction, and no student assigned to this treatment verbalized aloud while solving the practice problems. Prior to students' solving problems on their own, the proctor remarked, “For the rest of this period you'll be working problems on your own. As you work problems, remem~ her to follow the steps shown on this first page." During the periodic monitoring of these students, proctors did not observe any instances of oven verbalization. At the end of each training session, the proctor questioned students about their work (e.g., “What kinds of things did you think about while solving prob- lems‘.’"). All students who participated in this study received periodic monitoring by their proctor while individually solving problems during each of the six training sessions. Each proctor monitored the performance of her students five times (about every 6—7 min) during each of the six training sessions (30 times total) by walking up to each student and asking, “What page are you working on?“ Students then replied with the page number. The attributional treatments were distinguished by the proctor‘s statement following the student’s reply. During the first three training sessions, the proctor remarked, “You‘ve been working hard," to students assigned to the first- half-efi'ort—flerfiack treatment. The proctor delivered this state- ment rather matter-of-factly and without accompanying social re- inforcement (e.g., smiles or pats), after which the proctor immediately departed. During the last three sessions, the proctor did not deliver effort feedback but, instead, acknowledged the student’s reply with performance feedback (e.g., “That's fine.“ or "OK"? and then departed. Performance feedback was deliv- ered during the second half of training to preclude students from interpreting the discontinued effort feedback to mean that they were not performing as well as before, which could have influ- enced self-efficacy and skill development and thereby masked potential effects of the effort feedback (Schunk, 1984). In sum- mary, students assigned to this treatment received 15 statements of effort fwdback spread over the first three training sessions. Subjects assigned to the second-halfiefi'onfeedback treatment received only performance feedback during the first three sessions. During the second half of the training program (Sessions 4—6). 3Students who were having difficulty solving problems could also consult the proctor during the periodic monitoring conducted in conjunction with the attributional feedback. Of the 90 students in the final sample. 10 consulted the proctor at various times learn what students mink about as they solve problems. From during the training program; they were proportionately distributed now on. I'd like you 10‘ solvaroblems~~With6ut talkingouteasse throughout the treatment conditions. STRATEGY TRAINING the proctor delivered effort feedback instead. These students. therefore. also received 15 statements of effort feedback. but the statements were spread over the last three training sessions. Stu— dents assigned to the nowfi’on-feedback treatment received per formance feedback during all six training sessions. The proctor never delivered effort feedback. Posttest. Each student received the posttest from the same tester who had administered his or her pretest. The tester was not aware of the student's treatment assignments for verbalization and effort feedback or of how the student performed during the training program. Tests and training materials were scored by an adult who had not participated in the data collection and who was un- familiar with the purpose of the study. Students' attributions for their problem solving during training were assessed after the last session. The procedures were similar to those Mwapmamrmefiasked-wbjeots to think about their work during the training sessions and mark how much they thought each factor helped them solve problems. Self-efficacy and subtraction skill were assessed on the next day. The instruments and procedures were identical to those of the pretest except that the parallel form of the skill test was used to eliminate possible problem familiarity. Results Means and standard deviations of all measures are pre- sented by treatment condition in Table 1. Preliminary anal- yses of variance were conducted by using two experimental Table 1 205 factors. verbalization (continuous, dlSCOI‘IUnUCd. or none) and effort feedback (first half. second half. or none). These analyses revealed no significant between—conditions differ- ences on any pretest measure or on any subject measure (gender, age, standardized mathematical achievement scores, intelligence scores). There also were no significant differ- ences on any pretest or posttest measure due to tester or school. Self~Eflicacy and Skill Intracondition changes (pretest to posttest) on each mea- sure were evaluated by using the t test for correlated scores “mi-neg»! 