Elsevier

Learning and Instruction

Volume 20, Issue 1, February 2010, Pages 18-29
Learning and Instruction

Writing learning journals: Instructional support to overcome learning-strategy deficits

https://doi.org/10.1016/j.learninstruc.2008.12.001Get rights and content

Abstract

Although writing learning journals is a powerful learning tool, instructional support is needed to overcome deficits in the use of self-regulated learning strategies. In a 2 × 2 experimental design with high-school students (N = 70), we analysed the effects of two modes of instruction (namely, informed prompting and learning-journal example) along with prompts. Informed prompting that provided background information on the prompted strategies enhanced learning in the training and transfer session. A learning-journal example that modelled the application of the strategies primarily fostered the strategy used in the training session and learning in the transfer session. Theoretically, the results provide support for the self-regulation view of writing-to-learn.

Introduction

Writing is a very common follow-up course-work activity. Frequently, students are required to perform writing assignments such as note taking, essays, term papers, summaries, scientific reports, learning journals, etc. In these cases, teachers are acting (implicitly) on the assumption that writing automatically fosters learning. This is, however, not always true. Research on writing-to-learn has revealed that writing does not automatically contribute to learning. Rather, writing affects learning positively if specific cognitive and metacognitive strategies of self-regulated learning are explicitly supported by the writing task. The present study is concerned with the instructional supports that render writing a productive learning activity, especially when high-school students write learning journals.

In education, there exists the long-lasting belief that writing automatically contributes to learning, the strong-text-view of writing-to-learn (see Emig, 1977). However, current research in writing-to-learn has not provided such a clear relationship between writing and learning (Ackerman, 1993, Bangert-Drowns et al., 2004, Klein, 1999). On the one hand, studies in writing-to-learn provide evidence that writing can contribute to learning (Applebee, 1984, Tynjälä et al., 2001). For example, Mason (2001) found that writing served as a tool to reason on, monitor, and communicate conceptions and understandings of science topics. Similarly, Gunel, Hand, and McDermott (2009) showed that writing-to-learn supported students' understanding of science concepts, especially when students wrote for peers or younger students as compared to writing for teachers or parents. With respect to learning scientific principles through analogies, writing conditions produced greater learning gains compared to a speaking-only condition (Klein, Piacente-Cimini, & Williams, 2007). On the other hand, recent research reviews and meta-analytic studies revealed that writing-to-learn effects are often inconsistent and typically rather small (see Ackerman, 1993, Klein, 1999). According to a meta-analysis by Bangert-Drowns et al. (2004), writing did not inherently enhance learning. Rather, most writing assignments yielded small effects, with an average effect size of 0.20.

These results raise the question of which variables moderate the effects of writing-to-learn. To answer this question, Bangert-Drowns et al. (2004) considered the impact of writing on discrete learning processes. The authors argued that writing contributes to learning by supporting beneficial cognitive and metacognitive strategies of self-regulated learning, the self-regulation view of writing-to-learn (see also Nückles, Hübner, & Renkl, 2009). Actually, Bangert-Drowns et al. (2004) identified metacognitive prompts that stimulated metacognitive processing (e.g., monitoring, self-regulatory processes) as a significant predictor of the learning effects of writing. In a similar vein, Berthold, Nückles, and Renkl (2007) found that cognitive learning strategies mediated learning outcomes while writing learning journals. Additionally, in a recent study by Nückles et al. (2009), knowledge acquisition while writing learning journals was highest when students received cognitive and metacognitive prompts for their writing. Hence, according to the self-regulation view, writing enhances learning if beneficial cognitive and metacognitive learning strategies are triggered by the writing task.

Following the self-regulation view of writing-to-learn, learning journals can be considered as a writing task that fosters beneficial cognitive and metacognitive learning strategies. A learning journal is a writing assignment, typically to be performed as a follow-up course-work activity. After attending a lecture or a course, students are asked to write down their reflections on the previously encountered materials. Research revealed that students who wrote learning journals gained significantly more knowledge compared with students who did not write learning journals (Connor-Greene, 2000, Wong et al., 2002). Furthermore, learning journals proved to be superior over other writing tasks, such as writing a summary (Cantrell, Fusaro, & Dougherty, 2000) or writing a scientific report (McCrindle & Christensen, 1995). In comparison to other writing tasks (e.g., summary and scientific report), learning journals are explicitly intended to induce productive cognitive and metacognitive strategies of self-regulated learning.

