Combined strategy and attributional training for poor readers Lorna K.S. Chan Department of Education The University of Newcastle Paper to be presented at the 1993 AARE annual conference, Fremantle, Western Australia, 22-25 November, 1993. Author Note: This paper reports on a research project funded by a 1991 ARC Small Grant. ABSTRACT This paper reports on a study examining the effects of combined strategy and attributional training for poor readers through small-group intervention in a specific reading task context. Four Year 7 classes, consisting of 40 poor readers and 56 average readers, participated in the study. Students were randomly allocated to one of four instruction conditions involving different combinations of strategy instruction and attributional training. Instruction was provided in small groups of 6 to 8 students over nine one-hour sessions. Results confirm that strategy instruction and attributional training were particularly beneficial for poor readers. Combined strategy and attributional training for poor readers In the past few years research in the area of metacognitive instruction has shifted from examining the effects of strategy instruction on task performance to the search for instructional procedures of promoting maintenance and generalisation of the learned strategy (Chan, 1991a, 1991b, in press; Cole & Chan, 1990). One reason suggested for the common finding of lack of strategy generalisation relates to the often observed 'learned helplessness' among students with learning disabilities (Cullen, 1985), that is, the inappropriate belief that the causes of task successes or failures are beyond their personal control. Indeed there are increasing research findings which support a close relationship between metacognitive skills in strategy use, causal attributional patterns (beliefs about the causes of successes and failures) and task performance (e.g., Borkowski, Carr, Relinger & Pressley, 1990; Chan, 1992; Kurtz & Weinert, 1989; Paris & Winograd, 1990; Schneider, Borkowski, Kurtz & Kerwin, 1986; Schneider, Korkel & Weinert, 1987). It has been proposed that, to promote maintenance and generalisation of strategy use, strategy instruction should be supplemented by appropriate attributional training (e.g., Billingsley & Wildman, 1990; Borkowski, Weyhing & Carr, 1988; Chan, 1991a, 1991b, in press; Chan, Cole & Morris, 1990; Grolnick & Ryan, 1990; Licht & Kistner, 1986). One of the aims of the proposed study is to test this hypothesis that combining strategy training with attributional training is more facilitative of maintenance and generalisation of strategy use than strategy training alone. Past attributional training programs tend to emphasise effort attributions (see review by Licht & Kistner, 1986). These have not always been successful with students with learning difficulties because very often these students find themselves keep on failing in spite of increased effort (particularly if they do not know how to try harder). It has been maintained that a more fruitful approach may be to train these students to attribute failures to ineffective task strategy rather than to insufficient effort so as to avoid increasing feelings of helplessness (Jacobsen, Lowery & Ducette, 1986; Licht & Kistner, 1986). Further, more recently Fabricius & Cavalier (1989) provided evidence to suggest that children need to be metacognitively aware of the mechanisms by which strategies achieve their effects, and not just knowledge that a strategy can be effective, before generalisation of strategy use can be expected. This could be achieved by attributional training with regards to strategy use. But as yet little research has been done in the development of causal attributions in strategy use. A previous study by the author (Chan, 1992) indicated that while strategy attributions and strategy use were significant factors in mediating learning outcomes, they were not particularly prominent for students in primary and lower secondary classes nor for students with learning difficulties. This highlights the need for attributional training and strategy instruction for these group of students. The three pioneering studies on combined strategy instruction and attributional training either focused solely on effort attributions (Short & Ryan, 