971). Each of the three conditions of verbalization ‘ "and 'fliz‘tlir‘e‘e for effort feedback made significant irn‘prove- " ments in both self-efficacy and subtraction skill (all ps < .01 except p < .05 on self-efficacy for the no-feedback condition). Posttest self-efficacy and skill were analyzed with a 3 x 3 (Verbalization X Effort Feedback) multivariate analysis of covariance (MANCOVA), with the corresponding pretest measures as covariates. The MANCOVA yielded significant main effects for verbalization. Wilks's lambda = .642, F(4, 156) = 9.69, p < .001, and effort feedback, Wilks's lambda = .740, F(4, 156) = 6.34, p < .001; the Verbal- ization X Effort Feedback interaction was nonsignificant. Means and Standard Deviation: for All Measures as a Function of Experimental Treatment Verbalization hm“ Continuous Discontinued Effort feedback E‘— First Second None half half None Measure M SD M SD M SD M SD M SD M SD M Sel f-efficacy (average judgment per problem, 10 — 100) Pretest 56.4 28.7 55.3 27.0 Posttest 83.8 15.6 68.5 19.3 Skill (number of correct solutions for 25 problems) Pretest 8.8 6.0 8.3 7.9 Posttest 17.9 4.2 13.2 6.1 Ability (0 - 100) Pretest 57.0 34.3 51.3 29.7 Posttest 72.7 22.3 58.0 31.1 Effort (0 — 100) Pretest 73.3 28.6 73.7 27.7 Posttest 74.0 27.2 73.0 24.8 Task (0 — 100) Pretest 68.7 26.6 72.3 24.9 Posttest 66.7 23.5 72.0 24.3 Luck (0 — 100) Pretest 50.0 28.5 54.0 29.8 Posttest 44.7 30.0 53.3 28.6 Training performance (number of problems completed) 171.1 56.9 25.3 55.0 30.7 57.7 25.9 55.9 24.0 68.1 18.2 78.7 18.2 77.2 18.4 64.4 17.8 8.7 7.3 9.0 7.4 9.1 7.8 7.7 5.8 12.7 5.7 16.3 6.3 15.8 5.5 11.7 4.5 53.7 28.6 54.7 30.1 54.0 32.2 53.3 30.7 58.3 28.2 65.3 29.3 64.7 29.1 59.0 25.9 69.3 26.9 74.7 28.0 73.7 26.2 68.0 28.7 73.3 24.5 87.3 15.5 72.7 23.9 60.3 27.7 71.0 23.0 73.7 24.4 72.0 23 4 66.3 26.2 71.7 22.0 72.0 24.0 68.3 21 0 70.0 24.9 56.0 27.9 55.7 30.3 53.3 27.8 51.0 28.2 51.3 27.4 47.3 28.6 43.0 30.1 59.0 25.4 206 DALE H. SCHUNK AND PAULA D. COX Planned orthogonal comparisons (Kirk. 1982) applied to the posttest self-efficacy measure showed that verbalization conditions led to higher self-efficacy than the no—verbali- zation condition, (80) = 2.46, p < .05; continuous ver- balization led to higher self-efficacy than discontinued verbalization. t(80) = 4.03, p < .01; and providing effort feedback promoted self-efficacy more than not providing feedback, (80) = 4.11, p < .01, MSe = 218.04. On the measure of posttest skill, planned orthogonal comparisons revealed that the verbalization conditions dem- onstrated higher subtraction performance than the no-ver- balization condition, t(80) = 3.37, p < .01; continuous verbalization promoted skill more than discontinued ver- balization, ((80) = 4.81, p < .01; and effort feedback enhanced skillful performance more than = 5.14, p < .01, MS, = 14.30. Attributions Within-condition changes (pretest to posttest) on each attribution revealed a significant increase in effort attribu- tions for the first-half-effort-feedback treatment (p < .05). The four posttest attributions were analde with a MANCOVA having pretest attributions as covariates. This analysis yielded a significant main effect for effort fwdback, Wilks's lambda = .746, F(8, 148) = 2.92,p < .01; both the verbalization _ main effect and the Verbalization X Effort Feedback in- teraction were nonsignificant. Planned comparisons applied to the posttest measure of effort attribution showed that providing effort feedback led to higher effort attributions than not providing feedback, t(80) = 4.15, p < .01; stu- dents who received effort feedback during the first half of training judged effort as a more important cause of success than did subjects who received feedback during the second half, t(80) = 2.68, p < .01, M8,, = 450.05. Training Performance To determine whether treatments differentially affected students’ rate of problem solving during the training ses- sions, the total number of problems that students completed was analyzed with planned orthogonal