Anne Berger, Peter H. Ludwig University of Koblenz-Landau, Germany, Department of Educational Science The influence of teaching and student characteristics on writing papers and their achievement results of self-regulated learning Paper presented at the European Conference on Educational Research (ECER) From Teaching to Learning, organized by the European Educational Research Association (EERA) at the University of Gothenburg, Sweden, Network 27 Didactics learning and teaching, September 11th, 2008. One of the major goals of education is to enable students to acquire knowledge independently; to learn and to work self-regulated. At the latest when leaving high school and entering university, it is expected from students to know how to deal with self-directed learning environments. High school students distinguish each other between their selfregulation competence. Some need more support from teachers than others. Therefore, it is important for teachers to know how they can manage to do that. In the German Gymnasium, a high-achievement school type at the upper secondary level, comparable to the British grammar school or the French lycée, students are given a special complex task in grade 12 to develop their self-regulation skills. They have to write a paper of 10 to 20 pages (called Facharbeit ) over a period of six to ten weeks. These papers are worked out mostly at home and demand self-regulated working behavior at a high level. This task involves finding a topic to write about, planning and organising this work independently as well as self-managing (e.g. time, self-motivation, dealing with problems or coping with frustrating moments) and applying the rules of scientific work. Sometimes this includes conducting small experiments or using other empirical procedures. The goal of this assignment is to familiarize students with the principles and forms of independent and scientific working as a preparation for their subsequent college or university studies. These papers substitute a written examination in this form level. They are relevant for the general qualification for university entrance. The purpose of the study presented here is to contribute to the quality of teachers support for students self-regulated work on papers. Regarding the research questions it is intended to deliver evidence about the influencing factors of students independent work on a complex learning task. Further on, the relationship between strategic proceeding and the quality of the performance is examined as well as the impact of different motivational traits on the working process of a learning task. This project is part of the research program of the Graduate
School of Learning and Teaching Processes (UpGradE), founded by Andreas Helmke and Wolfgang Schnotz at the University of Koblenz-Landau, Germany. Theoretical background Most of the definitions of self-regulated learning have in common that learners have to initiate, organise and control their own learning behavior to be successful. Therefore, selfregulated processes operating on the cognitive, the metacognitive, and the motivational/volitional layer of the three-layer-model of self-regulated learning by Boekaerts (1999; Boekaerts et al. 2000) are constitutive (cf. Baumert et al. 1999). This field study is mainly based on the theoretical frame-model of external-regulated and self-regulated learning of Schiefele & Pekrun (1996; see figure). learner variables metacognitive knowledge skills, previous knowledge motivational orientations & beliefs metacognition: planning preparative resource mangement motivation internal regulation before learning during learning after learning metacognition: monitoring regulation cognition learning strategies resource management metacognition checking learning results self-evaluation motivation habitual volitional variables volition volition learning process: planning performance evaluation e.g. teacher behavior, enhancement, teaching methods, learning enviroment, exams, school organisation... external regulation learning outcome quality of the learning product Figure: The theoretical model of external- and self-regulated learning (Schiefele/Pekrun 1996) There is a broad variety of actual and habitual learner variables which influence the actual internal regulation of a specific working process (e.g. writing a paper). This process consists of three subsequent steps: before, during and after working. Each step involves different actions on the cognitive, the metacognitive, the motivational, and the volitional layer. This model displays just a rough simplified image of the real processes. In real educational 2
settings the internal features and activities of these steps probably influence, interact and merge into each other. Apart from the learner variables the internal regulation during task processing is also influenced by external regulation, e.g. the teachers behavior or the environmental conditions. The model postulates that all these aspects affect the quality of the learning outcome. Central research questions This study concentrates on the following central research questions: (1) Which strategies on the cognitive, metacognitive, and motivational/volitional layer are used by students while working on the paper? Does students behavior on the cognitive, metacognitive and motivational/volitional layer have an effect on the working result in terms of the quality of the paper? Which factors improve or decrease the quality of the paper? (2) Which teachers behavior (preparing for and supporting students in the work process) has an effect on the quality of the paper? (3) Do dispositional student variables (e.g. motivational orientations and general beliefs, volitional