Self-regulation is a competence humans need to meet the requirements of life-long learning. Therefore, it is important to analyse the possibilities to support self-regulated learning. In this symposium (chair: F. Perels, B. Otto and B. Schmitz, University of Technology Darmstadt, Germany), we present four different approaches that examine important aspects of self-regulated learning in school and university.

• C. Dignath et al., University of Frankfurt, Germany: The efficacy of self-regulated learning interventions at primary and secondary school level – a meta-analysis


• D. Whitebread & P. Coltman, University of Cambridge, UK: Towards a pedagogy for teaching thinking and developing metacognitive abilities in young children


• F. Perels et al., University of Technology Darmstadt, Germany: Improving self-regulated learning of preschool children. Evaluation of a training for kindergarten teachers


• P. Gradinger et al., University of Vienna/ University of Applied Sciences Linz, Austria: Improving self-regulated learning: Differential effects by VEL based on quantitative and qualitative data

Beginning with a meta-analysis investigating the efficacy of self-regulated learning interventions at primary and secondary school level the second paper reports findings from two studies focusing on different approaches to teaching ‘thinking’ and the development of self-regulatory and metacognitive abilities in young children. The third reported study also refers to self-regulated learning of young children by presenting the findings of a training programme for kindergarten teachers to support self-regulated learning of pre-school children. We close the symposium with another intervention study, which describes differential effects of a computer-based intervention within university context.

The presented studies will be discussed regarding their theoretical and educational significance (M. Schmidt, University of Technology Darmstadt, Germany).

Recently, research has increasingly focused on fostering self-regulated learning amongst young children. To consider this trend, this article presents the results of a differentiated meta-analysis of 77 studies on enhancing self-regulated learning amongst primary and secondary school students. Based on recent models of self-regulated learning, which consider motivational as well as cognitive and metacognitive aspects (see Boekaerts, 1999), effects of self-regulated learning on academic achievement, cognitive and metacognitive competencies, as well as on motivation were analyzed. To get a differentiated picture of the training effects on these diverse components of self-regulated learning, separated meta-analyses were conducted for every single outcome category. As the results show, self-regulated learning training proved to be effective already at primary school level. Subsequent analysis tested for the effects of several moderator variables, which consisted of study features and training characteristics. Most of the moderator analyses revealed inconsistent results amongst the various outcome categories. Therefore, the effectiveness of a certain training feature seems to depend on the school subject and the learning contents (e.g., mathematics or reading). The analysis yielded effects in favour of interventions conducted in the classroom instead of carried out by the researchers themselves. Age-related distinctions showed that younger and older students benefit most from interventions in a different way. Students achieved higher effects when training programmes included the instruction of metacognitive strategies. Concerning the instruction of motivational strategies, this result could be found only for the mathematics area. Moreover, the synthesis detected a general problem of very heterogeneous definitions of metacognition and self-regulation. Further research on fostering self-regulation competence amongst students should increasingly take into account the combination of metacognitive and motivational strategies instruction. This demand is closely linked to the need of a more consistent meaning of the terms metacognition and self-regulation.

Over the past years, more and more research on self-regulated learning in school settings was conducted whereby various models of self-regulated learning emerged. As different models emphasise different aspects of self-regulated learning, interventions to foster self-regulation have been developed with foci on diverse skills and competences. The resulting variety of training approaches and contents necessitates analyzing which aspects of interventions are successful. We conducted a meta-analysis in order to (1) give an account of existing trainings and (2) to draw general conclusions of the usefulness of several training characteristics.

Up to now, the effect of self-regulated learning has been investigated mainly in older students. However, recent research has increasingly focused on fostering self-regulated learning at primary school level. For this purpose a first meta-analysis was conducted including only interventions in grades 1 through 6. In a next step, data of studies conducted in secondary schools were analysed in order to seek for differences between the interventions at different school levels.