contrasts. These comparisons showed that verbalization treatments led to higher performance than the no—verbalization condition, ((81) = 2.61, p < .05, and that effort feedback led to more rapid problem solving than did no effort feedback, (81) = 2.74, p < .01, MSc = 1470.01. These differences were not at- tained at the expense of accuracy; identical results were obtained by using the proportion of problems solved cor- rectly (i.e., number solved correctly divided by total num- ber completed). Planned orthogonal comparisons applied to the number of problems that students completed during the first half of training revealed that subjects in the verbalization condi- tions solved problems more rapidly did those in the no- verbalization condition, ((81) = 2.55, p < .05; that effort- feedback treatments enhanced the rate of problem solving, compared with no feedback. ((81) 8 2.29, p < .05; and that first—half effort feedback led to more rapid problem solving than did second-half feedback, t(81) = 2.05, p < .05, MSe = 551.26). The same pattern of results was ob- tained when the proportion of problems solved correctly was used. The number of problems that students completed during the second half of training was also analyzed with planned comparisons, which yielded the following significant re- sults: The verbalization conditions outperformed the no- verbalization condition, (81) = 2.18, p < .05; continuous- verbalization students completed more problems than dis- continued verbalization subjects, ((81) = 2.51, p < .05; and effort-feedback students demonstrated more rapid prob- »7 ranrthes-moefeedbaekrsaidenazgrxsl) = 2.79, p < .01, MSe = 340.31. Identical results were obtained when the proportion of problems solved correctly was used. Correlational Analyses Product-moment correlations were computed among posttest self—efficacy, posttest skill, the four posttest attri- butions, and training performance (number of problems completed). Self-efficacy was positively related to skill, ability and effort attributions, and training performance (ps < .01). Skill was positively correlated with ability and ef- fort attributions and with training performance (ps < .01). The more problems that children completed during training, the higher were their ability (p < .01)' and effort attributions (p < .05), but the lower were their luck attributions (p < .05). Ability attributions and effort attributions were pos- itively related (p < .05). Discussion The present study shows that overt verbalization of the steps of problem solution and their application to problems facilitates task performance, self-efficacy, and skills. These findings are consistent with previous work demonstrating that verbalization often is beneficial for students who typ- ically perform in a deficient manner (Denney & Turner, 1979; Meichenbaum, 1977; Schunk, 1982) and support the idea that private speech can help to regulate task perfonn- ance (Vygotsky, 1962; Zivin, 1979). Learning disabled stu- dents often are inattentive to task instructions and display lackadaisical efforts while working at tasks (Licht, 1984; Torgesen & Licht, 1983). It has been suggested that ver- balization might assist these students to work in a more systematic manner (Hallahan et al., 1983). Although this study shows that overt verbalization is ben- eficial for training students to use a strategy, it does not specify the process by which verbalization promotes achievement outcomes. One possibility is that verbalization helps to focus students‘ attention on important task features and, as a form of rehearsal, assists strategy encoding and retention (Schunk, 1982) It is also possible that verbaliza- tion conveys to students a sense of personal control over learning outcomes because verbalization makes salient a STRATEGY TRAINING 207 strategy that can facrlrtate problem solving (Schunk. 1982). As students effectively use the strategy, they are apt to develop higher self—efficacy for continuing to perform well (Bandura, 1982a; Schunk, l985). Contrary to our prediction, discontinued verbalization did not enhance achievement outcomes as well as continuous verbalization. A lower level of problem solving during the second half of training, relative to that experienced by con- tinuous-verbalization students, should not have promoted self-efficacy or subtraction skills as’well. It is possible that, despite proctor instructions to the contrary, discontinued- verbalization students abandoned the strategic approach to problem solving when instructed to no longer verbalize aloud. They may have had difficulty internalizing the strategy; that cesses conveys that they are developing skills and that they can continue to perform well with hard work (Schunk. 