1984) or failed to distinguish effort attributions from strategy attributions (Borkowski, Weyhing & Carr, 1988; Reid & Borkowski, 1987). In the Borkowski, Weyhing & Carr (1988) study, for instance, subjects were taught a clustering-rehearsal strategy for a sort-recall task and to attribute their success to "I tried hard and used the strategy". Only the importance of "making an effort" and "trying hard" was emphasised. Further, in the causal attribution assessment subjects were asked whether they thought their success was "because you tried hard" without any reference to strategy use. Data from the author's recent study (Chan, 1992) however, clearly indicate that effort attributions are quite distinct from strategy attributions. Another issue in combined strategy and attributional training relates to the most efficient way of combining the two. Most attributional training programs employ general memory tasks that are not directly related to classroom learning. For example, subjects may be required to memorise a list of objects depicted in picture cards and the appropriate strategy to be taught may be one of grouping the pictures into several categories and then rehearsing the objects in each category (clustering-rehearsal strategy). This kind of recall task requires only the teaching of a simple, easy-to-learn strategy and the effects of strategy use can be immediately and clearly demonstrated, thus readily lending itself to attributional training. However, the learning of most classroom tasks (such as reading) requires strategies that may involve more learning time and practice before any benefits will become obvious. The self-questioning strategy (Chan, 1991b; Cole & Chan, 1990; Wong, 1985), for example, involves asking oneself questions during reading such as "What is the paragraph mainly about?, "Is there anything I am not sure about?", "How does it link with the previous paragraphs?", "Does it make sense?", "What is likely to come next?", thus requiring extensive time for learning. In this context a question needs to be asked: Can the goal be achieved by providing attributional training simultaneously with reading strategy training? Or is it necessary to provide attributional training in the context of a non-reading task (thus allowing strategy attributions to be established to some extent) before the combined strategy and attributional training on the reading task? To summarise, the present study aimed to examine the effects of combined strategy and attributional training through small-group intervention in a specific reading task context. In particular, the study was designed to test whether such combined training is more facilitative of maintenance and generalisation than strategy training alone and to identify the most effective way of combining the two types of training. METHOD A Subject-group x Instruction Condition x Testing Occasion repeated measures design was employed, with Testing Occasion being the within- subjects factor. Subjects Four Year 7 classes from a regional high school in New South Wales, Australia participated in the study. The four classes spanned the whole range of ability levels. However, the school is situated in a relatively low SES area, hence there is a relatively higher proportion of poor readers. Students were classified into a group of poor readers and another group of average readers on the basis of their performance on the Reading Comprehension subtest of the Stanford Diagnostic Reading Test, Level II administered at the beginning of the school year. Students scoring in the bottom three stanines (below the 23rd percentile) were allocated to the poor readers group (n=40) and the remainder to the average readers group (N=56). Instruction Condition Students from each class were randomly assigned to one of four instruction conditions, as follows: 1. reading strategy + successive attributional training (SUAT) - subjects were given attributional training in strategy use on a sort- recall task (Phase 1) before combined strategy and attributional training on a reading task (Phase 2). 2. reading strategy + simultaneous attributional training (SIAT) - subjects were given only the combined strategy and attributional training on the reading task (Phase 2). In Phase 1 they just completed the sort-recall task without the attributional training. 