Abstract:
Formative feedback is considered an important and beneficial instructional strategy for fostering students’ learning (Hattie & Timperley, 2007; Shute, 2008; Wisniewski et al., 2020). Since providing individualized, immediate feedback is time-consuming and difficult to realize, especially in heterogeneous classes or individual learning settings (Lipnevich & Panadero, 2021; Shute, 2008), educational technologies have been discussed as a useful supplement to provide computer-based feedback (Narciss, 2013). However, there are contradictory findings regarding the effectiveness of computer-based feedback (e.g., Mertens et al., 2022). Previous research suggests that the moderating impacts of various factors on the effect of computer-based feedback can explain these inconsistent findings (e.g., Kluger & DeNisi, 1996; Kuklick & Lindner, 2023; Mertens et al., 2022; Narciss, 2008; Panadero & Lipnevich, 2022; Patchan et al., 2016; Strobl et al., 2019). Initial evidence suggests that various feedback-related (e.g., Lachner et al., 2017a; Mertens et al., 2022), learner-related (e.g., Kuklick & Lindner, 2023; Narciss, 2008, 2013), and context-related (e.g., Fyfe & Rittle-Johnson, 2016; Shirah & Sidney, 2023) factors might affect the effectiveness of computer-based feedback.
These influencing factors, however, have not be systematically investigated. Since previous research focused only on investigating individual feedback-related, learner-related, or context-related potential influencing factors, it was still unclear which factors significantly determine the effectiveness of computer-based feedback on learning. Answering this question requires, on the one hand, theoretical conceptualization and, on the other hand, empirical investigation of the factors of the feedback message itself, of learners, and of the orchestrating context. These factors, which are embedded in the learning process, were investigated, both separately and in relation to each other (Narciss & Huth, 2004; Shute, 2008; Strobl et al., 2019). In order to fulfill these requirements, this dissertation adopted a multidimensional perspective. First, important feedback-related (i.e., specificity of feedback, representation of feedback, level of elaboration, level of complexity), learner-related (i.e., prior knowledge, cognitive, metacognitive, and affective-motivational processes), and context-related (i.e., orchestration with prior instruction, sequence) potential influencing factors were identified based on the previous literature and the research findings. Second, a theoretical model was developed that encompasses the learning process using computer-based feedback and that was bordered by these potential influencing factors. Third, these factors were empirically investigated in different subject domains (i.e., writing and physics) and using different methodological approaches (theoretical conception, literature overview, meta-analysis, experimental studies).
In total, this dissertation comprises three articles for publication in international journals. The first is a theoretical paper in which the potential of computer-based feedback tools to enhance learners’ expository writing quality was explored and potential feedback-related influencing factors have been identified. The second article is an empirical paper encompassing a pre-registered meta-analysis to examine the overall effectiveness of computer-based feedback tools in supporting students in improving their expository writing. Furthermore, the impact of the potential feedback-related factors identified in Paper 1 was examined as well as the potential impact of prior knowledge as a learner-related influencing factor. The third article focused on empirically investigating the impact of further potential learner-related and feedback-related influencing factors as well as on context-related factors related to students’ learning in the domain of physics. Therefore, three pre-registered experiments were conducted that included manipulation of the orchestration of computer-based feedback with strategy instruction. Experiment 1 examined corrective feedback; Experiments 2 and 3 focused on elaborated feedback. In addition, Experiment 3 investigated the potential effect of the sequence in which computer-based feedback and strategy instruction were provided. In summary, through this multidimensional perspective investigating the effectiveness of computer-based feedback and how it is affected by feedback , learner-, and context-related factors, it was possible to adapt the new theoretical model adequately and provide theoretical implications and suggestions for future research accordingly. In addition, based on new evidence, this dissertation was able to provide recommendations for practice regarding what to consider when implementing and orchestrating computer-based feedback to foster learning effectively.