1984). The perception of skill impr0vement can raise self-efficacy and lead to greater skill development (Schunk, 1985). It is somewhat surprising that there was no difference in self-efficacy or skill between the two conditions of effort feedback. We thought that effort feedback for early suc- cesses would be viewed as credible by students but that discontinuing effort feedback would decrease the salience of effort as a cause of success. The perception of less effort for success raises self-efficacy more than greater effort being required does (Bandura, 1982b). Conversely, effort feed- back for later successes might lead students to question their capabilities because they could wonder why they still had is, theymay nothave produced or to succeed (Schunk, 1984). Such self-doubts tions to regulate their performances (Wilder et al., 1984). They also may have believed that although the strategy was useful, other factors (e.g., effort) were more important for solving problems. Children often have naive ideas about when a strategy may be useful (Wellman, 1983). Brown and her colleagues have emphasized that cogni- tive-skill training needs to include the following three com- ponents: instruction and practice in applying a strategy, training in self-regulated implementation and monitoring of strategy use, and information on strategy value and on the range of tasks to which the strategy can be applied (Brown, Campione, & Day, 1981; Brown & Palincsar, 1982; Brown, Palincsar, & Armbruster, 1984). When students receive only the first (skill-training) component, as in the present study, they may not use the strategy on their own because they do not fully understand how and when to apply the strategy or that strategy use greatly improves their performance (Baker & Brown, 1984). Regarding the latter point, explicitly link- ing strategy use with better performance may enhance the effects of strategy training. For example, the trainer could remark after a student correctly solved a problem, “That's correct. You got it right because you applied the steps in the right order." One suggestion for facilitating students' self-regulated strategy use is to have them cognitively transform the strat- egy (Borkowski & Cavanaugh, 1979). Greater cognitive activity can lead to better strategy encoding, retention, and retrieval. A procedure that has been effectively employed to develop self-regulation is self-instructional training, which comprises modeling, guided practice, faded self-guidance (i.e., verbalizations faded to whispers), and covert (silent) self-instruction (Meichenbaum, 1977). There is evidence that this procedure can help sudents with cognitive deficits (e.g., educable mentally retarded, learning disabled, and remedial), who may not make proper use of verbal media- ‘ tors to regulate their task performances (Harris, 1982; John- ston et al., 1980; Whitman & Johnston, 1983; Wilder et a1. , 1984). This study also demonstrates that effort-attributional feedback for students' problem-solving successes led to higher self-efficacy and subtraction skills. As students solve prob- lems, they begin to develop self-efficacy for performing well. Telling them that effort is responsible for their we should not result in high self—efficacy" (§Eliirnk, 1985). One possible explanation for these results is that because students had a learning disability in mathematics, they probably had to expend effort to solve problems throughout the training program. Receiving effort feedback for later successes may have seemed just as credible to these stu- dents as early effort feedback seemed to subjects receiving it during the first half of training. Rather than questioning their capabilities, students who received later effort feed- back may have interpreted it as indicating that they were becoming more skillful. It is interesting that only students in the first-half-effort- feedback treatment showed a significant gain in effort at- tributions, which suggests that early effort feedback served to highlight the role of effort as a cause of success. This finding is noteworthy because researchers have demon- strated that learning disabled students are less likely to at- tribute outcomes to effort