On the cognitive level, students should employ strategies such as the organisation and elaboration of the learning material. Organisational strategies (e.g., identifying main points and structuring the learning material) help to establish so-called internal links, that is, finding a meaningful structure of the learning contents (Mayer, 1984, Weinstein and Mayer, 1986). Elaboration strategies (e.g., generating examples and using analogies) serve to build external links that relate the new material to the learner's prior knowledge (Mayer, 1984). Additionally, on the metacognitive level, students should continuously monitor their learning in order to prevent illusions of understanding (Chi, Bassok, Lewis, Reimann, & Glaser, 1989) and strive to bridge gaps in understanding by accomplishing remedial organisation and elaboration strategies. In summary, according to the self-regulation view of writing-to-learn, the application of cognitive and metacognitive learning strategies while writing a learning journal is crucial for learning success.

The use of learning strategies is strongly linked to the learners' age. Research on strategy development revealed that preschool children are often acting passively and rather non-strategically in memory tasks (Flavell, Beach, & Chinsky, 1966). By the end of elementary school, spontaneous and effective uses of memory strategies such as rehearsal and categorisation can be observed (Schneider & Bjorklund, 1998). In order to cope with increased study requirements, high-school students from the ages of 10 to 16 years are developing more sophisticated cognitive and metacognitive learning strategies (Nolen and Haladyna, 1990, Pintrich and De Groot, 1990, Zimmerman and Martinez-Pons, 1990). The developmental patterns found with regard to high-school students on sophisticated learning strategies are very similar to those found for primary-school students with regard to simple recall strategies (Brown, Bransford, Ferrara, & Campione, 1983). Typically, there is a sporadic emergence of beneficial strategies in an early stage of development, resulting in an increased and stable tendency to use the strategy. For example, with regard to elaborative strategies, Beuhring and Kee (1987) found that 5th graders use elaborations less frequently than 12th graders. In a similar vein, Zimmerman and Martinez-Pons (1990) revealed that 5th graders reported significantly less self-regulated learning strategies compared to 8th graders. In turn, 11th graders surpassed 8th graders in different measures of self-regulated learning.

However, faced with complex academic tasks such as writing, studying texts, or scientific reasoning, even older learners typically demonstrate strategy deficits (Brown et al., 1983, Winne, 2005). For example, Rachal, Daigle, and Rachal (2007) came to the conclusion that students are often not prepared to consistently use effective learning strategies once they begin college. Similar results were found for writing learning journals. Although learning journals allow for the application of cognitive and metacognitive strategies, university students typically do not show such strategies to a satisfactory degree without instructional support (Nückles, Schwonke, Berthold, & Renkl, 2004). Hence, research revealed that instructional support is needed in order to enhance self-regulated learning strategies. Furthermore, empirical evidence suggests that the younger the students are, the more important instructional support is in order to overcome strategy deficits.

Work on strategy development identified different reasons for deficiencies in strategy use (Flavell et al., 1966). A mediation deficiency implies that students are not able to use a strategy to improve their task performance, because they do not possess the necessary cognitive requirements (mediators) to benefit from the strategy. However, even being able to apply a strategy efficiently does not guarantee their actual use. A production deficiency will be diagnosed if learners are capable of using a strategy with good effects, but do not do so spontaneously. The reasons for production deficiencies are seen in insufficient meta-knowledge about fields of application and functional value of strategies (Brown, 1978, Flavell, 1978). Finally, recent studies of strategy development introduced another deficit to mediation and production deficiencies, the utilisation deficiency (see Miller, 1990). By definition, utilisation deficiency occurs when “a spontaneous strategy or a semispontaneous (e.g., prompted) trained strategy may not help recall” (Miller, 2000, p. 1014). In other words, unfamiliar strategies do not necessarily have a positive impact on learning outcomes during the initial stage of usage. This is due to the fact that the available cognitive capacity is largely devoted to the application of the unfamiliar strategy, leaving few capacities left for content learning. Depending on what kind of deficit is predominant, different kinds of instructional support will be helpful.