In current research there is a clear agreement that applying self-regulation to learning contexts has a positive impact on learning behaviour and academic achievement (Schunk, 2005). The decline of debates on life long learning and key competences acquisition at school as well as the large number of intervention studies aimed at training self-regulated learning in students suggest the importance of summarizing these empirical findings. Rather than developing further training programs to foster self-regulation skills it appears useful to provide an overview of existing interventions and to investigate which kind of training works best and which aspects have proved their worth.

Due to the great variety of underlying models and definitions of self-regulated learning, we first had to define self-regulated learning in order to select studies on interventions aiming at fostering the same construct. With Boekaerts (1999) we define self-regulation as an interaction of cognitive, metacognitive and motivational processes which work together in the information processing system. This meta-analysis thus contains studies based on conceptions of self-regulation which include at least two aspects of this three-layered model. The definition of self-regulated learning was detected with the help of the theoretical background of the article and the operationalization of the variables to measure the construct.

For the literature research the computerized databases PsycInfo, ERIC and Psyndex were searched using 45 keywords such as learning strategies, self-regulatory strategies, metacognitive skills, self-regulated learning, etc. The search included only studies published after 1992, given that a meta-analysis on self-regulated learning interventions already exists from the year 1992. Additionally, we identified articles in bibliographies given in primary studies or in review articles.

Only studies aiming at training self-regulated learning at primary or secondary school were considered in this meta-analysis. Studies also had to employ a control group design with longitudinal measurement and had to provide sufficient descriptive data to calculate effect sizes. A data coding form was developed to assure greater accuracy in the data collection process.

From 274 studies published in this domain between 1992 and 2005, 104 were conducted at primary school level, 102 at secondary school level and 68 with students with learning disabilities. Only studies that employed primary school students without learning disabilities were included in the coding process of the first analysis. However, most of these studies did not meet the criteria: only 47 intervention studies were eligible and thus included in this analysis. In a second step, 77 secondary school interventions were integrated following the same criteria as the ones mentioned above.

Mean values, standard deviations and sample sizes were extracted from each primary study to compute effect sizes based on the standardized mean difference (see Hedges & Olkin, 1985).

For comparing and combining results from the primary studies, the weighted mean effect size as well as homogeneity tests were computed. To explain the variability of effect sizes, a meta-analytic Regression was conducted (see Lipsey & Wilson, 2001) in order to examine relationships between study outcomes and characteristics of the studies, as study design, instruments to measure self-regulation, or duration and kind of treatment.

Studies were combined concerning the domain of self-regulation strategies (cognition, metacognition, motivation) and the school subject the intervention aimed at. Tests showed that results of different studies were heterogeneous for the different outcome categories.

Effect sizes turned out to be higher for primary school students, but in most cases, this difference between primary and secondary school level was not significant. Concerning the impact of moderator variables, the results of both analyses show similar patterns. However, there are differences between both school levels: Most of the moderator analyses for primary school studies revealed significant results for, whereas for secondary school, most of the differences go in the same direction, but they are less frequently significant. Concerning the inclusion of metacognitive strategy instruction in a training, there are some differences: Nearly for all groups, effect sizes are higher if metacognition was trained. The instruction of metacognitive strategies seems to have a positive impact on training success. Another interesting point is the influence of the direction of training: Teacher-directed programmes are more effective across the outcomes “academic performance mathematics”, “metacognitive strategies” and “attribution”. Self-directed programmes are more effective concerning academic performance in science as well as unspecified academic performance. Cross-classified comparisons should provide explanations for these differences between programmes conducted in the context of mathematics or reading and writing. These analyses revealed that training programmes conducted in a mathematical context are mostly self-directed while text comprehension programmes are mostly teacher-directed. This is just one example for the difference of training characteristics between mathematical context and context of text comprehension.

Which training works best seems to depend on the context and the learning content. Altogether, programmes that are implemented in a regular classroom setting seem to be more effective. Even if older children benefit more from interventions in the academic area, it does make sense to train students in self-regulated learning strategies as early as first grade, regarding the effects on motivation and strategy use.

This paper reports findings from two studies focusing on different approaches to teaching ‘thinking’ and the development of self-regulatory and metacognitive abilities in young children.