than are their nondisabled peers (Butkowsky & Willows, 1980; Licht, 1984; Pearl et al., 1980). Training procedures that help learning disabled stu- dents attribute outcomes to effort have important teaching implications. No Verbalization x Effort Feedback interactions were obtained on any measure. Given the difficulty of the task for the present sample, only limited gains in subtraction skills and self-efficacy may have been possible. Had the study been conducted over a longer period, it is possible that continuous verbalization plus second-half effort feed- back might have led to the largest increases in self-efficacy and skills, assuming that students still needed to expend effort to succeed. The lack of interactions should not imply that verbali- zation and effort feedback are interchangeable procedures. Verbalization is useful for training students to systemati- cally use a task strategy, whereas effort feedback can mo- tivate students to continue working diligently at the task. No amount of effort feedback will promote self-efficacy and skills if students do not understand how to apply a task strategy. Effort feedback is useful as an adjunct to a sound instructional program. To sound a precautionary note, however, we believe that effort feedback for the same task over an extended period is not necessarily desirable, even with learning disabled 208 DALE H. SCHUNK AND PAULA D. COX students. The present task likely engendered a self~focus'. students worked alone and could have compared their pres- ent performance to how they had performed previously. As students become more skillful over time, they ought to solve problems with less perceived effort. In a resource room, students also could compare their performances with those of their peers. Students actually might feel less efficacious if they continually received effort feedback because they might wonder why they had to work hard to succeed when ' their peers demonstrated comparable performance but did not receive effort feedback. Future research needs to explore what effort-attributional feedback means to students. In school, the meaning of at- tributional feedback stems largely from interactions with / teacher'sTeachers often combine effortéwsr'dfiamiset‘ '- r of encouraging learning disabled students to persevere at tasks (e.g., “That’s good. You’re really working hard."). Praise can convey how the teacher views student abilities (Weiner, Graham, Taylor, & Meyer, 1983). Especially when students believe that a task is easy, praise combined with effort information signals low ability. The present results suggest that students did not interpret effort feedback as indicating low ability; first-half-feedback students did not place less emphasis on ability as a cause of success. Effort feedback over an extended period might imply lower ability among learning disabled students if they believed that their skills had improved considerably. Consistent with previous similar research, this study sup- ports the idea that although self-efficacy is influenced by prior performances, it is not merely a reflection of them (Schunk, 1982, 1984). Students who received effort feed— back during the first half of training solved more problems during the first half of training than students who received effort feedback during the second half of training; students in these two conditions did not differ, however, in their self-efficacy judgments. This finding is not surprising. Ef- ficacy appraisal is an inferential process that involves judg- ing the relative contributions of factors such as attributions, amount of external aid received, situational circumstances under which the peformances occurred, and changes in per- formance patterns (Bandura, 1982b; Schunk, 1985). The present results have implications for teaching. Learning disabled students who were deficient in subtraction skills benefited from verbalizing aloud while solving problems and from receiving feedback that linked their successful problem solving with their efforts. Both procedures can easily be implemented in resource rooms. At the same time, the utility of verbalization as a remedial procedure will be enhanced if research demonstrates that verbalizations can effectively be faded to a covert level; many students ver- balizing simultaneously could prove distracting to some. Teachers also need to know how other forms of attributional feedback (e.g., ability) affect students‘ self-efficacy. For example, Schunk (1984) found with children in regular classes that ability feedback for early successes enhanced self-ef- ficacy and skillful performance better than effort feedback did. Understanding how learning disabled students use pri- vate Speech and interpret attributional feedback would have important implications for teaching. References Andrews. G. R., & Debus, R. L. (1978). Persistence and the causal perception of failure: Modifying cognitive attributions. Journal of Educational Psychology, 70, 154—166. Asamow. J. R., & Meichenbaum, D. (I979). Verbal rehearsal and serial recall: The mediational training of kindergmen chil- dren. Child Development, 50, 1173-1177. Baker, L.. & Brown, A. L. (1984). Metaoognitive skills and reading. In P. D. Pearson (Ed.), Handbook of reading research (pp. 353—394). New York: Longman. Bandura. A. (l982a). The self and mechanisms of agency. In J. Suls (Ed.), Psychological perspectives on the self (Vol. I, pp. 3—39). Hillsdale, NJ: Erlbaum. Bandm. A. (1982b). Self—efficacy mechanism in human agency. " ‘ cicanfiggghalogisr, 37h1122:l47,. . A., & Schunk, DH: (l98IV).Q.Cultivating competence, selfiefficacy, and intrinsic interest through proximal self-mo- tivation. Journal of Personality and Social Psychology, 41 , 586—598. Boersma. F. J.. & Chapman, J. W. (1981). Academic self-con- cept, achievement expectations, and locus of control in ele- mentary learning-disabled children. Canadian Journal of Behavioural Science, 13, 349—358. Borkowski, J. G., & Cavanaugh, J. C. (1979). Maintenance and generalization of skills and strategies by the retarded. In N. R. Ellis (Ed.), Handbook of mental deficiency. psychological the- ory and research (2nd ed., pp. 569—617). Hillsdale, NJ: Erl- Brown, A. I... Campione, J. C., & Day, J. D. (1981). Learning to learn: On training students to learn from texts. Educational Researcher, 10(2), 14—20. Brown, A. I.., & Palincsar, A. S. (1982). Inducing strategic leam- ing from texts by means of informed. self-control training. Top- ics in Learning & Learning Disabilities, 2, 1—17. Brown. A. I... Palincsar, A. S., & Armbruster, B. B. (1984). Instructing comprehension-fostering activities in interactive learning situations. In H. Mandl, N. L. Stein, & 'I‘. Trabasso (Eds.), Learning and comprehension of text (pp. 755—286). Hillsdale, NJ: Erlbaum. ‘ Butkowsky, I. S., & Willows, D. M. (1980). Cognitive-motiva- tional characteristics of children varying in reading ability: Evi- dence for learned helplessness in poor readers. Journal of Educational Psychology, 72, 408—422. ‘ Denney, N. W., & Turner, M. C. (1979). Facilitating cognitive performance in children: A comparison of strategy modeling and strategy modeling with oven self-verbalization. Journal of Experimental Child Psychology, 28, ll9~13l. Diener, C. l., & Dweck, C. S. (1980). An analysis of learned helplessness: II. The processing of success. Journal of Person- aliry and Social Psychala , 39, 940—952. Dweck. C. S. (1975). The role of expectations and attributions in the alleviation of learned helplessness. Journal of Personality and Social Psychology. 31. 674—685. Elig, T. W., & Frieze. I. H. (1979). Measuring causal attributions for success and failure. Journal of Personality and Social Psy— chology, 37, 621—634. Friend, 1., & Burton, R. (l98l). Teacher's guide: Diagnostic testing in arithmetic: Subtraction. Palo Alto, CA: Xerox Palo Alto Research Center. Grimm, J. A., Bijou, S. W., & Parsons, J. A. (I973). A problem- solving model for teaching remedial arithmetic to handicapped young children. Journal of Abnormal Child Psychology, 1. 26- 39. u . n. “um.” STRATEGY TRAINING 209 Hallahan. D. P.. Kneedler. R. D.. & Lloyd. .1. W. (1983). Cog- nitive behavior modification techniques for learning disabled children: Self-instruction and self-monitoring. In J. D. Mc- Kinney & L. Feagans (Eds), Current topics in learning disa- bilities (Vol. 1. Pp. 207—244). Norwood, NJ: Ablex. Harris. K. R. (1982). Cognitive-behavior modification: Applica— tion with exceptional children. Focus an Exceptional Children. 15. 1-16. Johnston, M. 13.. Whitman. T. L.. & Johnson. M. (1980). Teach- ing addition and subtraction to mentally retarded children: A self-instructional program. Applied Research in Mental Retar- dation. I. 141-160. Kelley. H. H., & Michela. J. L. (1980). Attribution theory and research. Annual Review of Psychology. 31. 