In the case of production deficiencies, prompts are a promising means of overcoming such deficits (King, 1992, Pressley et al., 1992). Prompts are questions or hints that are designed to induce productive learning behaviour. According to Reigeluth and Stein (1983), prompts can be conceived as “strategy activators”, because they elicit strategies that the learners are already capable of, but do not use spontaneously. With respect to journal writing in university courses, prompts proved to be effective. A study by Berthold et al. (2007) showed that providing undergraduate students of psychology with cognitive prompts or with a combination of cognitive and metacognitive prompts significantly increased learning. Similarly, Nückles et al. (2009) found that providing undergraduate university students with cognitive and metacognitive prompts while writing learning journals enhanced learning outcomes. A study by Schwonke, Hauser, Nückles, and Renkl (2006) showed that undergraduate students who received prompts that were adapted to the students' self-reported strategic behaviour gained significantly more knowledge than learners who received randomly selected, non-adapted prompts. In summary, when university students wrote learning journals, prompts were an effective means of encouraging students to elicit strategies that they would spontaneously apply, but to a rather unsatisfactory degree.

These results lead to the question of whether less advanced learners such as high-school students may benefit in the same way from cognitive and metacognitive prompts as university students do. A recent study suggested that prompts that proved to be beneficial for university students might not provide sufficient and effective instructional support for high-school students (Glogger, Schwonke, Holzäpfel, Nückles, & Renkl, submitted for publication). High-school students seemed to have no clear idea of how to respond to the prompts. To test the hypothesis that prompts might be more specific for younger learners, Glogger et al. (submitted for publication) compared non-specific prompts that had proven to be beneficial for university students (non-specific prompts) with prompts including additional scaffolds on specific learning strategies (specific prompts). Although specific prompts increased the quantity of cognitive learning strategies compared with non-specific prompts, they did not enhance the quantity of metacognitive strategies. Moreover, the specific prompts did not positively affect the quality of cognitive and metacognitive learning strategies. Hence, Glogger et al. (submitted for publication) came to the conclusion that the skills of high-school students are probably not sufficiently developed. Rather, younger and less advanced learners need more specific guidance and “sole” prompting (specific or non-specific) to use cognitive and metacognitive strategies when writing learning journals is not sufficient. Presumably, the strategy deficit is obviously more fundamental. These results suggest that high-school students need additional instructional support while writing learning journals with prompts. This raises the question of which type of additional instructional support along with cognitive and metacognitive prompts might be helpful.

According to Paris, Lipson, and Wixson (1983), students should have declarative and conditional strategy knowledge in addition to knowledge on how to perform a strategy. For example, learners ought to know that activities such as organising and elaborating the learning contents are crucial for deep understanding and sustained retention (declarative knowledge). Additionally, students should possess conditional knowledge about strategies, including knowing when and why strategies should be applied (Paris et al., 1983). Thus, given the importance of such declarative and conditional meta-knowledge (also called metastrategic knowledge, Zohar & Peled, 2008), simply presenting prompts that generally advise to use certain strategies may not be sufficient in helping students overcome their production deficiencies (Brown, 1978, Flavell, 1978). These theoretical considerations stimulated empirical studies that examined the effectiveness of providing students with relevant metastrategic knowledge, which encouraged students to apply beneficial learning strategies. For example, with regard to rehearsal strategies, there is empirical evidence that instructions including such information (e.g., informed feedback, informed training) enhanced learning outcomes. Paris, Newman, and McVey (1982) found that children from primary school who received brief explanations on the reasons why memory strategies would aid in remembering pictures (informed training) demonstrated a higher use of these strategies and a greater recall than children without such information. An experimental study by Simon, McShane, and Radley (1987) with primary-school students showed that an informed-training group with a detailed explanation of why certain strategies were useful outperformed a blind-training group on a problem-solving computer program.