There is a considerable literature supporting the crucial significance of metacognitive abilities for the development of effective thinking and learning. Other research has made claims about the efficacy of 'teaching thinking' programs and approaches in stimulating metacognitive development.

In much of the literature there is agreement that metacognitive abilities begin to develop around the age of 8 years. With some notable exceptions, therefore, much of the work examining the relation between pedagogy and metacognition has focused on older Primary and Secondary aged school children. However, more recent work, by ourselves and others, has shown that the self-regulatory and metacognitive abilities of younger children (i.e. from 3-7 years) have been systematically underestimated by self-report and experimental methodologies. Observational approaches in naturalistic settings have revealed much earlier origins of the abilities to monitor and regulate cognitive activity.

The studies reviewed in this paper have explored the extent to which :

·        young children are capable of developing metacognitive abilities and reporting upon their own thinking

·        pedagogical programmes can be developed which facilitate self-regulatory and metacognitive abilities in young children

·        we can identify the key elements in such practices which appear to be related to the development of young children’s self-regulatory and metacognitive abilities

AIMS

The aims of the two main projects reported in this paper were complementary, exploring between them aspects of metacognition in young  children between three and seven years of age. The first study reported, the CIndLe (Cambridge Independent Learning) project, aimed to determine the metacognitive, self-regulatory capabilities of 3-5 yr old children. The second, known as the META (Metacognition & Thinking Abilities) project, developed ideas from the earlier work by aiming to investigate the effectiveness of approaches to teaching thinking in developing metacognitive abilities of 4-7 year old children.  The two studies together aimed to explore the following research questions:

1.      To what extent are young children capable of developing metacognitive & self-regulatory abilities, and reporting on their own learning?

2.      Can pedagogical programmes and practices be developed which facilitate self-regulatory and metacognitive abilities in young children ?

3.      Can we identify the key elements in such practices and programmes ?

METHODOLOGY/RESEARCH DESIGN

The CIndLe (Cambs Independent Learning in the Foundation Stage) was a two year project involving practitioners from 32 different English Nursery and Reception classrooms/settings. During the project, pedagogies derived from previous literature were developed by practitioners which were likely to stimulate metacognitive activity in young children eg: co-operative group work, reciprocal teaching, self explanations, self assessment and debriefing.

Practitioners recorded their observations and researchers video-recorded sequences of activity. From both of these data sources  ‘events’ exemplifying aspects of metacognitive behaviour were identified. A framework for analysis of this observational data, initially based upon existing models of metacognition and self-regulation, was developed.

The second study, the META project, involved practitioners in adapting ‘Teaching Thinking’ programmes to the 3-7 age-group. Approaches explored included Mind-mapping (Caviglioni & Harris, 2000), TASC (Thinking Actively in a Social Context) problem-solving (Wallace, 2002) and Philosophy for Children (Lipman, 1985).. Teachers collected observational data (notes/photos), recorded dialogue and collated products of children’s work. In addition a small experimental study was conducted of the impact of Philosophy for Children sessions on 4-5 year old children with below average linguistic development.

FINDINGS

Research  Q. 1.: Incidence of metacognitive & self-regulatory behaviours in 3-7 year old children.

Analysis of collected data in the CIndLe study resulted in the identification and analysis of 705 self-regulatory events, of which 582 (i.e. 82.6%) contained an element of specifically metacognitive activity. The duration of these events varied from less than a minute to as much as 20-30 minutes, with an average rate of incidence of around 7 events per hour. Detailed analysis of a purposive sample of 196 of these events revealed extensive evidence of metacognitive knowledge, regulation of cognition and both motivational and emotional aspects of self-regulation. Within this, however, behaviours related to the regulation of cognition were easily the most numerous . These included the ability to monitor their own thinking, plan and carry out a deliberate approach or strategy and to identitfy and correct mistakes.

Research Q. 2: Effectiveness of pedagogical programmes and practices

Both the CindLe and META studies provided evidence that a range of pedagogical practices and programmes were effective in provoking and supporting metacognitive activity in young children. Within the paper a range of examples will be presented to illustrate the types of evidence collected including video taped episodes of children’s metacognitive activity and transcripts of exploratory and reflective dialogue.  An analysis will also be presented of the range of distinct cognitive experiences and requirements embedded within these practices and programmes, and of the prevalence of aspects of metacognitive activity in relation to each of these cognitive  experiences and requirements.