3. attributional training only (AT) - subjects were given only the attributional training on the sort-recall task (Phase 1). In Phase 2 they just completed the reading task without the combined strategy and attributional training. 4. strategy training only (ST) - subjects were given only strategy training in both phases without any attributional training. A schematic representation of the common and distinguishing elements among the four instruction conditions is presented in Table 1. Instruction Program Instruction was provided in small groups of 6 to 8 students over nine one-hour sessions, 1 session for Phase 1 and 8 for Phase 2. Self- instructional training (Chan, 1991b; Cole & Chan, 1990) was employed for teaching the self-questioning strategy. Self-instructional training is a metacognitive training technique that involves 'thinking aloud' modelling of strategy use, external- and self-guided practice, and fading procedures to promote internalisation of strategy use. In each of the eight sessions in Phase 2, an expository passage of about 350 to 450 words in length and written at a Grade 6 readability level was used for the reading task. Testing Occasions Measures of reading comprehension performance were taken at the beginning (pretest) as well as the completion (posttest) of Phase 2, and four weeks after (maintenance test). All subjects were also administered the Causal Attributions and Reading Strategies scales both before and after the instruction program. Dependent Measures Measures of reading comprehension competence, knowledge and reported use of reading strategies, and attributional beliefs were employed. Reading comprehension was assessed using reading passages similar to those used for the instruction program, each followed by six questions requiring short answers. Three parallel forms were employed for the pretest, posttest and maintenance test measures. Knowledge and use of reading strategies were assessed using the Reading Strategies Scale (Chan, 1922). There are 20 items, each describing a student using a particular strategy for reading, such as "summarising the main ideas in her own words as she reads". Negative items describing ineffective strategies are also included. After each description, students are required to rate the strategy on two separate four-point scales in terms of firstly, how helpful they consider that strategy to be; and secondly, how often they read that way. Ratings on the two questions were summed separately, providing a 'knowledge' score and a 'reported usage" score, each ranging from 20 to 80, with a higher score indicating greater knowledge or usage. Reliability estimates on the two subscale scores for Year 7 students are 0.86 on the knowledge scale and 0.85 on the usage scale. Attributional beliefs are assessed using the Causal Attribution Scale (Chan, 1992). This is a ten-item scale designed by the researcher to assess students' tendency of attributing their school success and failure expereinces to the four likely reasons of effort, ability, strategy use and luck. Five items describe success incidents (such as doing well on a test) and the other five, failure incidents. For each item, four different reasons are listed and students are required to rate each on a four-point scale to indicate how true they consider that particular reason to be for them. The ratings for each of the four attributions (effort, ability, strategy use and luck) are summed across the five success items and the five failure items respectively, thus giving eight separate sub-scores, each ranging from 5 to 20. For example, a high score on the Failure-ability subscale indicates greater tendency of attributing school failure experiences to lack of ability. Cronbach's coefficient alphas computed on the eight subscales for Year 7 ranged from 0.63 to 0.83. RESULTS AND DISCUSSION The means and standard deviations of the reading comprehension competence, causal attributions and reading strategies scales scores are presented in Tables 2 and 3. All statistical analyses were conducted on SPSSx, Release 3.1. Reading Comprehension Competence Data were analysed using a Subject-group (2) x Instruction Condition (4) x Testing Occasion (3) repeated measures ANOVA. Three simple contrasts