Is adaptation to task complexity really beneficial for performance? Stephanie Pieschl a, * , Elmar Stahl b , Tom Murray c , Rainer Bromme a a University of Muenster, Institute of Psychology, Fliednerstr. 21, 48149 Muenster, Germany b University of Education, Kunzenweg 21, 79117 Freiburg, Germany c University of Massachusetts, Amherst, MA and Perspegrity Solutions, 252 Chesterfield Rd, Westhampton, MA, USA article info Article history: Received 28 January 2011 Received in revised form 27 August 2011 Accepted 30 August 2011 Keywords: Self-regulated learning Metacognition Adaptation Hypertext Performance abstract Theories of self-regulated learning assume that learners flexibly adapt their learning process to external task demands and that this is positively related to performance. In this study, university students (n ¼ 119) solved three tasks that greatly differed in complexity. Their learning processes were captured in detail by task-specific questionnaires and computer-generated log files. Results indicate that students adapted almost all learning processes significantly to task complexity. For example, students accessed more hypertext pages for complex tasks than for simple tasks. However, this kind of adaptation was not consistently related to performance. For variables capturing learners’ self-regulation, such as the number of accessed hypertext pages, more pronounced adaptation was significantly and positively related to performance even when learners’ general processing depth was statistically controlled. Results were less consistent for variables capturing learners’ self-monitoring, such as their judged task complexity. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction Adaptation is a central mechanism proposed by most theories of self-regulated learning (Pintrich, 2004; Winne & Hadwin, 1998; Zimmerman, 2002). Based on processes of self-monitoring and self-control (Nelson & Narens, 1994), skillful self-regulated learners are assumed to adapt to a multitude of internal and external cues to optimize their learning processes (Hadwin, Winne, Stockley, Nesbit, & Woszczyna, 2001). Accordingly, adaptation is also hypothesized to be associated with superior performance. Given the pervasive nature and importance of these basic assumptions, the lack of studies explicitly investigating these issues is surprising, especially studies conducted in relatively authentic learning settings. The current study will explore these questions focusing on learners’ adaptation to task complexity as an illustrative example. 1.1. Metacognitive self-monitoring, self-regulation, and adaptation Models of self-regulated learning conceptualize adaptation on different levels of granularity. Fine-grained adaptation could refer to adapting study strategies within the enactment stage of learning. Large-grain adaptation, on the other hand, could have a forward reaching nature affecting learners’ more permanent trait-like approaches to learning, such as their learning style. In this paper, we focus exclusively on adaptation on a moderate level of granu- larity that is consistent with one specific kind of adaptation sug- gested by the COPES model of studying (Winne & Hadwin, 1998): Adaptation that addresses coordinating activities across several stages of studying and results in large scale adjustments of learning. One example of this kind of adaptation is learners’ between-task adaptation to task complexity. In the current study we investigate this phenomenon by confronting learners with three tasks of differing complexity and by investigating their between-task changes in their learning processes. Independent of the assumed level of granularity, all conceptu- alizations of adaptation propose that it is based on learners’ self- monitoring and learners’ self-regulation. Metacognitive self- monitoring implies bottom-up information processing; learners monitor object-level information such as their own knowledge, their cognitive operations, their learning processes, or their prod- ucts of learning (Nelson & Narens, 1994). The COPES model posits that during monitoring learners compare these perceptions with their internal standards for learning and that this comparison results in evaluations (Winne & Hadwin, 1998). Learners’ internal standards can be influenced by monitoring processes: Within the first stages of learning, learners perceive potential constraints, available resources, and the given goal of the task by monitoring external conditions such as task complexity. As a result, learners * Corresponding author. Tel.: þ49 251 8331386; fax: þ49 251 8339105. E-mail addresses: pieschl@uni-muenster.de (S. Pieschl), elmar.stahl@ph- freiburg.de (E. Stahl), tmurray@cs.umass.edu (T. Murray), bromme@uni-muenster. de (R. Bromme). Contents lists available at SciVerse ScienceDirect Learning and Instruction journal homepage: www.elsevier.com/locate/learninstruc 0959-4752/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.learninstruc.2011.08.005 Learning and Instruction 22 (2012) 281e289