Research Q.3: Key elements in such practices and programmes

Evidence from both studies has also suggested that there are a number of key pedagogical elements within the various practices and programmes explored that supported and facilitated metacognitive activity. Quantitative analysis of the CindLe events showed that metacognitive behaviours were most in evidence when activities were child initiated, were carried out in a small group and involved extensive collaboration and talk.

 The more qualitative data provided by the META project supported these indications and suggested that Mind-mapping and Philosophy for Children were particularly effective in provoking metacognitive dialogue. This arose from their provision of:

·        opportunities for shared problem–solving,

·        a context for the development of a structured ‘meta language’,

·        a framework for sustained shared thinking (Siraj-Blatchford et al, 2002), exploratory talk (Littleton et al, 2005) and self-explanations (Seigler 2002).

THEORETICAL AND EDUCATIONAL SIGNIFICANCE

There is a considerable and longstanding literature supporting the crucial significance of metacognitive abilities for the development of effective thinking and learning. Other research has made claims about the efficacy of 'teaching thinking' programs and approaches in stimulating metacognitive development. The development of thinking skills is now also enshrined within the English National Curriculum.

In much of the literature there is agreement that metacognitive abilities begin to develop around the age of 8 years. With some notable exceptions, therefore, much of the work examining the relation between pedagogy and metacognition has focused on older Primary and Secondary aged school children. However, more recent work, by ourselves and others, has shown that the self-regulatory and metacognitive abilities of younger children (i.e. from 3-7 years) have been systematically underestimated by self-report and experimental methodologies. Observational approaches in naturalistic settings have revealed much earlier origins of the abilities to monitor and regulate cognitive activity.

The findings from the studies reported in this paper support this new position, providing extensive evidence that 3-7 year old children are developing their metacognitive abilities, knowledge and motivations, particularly in relation to the ongoing monitoring and regulation of cognition. 

Self-regulation is a competence humans need to meet the requirements of life-long learning, which could be supported already for preschool children. Therefore, the aim of the presented study was to test the effects of a self-regulation training for kindergarten teachers on both their own self-regulation as well as the improvement of self-regulation in the preschool children they were teaching. The training for the kindergarten teachers was based on the process model of self-regulation by Schmitz and Wiese (2006) and also included methods to foster self-regulation in preschool children. 35 kindergarten teachers (19 in the experimental group and 16 in the control group) and 97 children (48 in the experimental group and 49 in the control group) of German kindergartens took part in this study. For the training evaluation, a control group design was applied using a self-regulation questionnaire for the kindergarten teachers and an interview for the preschoolers. The results obtained by means of analyses of variance show that on the level of the kindergarten teachers their self-regulation as well as the use of methods to support self-regulation in preschool children improved significantly (self-regulation, p<.05; methods, p<.10). The self-regulation of the preschoolers whose kindergarten teachers took part in the training improved significantly as well (p<.00). Therefore, the conclusion may be drawn that self-regulated learning of preschool children can be supported if their kindergarten teachers increase their own self-regulation knowledge and competence. The results of this study have important implications for the curriculum of kindergarten instruction, which is in change in Germany.

Self-regulation is a competence humans need to meet the requirements of life-long learning, which should be supported already at preschool children. Therefore, the main goal of the study was to improve self-regulative competences of preschoolers. A training programme for kindergarten teachers was developed and evaluated on the level of the kindergarten teachers as well as on the level of the preschool children.

The training contents are taken from the self-regulation model of Schmitz and Wiese (2006) who divide the self-regulation process into three phases: preaction, action, and postaction.