were set up for the Instruction Condition factor, comparing the SUAT with each of the other three conditions. Helmert contrasts were used for the Testing Occasion factor, one comparing the pretest with the posttest and maintenance test combined, and the other comparing the posttest with the maintenance test. Results indicated a significant three-way interaction, which was located in the interaction of Subject-group x SUAT versus AT contrast x pretest versus combined posttest and maintenance test contrast, F (1,87) = 4.11, p<0.05. As depicted in Figure 1, a differential pattern of the effects of the various instruction conditions for the poor and average readers was observed. For the poor readers, higher levels of performance on the posttest and maintenance test relative to the pretest was demonstrated by students receiving reading strategy instruction (SUAT, SIAT, ST) compared to those not receiving instruction on reading strategy (AT - attribution training only). For the average readers, on the other hand, only students receiving combined strategy and attributional training (SUAT and SIAT) improved on the posttest, but the improved performance was not maintained. The findings suggest that a combination of reading strategy instruction with successive attributional training (SUAT) seems the most effective in improving the poor readers' comprehension performance on the posttest, though such improvement was somewhat attenuated on the maintenance test. For longer term maintenance, all three instruction conditions involving the teaching of reading strategies were effective in enhancing the reading comprehension performance of the poor readers. Knowledge and Use of Reading Strategies Data on the two measures were analysed using a Subject-group (2) x Instruction Condition (4) x Testing Occasion (2) repeated measures MANOVA. The multivariate results indicated a significant three-way interaction, F(6,156)=2.76, p<.01, which was located only in the reported usage measure, F(3,79)=3.86, p<0.02. This significant three-way interaction for the usage measure was focused on the interaction involving the SUAT versus AT contrast, F(1,79)=10.16, p<0.01. Examination of Figure 2 revealed that apart from the poor readers receiving attributional training only (AT), all subjects (both poor and average readers) in the other instruction conditions reported greater use of reading strategies after the instruction program. Apart from subjects in the AT instruction condition, all the other subjects received training in the use of the self-questioning strategy for reading. The findings suggest that the reading strategy instruction has resulted in increased reported use of reading strategies. Subjects receiving attributional training only were taught the clustering-rehearsal strategy to improve item recall. Learning of this non-reading strategy seems to have facilitated increased reported use of reading strategies among the average readers, but not among the poor readers. The multivariate statistics were also significant for the occasion main effect, F(2,78)=23.99, p<.001. Univariate results revealed that the occasion main effect was significant for both the knowledge and reported usage measures, F(1,79)=47.62, p<.001 and F(1,79)=14.36,p<.001, respectively, indicating that all subjects, regardless of subject-group or instruction condition, had greater knowledge and reported greater usage of reading strategies after than before the intervention. Attributional Beliefs Data on the four success and four failure attribution measures were analysed using two separate Subject-group (2) x Instruction Condition (4) x Testing Occasion (2) repeated measures MANOVAs. Multivariate results for the four success attributions indicated that only the subject-group main effect was significant, F(4,84)=13.34, p<.001. For failure attributions, significant multivariate statistics were obtained for the Instruction Condition x Testing Occasion interaction, F(12,245)=1.77, p<.05, and the subject-group main effect, F(4,83)=8.12, p<.001. Univariate statistics revealed that the significant Instruction Condition x Testing Occasion interaction on the failure attributions was focused on the strategy and ability measures, F(3,86)=2.74, p<.05 and F(3,862.71, p<.05, respectively. The