In summary, with regard to rehearsal strategies and primary-school students, declarative and conditional strategy knowledge proved to be beneficial. Hence, the question arises whether more advanced learners (high-school students) would also benefit from the provision of metastrategic knowledge in addition to prompts (so-called informed prompting) when trying to apply rather sophisticated and complex strategies such as organisational, elaborative or metacognitive strategies in writing learning journals.

If mediation deficiencies are predominantly responsible for the failure to use strategies both prompts and the provision of metastrategic knowledge may not be sufficient. When students have a mediation deficiency, they do not use a certain strategy because they do not possess the cognitive requirements to apply the strategy. Hence, in this case, instructional support that fosters the acquisition of the lacking cognitive skills is required. Regarding cognitive skill acquisition, worked-out examples proved to be effective (Atkinson, Derry, Renkl, & Wortham, 2000). Traditionally, worked-out examples provide algorithmic solution steps that lead to a final solution. Although research on worked-out examples was concentrated on algorithmic domains for a long time (e.g., mathematical subdomains), recent empirical studies investigated learning from examples in non-algorithmic domains as well (Gentner et al., 2003, Rourke and Sweller, 2009, Rummel and Spada, 2005, Schworm and Renkl, 2007). In the case of writing learning journals, a non-algorithmic worked-out example can be provided in the form of an example of a well-written learning journal, showing the critical features that make a learning journal well written (e.g., well-organised, including many examples, high amount of metacognitive statements, etc.). Research revealed that learning from written examples or learning by observing a model performing a writing task proved to be successful (see Braaksma et al., 2002, Ellis et al., 2005, Graham and Perin, 2007). For example, Couzijn (1999) found with regard to 9th graders that learning-to-write argumentative texts by observing a model was superior to a learning-by-doing condition. Braaksma, van den Bergh, Rijlaarsdam, and Couzijn (2001) identified effective activities of observational learning when learning-to-write an argumentative text. Amongst other things, they found that focusing on a written example when elaborating on the performance of a model positively contributed to the acquisition of writing skills.

In summary, providing high-school students with written examples proved to be beneficial with regard to the acquisition of writing skills. Hence, this raises the question of whether providing high-school students with a learning-journal example along with cognitive and metacognitive prompts would help them to overcome mediation deficiencies by supporting the acquisition of cognitive skills in order to use the prompted cognitive and metacognitive strategies.

The studies described in the previous paragraph used worked-out examples and modelling to facilitate students' acquisition of writing skills. The focus was on instructing students on how to produce argumentative text, such as discussing and integrating opposing positions in an essay. Hence, the main goal was to help students acquire a certain genre or rhetorical format (Klein, 1999). It is obvious that writing to instantiate a particular genre (e.g., a scientific report, an analytic essay) places high demands on the writer and is likely to overtax novice writers (Torrance, Fidalgo, & García, 2007). Therefore, instructional approaches to learning-to-write typically include rather extensive and long-term interventions. For example, in the study by Braaksma et al. (2002) the courses on writing argumentative texts consisted of two instructional sessions lasting about 60 min (first session) and 30 min (second session). Couzijn (1999) implemented four lessons on writing argumentative text over two days with each lesson lasting 1 h.

In contrast to these learning-to-write interventions (see Braaksma et al., 2002, Couzijn, 1999), our approach to writing learning journals does not aim to convey a sophisticated rhetorical format. Unlike genres such as scientific articles, essays or argumentative texts, learning journals do not have a fixed rhetorical structure (see Nückles et al., 2009). Rather, the writing of a learning journal is very free and expressive and allows the learner to personally select which aspects of a learning episode require deeper reflection. Hence, contrary to rhetorically high demanding writing tasks such as writing argumentative texts, the writing of learning journals is rhetorically less demanding. Therefore, the introduction into journal writing is not intended to equip students with ample knowledge on rhetorical formats and the corresponding writing skills. In other words, we suggest that writing learning journals should not be conclusive and extensive interventions into writing, as known from the learning-to-write literature (Glaser and Brunstein, 2007, Graham and Harris, 2003). Rather, our introductions should efficiently help students use writing as a beneficial medium to facilitate self-regulated learning strategies. However, as previous research suggests (Bangert-Drowns et al., 2004, Berthold et al., 2007, Nückles et al., 2009), it is essential to explicitly support the application of such strategies. Against this background, we have developed brief instructions into journal writing that should help students efficiently apply cognitive and metacognitive strategies of self-regulated learning.