The presented training conception is based on a two-component approach. In each training session the participating kindergarten teachers learned self-regulatory strategies for their own learning processes (goal-setting, planning, motivation, volition, monitoring, self-reflection) and, at the same time, strategies for supporting the self-regulated learning of the preschoolers (support, encouragement, verbal rewards and modeling; self-regulation contents and methods to support). In general, our training conception follows the suggestions of Schunk and Ertmer (2000), who recommend that, beyond modeling, participants should be given greater responsibility for their own learning and self-regulation should be taught in content areas. 

To test the effectiveness of the training, a two-component design was applied: first, a control group design with repeated measures allowed calculating a pretest-posttest-comparison within and between the groups of kindergarten teachers (analyses of variance with repeated measures). The control group received the training after the experimental group. Based on the idea of the two-component approach, the preschoolers, whose kindergarten teachers took part in the study, were interviewed before and after the training.

35 kindergarten teachers from several German kindergartens took part in the study. Complete data could be used from 19 kindergarten teachers belonging to the training group and 16 participants of the control group. Three preschoolers per kindergarten teacher participated in the study. Concerning the preschoolers in the training group, 48 data sets were complete, whereas the study included 49 children of the control group.

A pretest consisting of a questionnaire was conducted with the kindergarten teachers two weeks before the intervention. At the same time, the preschoolers were interviewed. The training group was trained on five afternoons (one afternoon per week) and each unit lasted 120 minutes. During the training period, the kindergarten teachers were asked to keep a diary for every workday. We expected this kind of self-reflection to support the training progress. Two weeks after the intervention, the kindergarten teachers filled out the questionnaire for a posttest and the preschoolers were interviewed again. Additionally, the kindergarten teachers documented their observation of the preschoolers on an observation sheet at the time of the posttest. The kindergarten teachers of the control group received a training a few months after the experimental group and after the posttest.

The differences between experimental and control group were tested by means of analyses of variance with time as repeated measurement factor (there were no significant differences between the two groups before the intervention).

The results in a “test of knowledge about self-regulation” (including questions about methods, F(1, 33)=7.16, p<.01) as well as for the scale “overall self-regulation” (combines the preaction, the action and the postaction phase, F(1, 33)=5.29, p<.05) improved significantly in the experimental group. The control group did not improve in this period. The scale “methods” included items concerning the kindergarten teachers’ use of support, encouragement, verbal rewards and modeling for the preschoolers. The use of these methods increased for the kindergarten teachers of the training group (F(1, 33)=3.38, p<.10). No significant changes were found for the control group in the same period.

The interview with the children was evaluated by means of analyses of variance with time as repeated measurement factor as well. In case of pretest differences between experimental and control group analyses of covariances were computed.

The children in the experimental group improved significantly with respect to the most important variables concerning self-regulation. Highly significant results were received for all phases of self-regulation, when children were asked to describe these in two different examples. Consequently, the children in the experimental group were able to describe phases of activities in a more detailed and a more exact way after their kindergarten teachers had been trained. This shows that they could also differentiate between the phases of self-regulation. The scale “overall self-regulation“, which combines the preaction, the action and the postaction phase, showed a significant improvement of the experimental group (F(1, 95)= 91.88, p<.00). The control group did not change their level.

Having in mind the importance of self-regulated learning for life-long learning processes, the evaluation of the training showed that preschool children’s learning could be improved by training their kindergarten teachers. This is an important result since the kindergarten teachers spend every weekday with the preschoolers and therefore have a great impact on the learning progress of these children. An important aspect of the two-component approach is the possibility to improve self-regulation of children without training them. The results of this study have important implications for the curriculum of kindergarten instruction, which is in change in Germany.

Schools and universities should promote educational motivation and competences for life long learning of their students (see Commission of the European Communities, 2000; Spiel & Schober, 2002), in short self-regulated learning. But still, empirical results show sometimes an insufficient learning culture in the central institutions of education.

E-learning is one promising possibility to establish a new learning and teaching culture. Moreover e-learning gives a set of other advantages, for example easy accessibility of disadvantaged people.

We will present here the Vienna E-Lecturing (VEL)-program, which was developed, implemented and evaluated by a research team. VEL is a long term intervention of one year duration and is totally integrated in a compulsory lecture at the university VEL is designed as a blended learning concept for big courses at university. It follows up four goals: imparting (1) professional content (research methods and evaluation), (2) cooperative learning, (3) self-regulated learning and (4) e-competences.