significant interaction was located in the SUAT versus AT contrast x Testing Occasion interaction for the strategy attribution measure, F(1, 86) = 6.96, p<0.01; but in the SUAT versus SIAT contrast x Testing occasion interaction for the ability attribution measure, F(1,86) = 5.62, p<0.02. These two significant interaction effects are graphed in Figure 3. These results indicate that regardless of subject- group, only the AT condition was effective in increasing students' likelihood of attributing failure to use of ineffective strategies and the SIAT condition was particularly effective in reducing students' likelihood of attributing failure to lack of ability. Regarding the subject-group main effects, the associated univariate F statistics were significant for the effort [F(1,87) = 32.33, p<0.001], ability [F(1,87) = 5.20, p<0.03], and luck [F(1,87)=22.87, p<.001] attributions for success; and for the ability [F(1,86) = 24.06, p<0.02] and luck [F(1,86) = 7.74, p<0.01] attributions for failure. Examination of the means indicate that regardless of instruction condition and testing occasion, the poor readers were less likely than the average readers to attribute success to effort or ability but were more likely to attribute success to luck, or attribute failure to lack of ability or bad luck, a causal attribution pattern typical of the learned helplessness phonomenon. Conclusion Several key findings emerged from this study. First, in terms of effects on reading comprehension, a combination of reading strategy instruction with successive attributional training (SUAT) seems the most effective in improving the poor readers' comprehension performance on the posttest, though such improvement was somewhat attenuated on the maintenance test. For longer term maintenance, all three instruction conditions involving the teaching of reading strategies were effective in enhancing the reading comprehension performance of the poor readers. Secondly, regarding effects on knowledge and reported use of reading strategies, all subjects, regardless of subject-group or instruction condition, had greater knowledge of reading strategies after than before the intervention. Further, apart from the poor readers receiving attributional training only, all subjects (both poor and average readers) in the other instruction conditions reported greater use of reading strategies after the instruction program. Thirdly, findings on causal attributions reveal that regardless of instruction condition and testing occasion, the poor readers were less likely than the average readers to attribute success to effort or ability but were more likely to attribute success to luck or to attribute failure to lack of ability or bad luck, a causal attribution pattern typical of the learned helplessness phonomenon. Further, regardless of subject-group, only the AT condition was effective in increasing students' likelihood of attributing failure to use of ineffective strategies and the SIAT condition was particularly effective in reducing students' likelihood of attributing failure to lack of ability. These findings taken together confirm that strategy instruction and attributional training were particularly beneficial for poor readers. All three instruction conditions involving strategy instruction were effective in enhancing poor readers' reading comprehension performance and reported use of reading strategies. Further, combining strategy instruction and attributional training would reduce students' likelihood of attributing failure to lack of ability, but increase students' likelihood of attributing failure to use of ineffective strategies. Attributing failures to use of ineffective strategies (that can be controlled by the student) but not to lack of ability (which is beyond our personal control) would imply that future failures could be avoided with use of more effective strategies, thus improving perceptions of personal control over learning outcomes and reducing the learned helplessness tendency. It is more beneficial to convince poor learners that their failures are attributable to use of ineffective strategies rather than to insufficient effort because very often increased effort by poor learners may not necessarily result in success unless they know how to work harder. Indeed, repeated