The following hypotheses and research questions were addressed: Providing high-school students with a learning-journal example might specifically encourage them to extensively apply the prompted cognitive and metacognitive strategies. As is well known from observational learning, the opportunity to see a model typically raises the frequency of the observed target behaviour (Bandura, 1986). Hence, we predicted that students who received a learning-journal example would apply a greater amount of cognitive and metacognitive strategies in their journal entries compared to students without a learning-journal example (Hypothesis 1a). However, what might be of interest is to what extent the information about why and when to use the prompted strategies would raise the quantity of the strategies. Hence, we were interested in whether the informed prompting would affect the amount of the prompted cognitive and metacognitive strategies in the written learning journals. Hence, we assumed that the provision of declarative and conditional metastrategic knowledge might help the students apply the cognitive and metacognitive strategies in a more effective way (Hypothesis 1b).

Regarding learning outcomes, predicting the effects of the learning-journal example is more complicated: Simply raising the frequency by which a certain strategy is applied might not necessarily improve learning success. For example, in terms of an utilisation deficiency (Miller, 1990), unfamiliar strategies do not necessarily have a positive impact on learning outcomes during the initial stage of usage. Thus, providing students with a learning-journal example should encourage them to practice the prompted strategies in the training session. However, it is possible that some practice of the modelled strategies is needed before an improvement in learning outcomes may occur. Thus, we predicted that participants who studied a learning-journal example would gain more knowledge than participants without a learning-journal example in the transfer session, but not necessarily in the training session (Hypothesis 2a). As regards the informed prompting, we assumed that the provision of metastrategic knowledge would help the students use the prompted strategies in a more effective way and thereby enhance learning outcomes. Hence, students who received the informed prompting should outperform students who did not receive this mode of instruction as regards their learning outcomes (Hypothesis 2b).

Regarding the combination of informed prompting and a learning-journal example, we assumed that both modes of instruction would complement each other. Giving students an example of a journal entry should stimulate the application and practice of the depicted strategies. Providing them together with metastrategic information about why and under which conditions to apply the strategies should help them apply the strategies in a qualitatively superior fashion (Hypothesis 3a). It is also plausible to assume that the combination of the two modes of instruction might have either an additive or even over-additive effect on learning outcomes (Hypothesis 3b).

Section snippets

Sample – design

To test the effects of two different modes of instruction (informed prompting and learning-journal example) in addition to cognitive and metacognitive prompts, an experimental study was conducted.

Students from different German secondary schools (N = 70) participated in this experiment (mean age = 17.62, SD = 0.86; 25 male, 44 female, 1 unknown). They received 15 euros for their participation. The students were randomly assigned to one condition of a 2 × 2 factorial design (see Fig. 1), where one factor

Results

An alpha-level of 0.05 was used for all statistical analyses. Partial η2 was chosen as effect size measure.1 Table 2 presents the descriptives for all measures of the study.

Regarding the pretest scores, the overall mean was 1.21 (SD = 0.24; the highest value that could be accomplished was 5, see Method section). Thus, the learners had very low

Summary of the effects

With respect to Hypothesis 1a, namely that the learning-journal example would raise the amount of the prompted strategies, the results we found for the organisation of the written learning journals did not confirm our prediction. Providing students with a well-organised learning-journal example did not affect the organisation of their own learning journals positively. However, in line with Hypothesis 1a, the learning-journal example had strong effects on the amount of elaborative and

Acknowledgements

This research was supported by the Deutsche Forschungsgemeinschaft (DFG; [German Research Foundation]) with a project grant awarded to Matthias Nückles (contract NU 129/2-1). We would like to thank Bettina Brockhaus and Julian Roelle for their assistance in coding the learning journals and learning outcomes. We also wish to thank Ella Micheler for her proofreading.