457—501. Kirk, R. E. (1982). Experimental design: Procedures for the be- havioral sciences (2nd ed.). Monterey, CA: Brooks/Cole. Licht. B. G. (1984). Cognitive-motivational factors that contribute to the achievement of leaming-disabled children. In J. K. Tor- gesen & G. M. Senf (Eds). Annual review of learning disa- bilities (Vol. 2. pp. 119—126). New York: Professional Press. Matuyama, G. (1982). How should attributions be measured? A reanalysis of data from Elig and Frieze. American Educational Research Journal. 19. 552—558. Meichenbaum, D. (1977). Cognitive behavior modification. New York: Plenum Press. Pearl. R.. Bryan, T.. & Donahue, M. (1980). Learning disabled children‘s attributions for success and failure. Learning Disa- bility Quarterly, 3. 3-9. Schunk, D. H. (1982). Verbal self-regulation as a facilitator of children's achievement and self-efficacy. Human Learning. 1 265-277. Schunk. D. H. (1984). Sequential attributional feedback and chil- dren's achievement behaviors. Journal of Educational Psy~ chology, 76, 1159—1169. Schunk. D. H. (1985). Self-efficacy and classroom learning. Psy- chology in the Schools, 22, 208-223. v Texas Education Agency ( I983). State Board of Education rules for handicapped students (TEA Document No. A04 7l0 01). Austin. TX: Author. Torgesen. J. K.. & Licht. B. G. (1983). The learning disabled child as an inactive leamer: Retrospect and prospects. In J. D. McKinney & L. Feagans (5:15.). Current topics in learning disabilities (Vol. 1. pp. 3—31). Norwood, NJ: Ablex. Vygotsky. L. S. (1962). Thought and language. Cambridge, MA: MIT Press. ‘ Wechsler. D. (1974). Wechsler Intelligence Scale for Children: Form R. New York: Psychological Corporation. Weiner, B. (1979). A theory of motivation for some classroom experiences. Journal of Educational Psychology. 71, 3—25. Weiner, 8.. Graham. 8., Taylor. S. E.. & Meyer. W. (I983). Social cognition in the classroom. Educational Psychologist. 18, 109-124. , ‘ “M Wellman, H. M. (1983). Metamemory revisted. In M”; r. H. Chi (Ed.). Trends in memory development research (pp. 31—51). Basel, Switzerland: S. Karger. Whitman, T.. 8: Johnston, M. B. (1983). Teaching addition and subtraction with regrouping to educable mentally retarded chil- dren: A group Sclf~instructional training program. Behavior Therapy. 14. 127—143. Wilder. I... Draper, T. W., & Donnelly, C. P. (1984). Overt and covert verbalization in normal and learning disabled children’s problem solving. Perceptual and Motor Skills, 58, 976—978. Winer, B. J. (1971). Statistical principles in experimental design (2nd ed.). New York: McGraw~HilL Woodcock. R., & Johnson. M. (1977). Woodcock-Johnson Psy- cho-Educational Battery. Boston: Teaching Resources. Zivin, G. (1979). The development of self-regulation through pri- vate speech. New York: Wiley. Received May 9, 1985 Revision received January 13, 1986 I Integrating Personality and Social Psychology: Call for Papers The editors of the Journal of Personality and Social Psychology intend to publish a special issue devoted to papers demonstrating that social behavior is best understood by integrating the diverse concerns of the three sections of JPSP. Papers issues crossing the boundaries of individual diff are invited that deal with particular substantive erences, social cognition. and interpersonal rela- tions. Papers should represent practical demonstrations that the diverse concerns of this journal belong together in a full understanding of social behavior. We seek previously unpublished con- tributions. primarily empirical studies, but we are also amenable to syntheses of long-term research programs and to innovative theoretical statements. Contributions intended for the special issue should be sent to the guest editor: John F. Kihlstrom. PhD W. J. Brogden Psychology Building University of Wisconsin 1202 West Johnson Street v Madison, Wisconsin 53706 Interested authors should send an abstract of their article to the guest editor by September 1. 1986 and plan to submit a completed manuscript by December 1. 1986. at which time the paper will bec0me subject to the usual peer review process. Journal of Personality and Social Psychology editors: Attitudes and Social Interpersonal Relations Personality Processes and Cognition and Group Processes Individual Difi‘erences Steven J. Sherman Harry T. Reis lrwin G. Sarason Charles M. Judd Norbert L. Kerr Edward F. Diener Warren H. Jones ...
View Full Document

Page1 / 9

Strategy and attributional learning - Md 0! bearer-ond KM...

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online