failure in spite of increased effort would only further reinforce beliefs of lack of ability, thus increasing feelings of learned helplessness. The findings of the present study demonstrate that teaching poor readers use of effective reading strategies while convincing them that reading successes and failures are attributable to use of effective or ineffective strategies not only improves their comprehension performance and increases use of reading strategies, but also reduces their perceptions of learned helplessness. Finally, overall findings suggest that the most efficient way of combining attributional training with reading strategy instruction is to provide attributional training in the context of a non-reading task (thus allowing strategy attributions to be established to some extent) before the combined strategy and attributional training on the reading task. Most reading strategies require sufficient time for learning and practice before their benefits in enhanced performance can be demonstrated. Attributional training could be facilitated with initial teaching of simple and easy-to- learn strategies such that their effects on performance can be more easily and convincingly demonstrated. This will allow strategy attributions to be established to some extent before continuing the combined strategy and attributional training on the reading task. REFERENCES Billingsley, B.S. and Wildman, T.M. (1990). Facilitating Reading Comprehension in Learning Disabled Students: Metacognitive Goals and Instructional Strategies. Remedial and Special Education, 11(2), 18- 31. Borkowski, J.G., Carr, M., Rellinger, L. & Pressley, M. (1990). Self- regulated cognition: Interdependence of metacognition, attributions and self-esteem. In B.J. Jones & L.Idol(Eds.). Dimensions of thinking and cognitive instruction (p.53-92). Hillsdale:Lawrence Erlbaum. Borkowski, J.G., Weyhing, R.S. and Carr, M. (1988). Effects of Attributional Retraining on Strategy-Based Reading Comprehension in Learning-Disabled Students. Journal of Educational Psychology, 80(1), 46-53. Chan, K.S. (1991a). Metacognition and remedial education. Australian Journal of Remedial Education, 23(1), 4-10. Chan, K.S. (1991b). Promoting strategy generalisation through self- instructional training in students with reading disabilities. Journal of Learning Disabilities, 24(7), 427-433. Chan, K.S. (1992). Causal attributions, strategy usage and reading competence. Paper presented at the AARE/NZARE Joint Conference, Deakin University, Geelong, Nov., 1992. Chan, K.S. (in press). Remedial education: Research and practice. Australian Journal of Remedial Education. Chan, K.S., Cole, P.G. & Morris, J.N. (1990). Effects of instruction in the use of a visual imagery strategy on the reading comprehension competence of disabled and average readers. Learning Disability Quarterly, 13(1), 2-11. Cole, P.G. & Chan, K.S. (1990). Teaching methods and techniques in special education. Sydney: Prentice-Hall of Australia. Cullen, J.L. (1985). Children's ability to cope with failure: Implications of a metacognitive approach for the classroom. In D.L. Forrest- Pressley et al. (eds). Metacognition, cognition and human performance, Vol.2 : Instructional practices. Orlando: Academic Press. Fabricius, W.V. and Cavalier, L. (1989). The Role of Causal Theories about Memory in Young Children's Memory Strategy Choice. Child Development, 60, 298-308. Grolnick, W.S. and Ryan, R.M. (1990). Self-Perceptions, Motivation, and Adjustment in Children with Learning Disabilities: A Multiple Group Comparison Study. Journal of Learning Disabilities, 23(3), 177-184. Jacobsen, B., Lowery, B. and DuCette, J. (1986). Attributions of Learning Disabled Children. Journal of Educational Psychology, 78(1), 59-64. Kurtz, B.E. and Weinert, F.E. (1989). Metamemory, Memory Performance, and Causal Attributions in Gifted and Average Children. Journal of Experimental Child Psychology, 48, 45-61. Licht, B.G. & Kistner, J.A. (1986). Motivational problems of LD children: Individual differences and their implications for treatment. In J.K. Torgesen & B.Y.L. Wong (Eds.), Psychological and educational perspectives on learning disabilities. Orlando: Academic Press. Paris, S.G. & Winograd, P. (1990). How metacognition can promote learning and