The theoretical background of VEL is provided by the process model of self-regulated learning (Schmitz, 2001).

To implement a blended learning concept, the VEL program consists of a systematic combination of virtual learning modules and face-to-face sessions.

Schools and universities should promote educational motivation and competences for life long learning of their students (see Commission of the European Communities, 2000; Spiel & Schober, 2002), in short self-regulated learning. But still, empirical results show sometimes an insufficient learning culture in the central institutions of education. Especially at university, there is not enough time for questions and discussions because of overcrowded lecture rooms. This gives way for a passive and "consuming" student behavior instead of the promotion of self-regulated learning as well as key qualifications (e.g. team working). Student learning is often effort-restricted, mainly performance-orientated instead of competence-orientated.

E-learning is one promising possibility to establish a new learning and teaching culture. Through the demand of structuring lectures newly, the application of new media for problem-oriented learning, active knowledge-construction and self-regulated learning are self-evident. Moreover e-learning gives a set of other advantages, for example easy accessibility of disadvantaged people (handicapped or working students, etc.).

Actually there is a broad offer of projects relating to e-learning at universities. Nevertheless new media are mostly used in a very simple manner (e.g. to publish learning products through the internet). Moreover many projects hardly ever base on well-founded didactical models. But especially these didactical theories are necessary, even though their realisation is time-consuming and complex. Besides that, technical know-how and enough resources for support are required.

The development, implementation and evaluation of well-founded e-learning projects demand high abilities in learning concepts, great efforts for the conceptual-design of the courses, etc. Therefore e-learning concepts may not be realized in the framework of university lectures but within research projects. Subsequent to such projects transfer into regular courses and generalisation to other contents should be devised to assure sustainability. Evaluation of effectiveness of didactical elements is inevitable to make effective and efficient transfer and generalisation possible.

One example for meeting these demands at university is the Vienna E-Lecturing (VEL) program, which is presented in this contribution.

VEL is designed as a blended learning concept for big courses at university. It follows up four goals: imparting (1) professional content (research methods and evaluation), (2) cooperative learning, (3) self-regulated learning (regulation of emotions, monitoring, learning strategies, etc.) and (4) e-competences (handling the platform tools, etc.).

The theoretical background of VEL is provided by the process model of self-regulated learning (Schmitz, 2001), which is in central parts empirically evidenced (e.g. Schmitz & Wiese, 2006).

To implement a blended learning concept, the VEL program consists of a systematic combination of virtual learning modules realized at a learning platform and face-to-face sessions. The online modules consist for instance of information input (postings of book chapters, links, etc. on the server), self tests (to check the declarative knowledge and motivate for further learning), and group work (in discussion forums to establish procedural knowledge). The face-to-face sessions consist of meet-the-expert"-units, intended to give the students a frame and to discuss very difficult topics of the lecture, and training sessions, to promote self-regulated learning. Additionally peer-tutors give tutorials, for instance to train students in team work and to handle the e-learning environment.

The main goals of the paper are: (1) to present the intervention VEL, (2) to demonstrate the (differential) effectiveness of VEL and (3) to give additional explanation of quantitative results through qualitative data. Data were collected before, during and after the intervention. Instruments comprised knowledge tests and questionnaires with closed and open questions. Additionally the learning behaviour was observed on the e-learning platform WebCT. Up to now 150 students participated in VEL and 150 students served as controls by attending a reduced e-lecturing course (without the promotion of cooperative and self-regulated learning, or e-competences). The summative evaluation shows positive effects of VEL regarding all four goals, namely professional content, cooperative learning, self-regulated learning and e-competences. Differential effects based on different baselines of self-regulated learning competences are hardly to find, but do exist. Preliminary results show beneficial effects especially for disadvantaged students, whereas well equipped students do not deteriorate. Qualitative Data can support results through clearing up the learning background. Overall, the application of VEL is recommendable for all regular students at university. At the end we will give an outlook for future applications.