References (57)

  • S.G. Paris et al.

    Learning the functional significance of mnemonic actions: a microgenetic study of strategy acquisition

    Journal of Experimental Child Psychology

    (1982)
  • A. Rourke et al.

    The worked-example effect using ill-defined problems: learning to recognise designers' styles

    Learning and Instruction

    (2009)
  • R. Schwonke et al.

    Enhancing computer-supported writing of learning protocols by adaptive prompts

    Computers in Human Behavior

    (2006)
  • M. Torrance et al.

    The teachability and effectiveness of cognitive self-regulation in sixth-grade writers

    Learning and Instruction

    (2007)
  • A. Zohar et al.

    The effects of explicit teaching of metastrategic knowledge on low- and high-achieving students

    Learning and Instruction

    (2008)
  • J.M. Ackerman

    The promise of writing to learn

    Written Communication

    (1993)
  • A.N. Applebee

    Writing and reasoning

    Review of Educational Research

    (1984)
  • R.K. Atkinson et al.

    Learning from examples: instructional principles from the worked examples research

    Review of Educational Research

    (2000)
  • A. Bandura

    Social foundations of thought and action: A social-cognitive theory

    (1986)
  • R.L. Bangert-Drowns et al.

    The effects of school-based writing-to-learn interventions on academic achievement: a meta-analysis

    Review of Educational Research

    (2004)
  • J.B. Biggs et al.

    Evaluating the quality of learning: The SOLO taxonomy

    (1982)
  • M.A.H. Braaksma et al.

    Observational learning and the effects on model-observer similarity

    Journal of Educational Psychology

    (2002)
  • M.A.H. Braaksma et al.

    Effective learning activities in observation tasks when learning to write and read argumentative texts

    European Journal of Psychology of Education

    (2001)
  • A.L. Brown

    Knowing when and how to remember: a problem of metacognition

  • A.L. Brown et al.

    Learning, remembering, and understanding

  • R.J. Cantrell et al.

    Exploring the effectiveness of journal writing on learning social studies: a comparative study

    Reading Psychology

    (2000)
  • J. Cohen

    Statistical power analysis for the behavioral sciences

    (1988)
  • P.A. Connor-Greene

    Making connections: evaluating the effectiveness of journal writing in enhancing student learning

    Teaching of Psychology

    (2000)
  • Cited by (103)

    • The effects of self-monitoring on strategy use and academic performance: A meta-analysis

      2022, International Journal of Educational Research
      Citation Excerpt :

      In school-based interventions, self-monitoring is typically mediated by journals and diaries. As a monitoring tool, they were used for capturing trend in the learning process, such as changes in task perception, interest, self-efficacy for self-regulation, and effort (e.g., Ferreira, Simão & da Silva, 2015; *Hübner, Nückles & Renkl, 2010; *Perels, Gürtler & Schmitz, 2005). Longitudinal studies (e.g., *Fabriz et al., 2014) showed that rating one's learning behaviors over a period of several weeks through a learning protocol improved self-regulation.

    • Self-regulated learning by writing learning protocols: Do goal structures matter?

      2021, Learning and Instruction
      Citation Excerpt :

      This research has also shown that the outlined enhancement of cognitive and metacognitive learning processes via optimizing the learning protocol task itself fosters learning outcomes. Specifically, both the increased quantity and the increased quality of learning processes were related to increases in subsequent performance on conceptual knowledge tests in previous studies (e.g., Hübner, et al., 2010; Nückles et al., 2009; Roelle, Nowitzki, & Berthold, 2017; see also; Glogger et al., 2012). In contrast to the wealth of studies that were targeted at optimizing the learning protocol task itself, there are hardly any studies that focused on the role of the context in which the learning protocol task is embedded.

    • Can Generative Learning Tasks be Optimized by Incorporation of Retrieval Practice?

      2020, Journal of Applied Research in Memory and Cognition
    View all citing articles on Scopus
    View full text