instruction. In B.J. Jones & L.Idol(Eds.). Dimensions of thinking and cognitive instruction (p.15-51). Hillsdale:Lawrence Erlbaum. Reid, M.K. and Borkowski, J.G. (1987). Causal Attributions of Hyperactive Children: Implications for Teaching Strategies and Self-Control. Journal of Educational Psychology, 79(3), 296-307. Schneider, W., Borkowski, J.G., Kurtz, B.E. and Kerwin, K. (1986). Metamemory and Motivaton: A Comparison of Strategy Use and Performance in German and American Children. Journal of Cross- Cultural Psychology, 17(3), 315-336. Schneider, W., Korkel, J. and Weinert, F.E. (1987). The Effects of Intelligence, Self-Concept and Attributional Style on Metamemory and Memory Behaviour. International Journal of Behavioral Development, 10(3), 281-299. Short, E.J. and Ryan, E.B. (1984). Metacognitive Differences Between Skilled and Less Skilled Readers: Remediating Deficits Through Story Grammar and Attribution Training. Journal of Educational Psychology, 76(2), 225-235. Wong, B. (1985). Self-questioning instructional research: A review. Review of Educational Research, 55, 227-268. Table 1. The Four Instruction Conditions. SUAT SIAT AT ST Phase 1 (1 one-hour session) sort-recall task X X X X clustering-rehearsal strategy X X X attributional training X X Phase 2 (8 one-hour sessions) reading task X X X X self-questioning strategy X X X attributional training X X Table 2. Means and Standard Deviation of the Reading Comprehension Pretest, Posttest and Maintenance Test Scores N Pretest Posttest Maintenance Test Mean SD Mean SD Mean SD Poor Readers SUAT SIAT AT ST 11 8 11 9 4.18 4.38 4.55 4.44 2.60 2.50 3.27 2.92 8.18 4.88 4.09 4.56 1.72 3.52 3.11 2.01 6.64 5.13 3.91 6.00 3.11 2.48 1.97 2.35 Average Readers SUAT SIAT AT ST 18 11 12 15 7.94 7.91 7.25 8.20 2.78 2.12 2.53 1.97 9.56 9.73 7.33 8.80 2.33 1.62 3.75 2.01 7.33 6.64 6.83 8.07 1.75 1.91 1.80 2.58 Table 3. Means and Standard Deviation of the Reading Strategies and Causal Attribution Scale Scores Instruction Condition Poor Readers Average Readers Pretest Posttest Pretest Posttest Mean SD Mean SD Mean SD Mean SD Knowledge of reading strategies SUAT SIAT AT ST 52.00 57.33 53.36 52.75 4.97 11.31 4.25 9.35 60.57 64.17 55.82 57.63 5.06 5.35 16.24 4.69 56.88 51.00 54.67 54.07 8.60 9.86 8.34 9.37 65.18 63.00 62.08 62.80 7.35 6.86 8.79 7.61 Reported usage of reading strategies SUAT SIAT AT ST 45.14 51.33 46.82 44.50 5.98 12.23 5.27 5.04 51.71 56.50 38.36 49.50 7.14 9.50 12.93 7.76 49.71 45.09 44.42 46.93 8.32 5.30 9.27 7.81 54.47 50.82 52.83 54.13 10.17 8.29 9.45 7.90 Effort attributions for success SUAT SIAT AT ST 14.20 13.75 13.91 13.38 2.78 4.23 3.96 4.57 14.50 14.88 14.36 12.13 3.75 3.98 4.15 4.67 17.39 17.27 17.39 16.88 1.88 1.42 2.47 2.58 16.50 16.91 17.39 17.25 2.85 2.26 2.59 1.81 Strategy attributions for success SUAT SIAT AT ST 13.40 13.63 11.73 12.63 2.55 3.54 2.83 2.56 14.10 14.88 12.64 11.13 3.90 2.23 3.75 4.36 14.06 14.00 12.77 14.31 2.34 3.89 3.70 3.07 13.00 14.46 14.00 15.13 2.93 3.78 2.83 2.36 Ability attributions for success SUAT SIAT AT ST 13.50 13.00 11.46 11.13 2.76 2.14 2.46 1.81 13.20 14.13 12.09 11.38 3.62 2.99 2.34 1.85 14.22 13.55 13.15 13.94 2.88 4.48 3.21 3.30 13.56 14.09 13.15 13.94 2.38 2.43 3.72 2.79 Luck attributions for success SUAT SIAT AT ST 11.00 10.88 11.09 10.50 2.87 1.96 2.51 1.60 9.50 11.63 10.55 10.88 2.01 3.34 2.25 2.03 8.06 9.91 8.77 8.13 1.55 3.94 2.71 1.96 8.50 8.46 8.77 9.00 2.48 2.54 1.96 2.89 Effort attributions for failure SUAT SIAT AT ST 10.67 11.75 10.09 10.63 3.12 3.92 2.12 2.39 9.89 11.50 11.09 9.88 3.95 4.28 2.17 2.69 12.22 8.18 11.85 11.13 4.18 3.19 3.98 4.27 12.44 9.27 11.85 10.88 3.85 3.49 3.94 3.24 Strategy attributions for failure SUAT SIAT AT ST 9.78 11.88 10.27 11.75 2.44 2.90 2.20 2.77 9.78 11.25 12.00 10.50 2.11 3.41 2.68 3.12 10.44 8.91 9.39 9.50 2.62 1.97 2.29 2.07 10.50 9.18 11.46 10.69 2.77 2.75 2.93 2.92 Ability attributions for failure SUAT SIAT AT ST 9.56 11.25 10.91 10.50 3.32 2.61 2.02 3.07 8.22 8.13 11.00 10.13 3.27 2.75 2.05 1.81 7.11 8.18 6.77 7.69 2.59 3.66 2.17 2.75 6.78 7.64 7.08 8.75 2.42 2.87 1.98 3.45 Luck attributions for failure SUAT SIAT AT ST 10.33 13.00 9.55 10.63 3.24 3.70 1.86 3.54 10.44 10.00 10.64 10.38 4.07 3.34 1.86 2.77 10.06 8.36 8.46 9.69 3.72 4.01 2.26 3.63 9.17 7.91 8.23 10.31 3.35 1.81 3.19 4.41