traits) have an effect on the quality of the paper? Design The research design and the content of the questionnaires is based on the model of Schiefele and Pekrun (1996, mentioned earlier in this article). Three student questionnaires at three measuring points were used. The first questionnaire was completed by the students before they started with the paper. It refers to goal-orientation, academic self-concept, outcome-anticipation regarding the paper, habitual volitional aspects, perceived parental educational behavior and educational background as well as the learning climate in the class. The second questionnaire was given after having finished and presented the papers, with a maximum of 14 days after that. This questionnaire includes a reflection of the working process in terms of cognitive, metacognitive and volitional aspects as well as actual motivational variables concerning the task. Furthermore, students were asked to assess the quality of school preparation and teacher support they received during the working process on the paper. The third measure took place after students received the grade for the paper. The students answered questions about the received grade and the perceived teacher feedback concerning the paper. 3
Transfer of research results to the schools Specific attention is paid to the transfer of the results of this study to the educational practice in the participating schools. The schools have been given the possibility to add schoolspecific questions to the questionnaires concerning the papers. All schools obtained an individual anonymized feedback about their own results of the study. The aim of this feedback is to support the schools in developing teaching processes concerning the focussed paper. Total sample The total sample comprises 10 schools ( Gymnasien ). Mostly all students of grade 12 of these schools participated in the study. 845 students completed the first questionnaire. In the second measure this sample was reduced to 785 students and at the third time to 638 students 1. The average age of the students in the first measure was 17.6 years (SD=0.75). As typical for Gymnasium there are slightly more girls in the sample (1 st measure: female: 55.4%; 2 nd measure: female: 56.3%). Selected research question The analysing work of the data is still in progress. For the present article the results of the teachers support and school preparation during the work process on the paper can be presented. Therefore, the following differential research question will be answered: Does the quality of the teacher support (as perceived by students) and the quality of school preparation have an effect on the quality of the paper in consideration of the student variables interest and fun on the paper, importance of the paper and the achievement level? These measures are not represented by scales but single items. Table 1 gives an overview over the focussed items. Positive effects of these predictor variables on the quality of the paper are expected. The quality of the paper is operationalized by the teacher assessment: the received grade for the paper. 1 This is not the final sample of the third measure; there are still data of two schools awaited. The study is still in progress. 4
Table 1: Item verbalization, answering categories, mean and standard deviation (SD) of the predictor variables. aspect item verbalization answering categories Mean SD quality of school preparation teacher support number of counseling interviews quality of counseling interviews quality of interim feedback information about quality of interim results quality of teacher support interest and fun fun at work interest at the topic importance personal importance importance of the grade general importance achievement level Overall, how well did you feel prepared for the paper by your school? How many counseling interviews did you have with your teacher? Overall, how helpful did you perceive the counselling interviews with your teacher for the work on your paper? With the feedback to my interim results I could... After the counseling interviews I knew how good my interim results were. Overall, I felt (...) by my teacher during the work on the paper Overall it was fun for me to work on the Facharbeit. How interested were you in the topic of your paper? How important was your paper to you? How important was the grade of the paper to you? I think it is important to write a paper in school. grade average of the last two exams in the subject of the paper (1) very good (2) rather good (3) not very good (4) not good at all 2.54.81 number 2.16 1.47 (1) very helpful (2) rather helpful (3) little helpful (4) not helpful (1) mostly work with (2) sometimes work with (3) mostly not work with (1) always (2) sometimes (3) never (1) very well supported (2) moderately supported (3) mostly neglected (1) true (2) rather true (3) hardly true (4) not true (1) very interested (2) rather interested (3) hardly interested (4) not interested (1) very important (2) rather important (3) not very important (4) not important at all (1) very important (2) rather important (3) not very important (4) not important at all (1) true (2) rather true (3) hardly true (4) not true German six-step number grade system, one is the best grade 2.25.89 1.63.74 2.01.70 1.94.74 2.55.89 1.62.76 1.94.79 1.69.76 2.35.91 2.89 1.03 5
