Instructional Science

Although digital slides (e.g., PowerPoint) have become a prevalent instructional tool in university settings, few studies have evaluated their efficacy by examining distinctions in students’ note-taking habits and working memory capacities. The present study explored how the learning outcomes of undergraduate students differ under varying conditions of slide accessibility (full, none, partial), taking into account their individual differences in note-taking and working memory. To this end, 240 samples were collected from six repeated sessions with 40 undergraduates. Reviewing notes was permitted before conducting tests to assess the external-storage effect. Types of note-taking were categorized into words and markers to measure their quantity: words were further sorted into words related to slides and words extending beyond slide content, while markers were divided into structuring, emphasizing, connecting, and summarizing. Working memory capacity was gauged using operation span and reading span tasks. Results from a mixed effects analysis indicated a significant association between full access to slides and higher recall test scores, whereas partial access was linked to superior scores in the higher-order performance test. The study found that both types of note words were significant predictors of learning success. However, the positive impact of note words was moderated negatively by connecting markers and reading span scores. This research, with its detailed conditions and covariates, sheds light on why previous studies have produced varied outcomes.

Language in educational contexts is characterized by complex and cognitively demanding features that can be challenging to use. Based on situated expectancy-value theory (SEVT), we assumed that these linguistic demands can lower students’ expectancies of performing well and the intrinsic value they place on tasks. This is particularly true of students with lower language abilities. Consequently, they may be less motivated to actively engage in academic tasks, potentially leading to lower academic achievement. To test this assumption, we linguistically varied an instructional statistics video into three conditions (easy, moderate, and difficult) and randomly assigned a total of 123 pre-service teachers to each condition. We measured their expectancies of success and intrinsic task value halfway through the instructional video and conducted an achievement test after the instruction. Drawing on path analysis, our results showed that different linguistic conditions had no significant effect on the students’ expectancy-value beliefs. However, we found a significant positive effect of language ability and a significant negative effect of the interaction between the linguistically difficult instruction and language ability on expectancies of success. Contrary to our expectations, this indicates that high linguistic difficulty is associated with lower expectancies of success among students with increasing language abilities. Nevertheless, this finding emphasizes the importance of considering the fit between contextual and individual features. However, it did not have an indirect effect on student achievement. The findings are further discussed, highlighting their implications for future research and delineating linguistic design in educational contexts.

As video lectures have become an important way for online learning, the learning strategy is one of the important factors that affect the quality of learning. This study investigated the effects of three generative learning strategies (learner-generated explanations, learner-generated drawings, and imagining) on learners’ cognitive load, judgment of learning, learning engagement, learning performance and learning efficiency, compared to the control group.A total of 160 undergraduate students were randomly allocated to one of the four groups (three experimental groups and one control group) and watched the same video lectures about chemistry. Students in the three experimental groups were asked to engage in learning activities tailored to the learning strategy prompts provided in the pause segments of the videos and to complete all measurements. Students in the control group just watched the video without any prompts.Compared with control and imagining, learner-generated explanations and drawings improved engagement, performance, and learning efficiency and showed distinct patterns across intrinsic, extraneous, and germane cognitive load. Judgments of learning improved only for learner-generated explanations, whereas imagining showed no reliable benefits.This study provides empirical evidence to the impact of different generative learning strategies on students’ video learning.

Self-regulated learning (SRL) is a cyclical process essential for activating independent learners. Because SRL does not develop spontaneously, teachers need to learn about and experience its components and effective practices to stimulate students’ SRL. To support teachers in SRL practice implementation, we propose a unique professional development model that combines direct and indirect instruction of SRL. The model is based on Authentic, Interactive, and Dynamic Experiences in SRL context (SRL-AIDE), and involves live-actor simulations, where participants are highly engaged and immersed in the experience.The study aims to shed light on the relationships and characteristics of practices that promote SRL directly by metacognitive strategic support (MCS), and indirectly by activating students’ independent learning with knowledge construction (KC). This is a quantitative study with a quasi-experimental design. Two teachers’ intervention groups with a pre/post design (N = 70) were compared: an experimental group exposed to the SRL-AIDE model focused on MCS vs. a comparison group exposed to a practice program focused on KC. The study examined in-class SRL practice implementation of two measures: explicitness level and duration in the lesson, and investigated the relations between the MCS and KC practices using real-time in-class measurement methods. Differences emerged between the groups in the MCS and KC practices by time. The experimental group showed greater gains in both practices. This study contributes to the field by providing quantitative evidence on the relationships between MCS and KC practices, which were previously described as supportive mainly in qualitative studies, offering new insights into how these practices work together to support SRL development. Implications for teacher education and future studies are discussed.

Teacher training often incorporates observable examples of focal teaching practices – models. Yet there is currently little empirical evidence on the effects of modelling. We tested the effects of video models on trainees’ skills, knowledge, and self-efficacy in relation to using an evidence-based teaching technique: retrieval practice. We recruited 89 first-year trainee teachers, gave them a document containing evidence-based guidance on how to use retrieval practice and then collected pre-test data on how well they were able to do this in a classroom simulator scenario. Participants were then randomised them to one of three groups: an active control group in which they restudied the document (no model), a video model of effective practice, or a similar video model annotated with the underpinning theory. We then collected post-test data in a second simulator exercise. Exposure to video models improved participants’ use of retrieval practice methods relative to no model. However, adding the annotation to the models did not yield additional benefits. Models did not improve teachers’ knowledge or self-efficacy. Findings support the theory that incorporating models in initial teacher training can help new teachers make use of evidence-based teaching practices.

Seductive details, which are interesting but irrelevant digressions included in a learning environment (e.g., fun facts), have been shown to impair the learning performance of students. However, to date, it is unclear whether the amount of seductive detail content is crucial for the extent of this detrimental effect, or whether it is more a phenomenon of presence or absence, as other studies have been inconclusive in this regard. Hence, in the present studies (N = 191 and N = 101), learners were presented with a varying amount of seductive detail content embedded within a learning text. Instead, the number of text interruptions by seductive content was kept constant between conditions. Both studies consistently showed a linear negative effect of the seductive detail amount on recall performance. Furthermore, Study 2 indicated a polynomial trend regarding transfer performance, suggesting that small seductive detail amounts could improve transfer, even though cross-validation showed that this model is rather unstable. Prior knowledge and working memory capacity did not significantly attenuate the seductive details effect. The results imply that the extent of the seductive detail effect is dependent on the amount of seductive detail content presented. Especially the inclusion of large amounts of digressions should be avoided by teachers and instructional designers, while the inclusion of small amounts might have rather negligible effects.