Description of the actual reduced sample The provisional analysed sample comprises 473 students (with 57.7 % females) of eight schools. The average student age is at the end of 17 (1st measure: 17.6; 2nd measure: 17.9). Some schools have not returned their questionnaires yet. Results Multiple regression analyses have been calculated including all 12 variables (see table 1) to predict the quality of the paper (grade), using the method "stepwise" 2. Only four of the 12 predictors have turned out to add significantly to the explanation of the grade differences. The other eight predictors do not show effects on statistically significant levels. Four regression models have been calculated with the significant predictors (table 2). Model 4 with four predictors has the highest explanation power with an adjusted multiple determination coefficient of R-square=.294 (p<0.001; see table 3). I.e.: The predictor variables achievement level, general importance, fun at work, and quality of teacher support explain together 29.4% of the variance in the grades of the papers. Table 2: Model summary and coefficients of the multiple regression analyses Model Summary e Mode l 1 2 3 4 R R Square Adjusted R Square,454 a,206,204 1,000,206 84,163 1 324,000,522 b,272,268,959,066 29,357 1 323,000,540 c,292,285,948,020 8,872 1 322,003,550 d,303,294,942,011 4,937 1 321,027 1,992 a. Predictors: (Constant), achievement level Std. Error of the Estimate b. Predictors: (Constant), achievement level, general importance c. Predictors: (Constant), achievement level, general importance, fun at work Change Statistics R Square Change F Change df1 df2 d. Predictors: (Constant), achievement level, general importance, fun at work, quality of teacher support e. Dependent Variable: grade of the "Facharbeit" Sig. F Change Durbin- Watson 2 stepping method criteria: F-probability (entry:.05; removal:.10) 6
Table 3: Coefficients of model 4 Mode l 4 (Constant) achievement level general importance fun at work quality of teacher support Unstandardized Coefficients B Std. Error Coefficients a Standardized Coefficients Beta -,161,226 -,715,475 Collinearity Statistics t Sig. Tolerance VIF,428,053,389 8,149,000,953 1,049,214,067,174 3,198,002,732 1,366,183,070,145 2,604,010,701 1,426,168,075,107 2,222,027,941 1,062 a. Dependent Variable: grade of the "Facharbeit" The achievement level turns out to be the most influential measure of the grades. It explains 20.4% of the variance of the grades (derived from the adjusted R-square of the simple regression model 1). The general importance of the paper adds 6.4% (difference between the adjusted R-squares of model 2 and 1). To have fun during working at the paper adds another 1.7% and the quality of the teacher support further 0.9% of the variance in the grade of the paper 3, when included in the regression model. According to model 4 (table 3) the power of the influence (beta-weight) is highest for the achievement level with β=.389**; followed by the perceived importance of the paper (β=.174**) and the fun at work (β=.145**). The lowest beta-weight has the predictor quality of the teacher support (β=.107*). Summary and perspectives The students previous achievement level is the strongest predictor for the quality of their paper. This result is not surprising; it reflects general findings in educational research very well. Generally high achieving students have good grades on this task as well. Students who find writing this special kind of paper in school to be important achieve slightly better grades than students who disagree. The importance of the grade for the student and the personal importance of the paper do not seem to be as determining as the attitude to these special kind of papers which are the first preparation for scientific work. Having fun during the working process on the paper is also relatively important for its quality. The interest in the topic of the paper is in contrary less important. Enjoying the work is more important than having an interesting topic to work on. In spite of the fact that the influence of the perceived teacher support is quite low it would be inequitable to consider the work of teachers concerning this kind of papers as ineffective. It has to be considered that there is a wide range of influences on the quality of these papers. From this point of view the support teachers are giving their students during the work process on the papers should not be underestimated. It also has to be considered that the questions 3 adjusted R Square Change 7
students answered in the study about teachers behavior and school preparation could not be that detailled. Qualitative interviews with students and teachers could eventually reveal the more specific didactic processes during the working process. These are the first analyses concerning the research question which are only giving a superficial overview yet. The study is still in progress. Deeper analyses, e.g. on the interactions with habitual student variables (e.g. self-concept, goal orientation) and with process variables (e.g. use of metacognitive, cognitive and volitional strategies) during the working process will be conducted as well as the issue which students profit the most from school preparation and teacher support. References Baumert, J. et al. (1999). Fähigkeit zum selbstregulierten Lernen als fächerübergreifende Kompetenz. www.mpib-berlin.mpg.de/pisa/cccdt.pdf. 19.08.2005. Boekaerts, M. (1999). Self-regulated learning. Where we are today. International journal of educational research, 31, 445-475. Boekaerts, M./Niemivirta, M. (2000). Self-regulated learning. Finding a balance between learning goals and ego-protective goals. In Boekaerts M./Pintrich P.R./Zeidner M. (eds): Handbook of self-regulation (417-450). San Diego: Academic Press. Schiefele, U. & Pekrun, R. (1996). Psychologische Modelle des fremdgesteuerten und selbstgesteuerten Lernens. In F. E. Weinert (Hg.): Psychologie des Lernens und der Instruktion. Enzyklopädie der Psychologie, Themenbereich D, Band 2 (249-278). Göttingen/ Bern/Toronto/Seattle: Hogrefe. 8