Reflection Principle: Difference between revisions

From ECT wiki
Line 6: Line 6:


=== Design Implications ===
=== Design Implications ===
Methods of reflections will typically not lead to generative processing of the learning material when reflections organized pictorially into mental maps, tables, graphic organizers, etc.  do not provide anchorage to the learning content just covered. A situation like this can be found in John Green’s Crash Course Psychology lecture, Perceiving is Believing. The video is already filled with distracting (extraneous) animations, taking up cognitive resources that should be prioritized for the learning material. However the video further fails to meet effective design standards by  relying on several bullet points to list the sub topics covered in the video solely by their name in the review ending segment (fig 1.). Additionally, John makes no point to verbally anchor relevant knowledge of the key components of perception, beyond reciting the names of the subtopics. If John Green decided to end the video by asking questions to the audience that were relevant and meaningful to the material discussed in the video, he might be able to muster some generative processing through internal reflection from his audience, but unfortunately he makes no such effort. [[File:Screenshot 2022-12-16 102416.png
Methods of reflections will typically not lead to generative processing of the learning material when reflections organized pictorially into mental maps, tables, graphic organizers, etc.  do not provide anchorage to the learning content just covered. A situation like this can be found in John Green’s Crash Course Psychology lecture, Perceiving is Believing. The video is already filled with distracting (extraneous) animations, taking up cognitive resources that should be prioritized for the learning material. However the video further fails to meet effective design standards by  relying on several bullet points to list the sub topics covered in the video solely by their name in the review ending segment (fig 1.). Additionally, John makes no point to verbally anchor relevant knowledge of the key components of perception, beyond reciting the names of the subtopics. If John Green decided to end the video by asking questions to the audience that were relevant and meaningful to the material discussed in the video, he might be able to muster some generative processing through internal reflection from his audience, but unfortunately he makes no such effort. [[File:Screenshot 2022-12-16 102416.png|thumb|170x170px|'''Figure 1''']]


=== Challenges ===
=== Challenges ===

Revision as of 14:27, 16 December 2022

Overview

The educational reformer John Dewey referred to reflective thought as, “Active, persistent, and careful consideration of any belief or supposed form of knowledge in the light of the grounds that support it, and the further conclusions to which it tends” (1932). The benefits of reflection pointed out by Dewey, has paved the way for educational psychologists like Roxana Moreno and Richard Mayer (2010) to consider reflection as a principle that could promote generative processing ( CTML Generative Processing Assumption). There are several types of methods of reflection. For instance, Elaborative Interrogation is a type of reflective process where students are prompted to answer “why” questions about material they had just read (Moreno et a., 2001; Seifert, 1993). Another type of reflective process, Self-Explanations, involves prompting learners to explain their answers to problems while they learn (Chi, de Leeuwenhoek, Chou & La Vancher, 1994). One important idea of the reflection principle is that learners who interact with multimedia environments intended to utilize the Feedback and Guided-Learning principles will vary in their generation of deep learning based on how they manage their cognitive resources through reflecting on their actions and feedback (Moreno & Mayer, 2010). In addition, the reflection principle predicts that forms of instruction that encourage students to reflect on correct solutions will likely lead to more meaningful learning than non-interactive forms of instruction that do not promote reflection.

Evidence

Moreno (2009) found that teachers and students had higher scores on a transfer test when prompted verbally to analyze and reflect on the connections between classroom animations and learned principles. To add, teachers and students who participated in the study displayed higher motivation to learn and took less time to observe animations than participants in the control group who were not prompted.

Design Implications

Methods of reflections will typically not lead to generative processing of the learning material when reflections organized pictorially into mental maps, tables, graphic organizers, etc. do not provide anchorage to the learning content just covered. A situation like this can be found in John Green’s Crash Course Psychology lecture, Perceiving is Believing. The video is already filled with distracting (extraneous) animations, taking up cognitive resources that should be prioritized for the learning material. However the video further fails to meet effective design standards by relying on several bullet points to list the sub topics covered in the video solely by their name in the review ending segment (fig 1.). Additionally, John makes no point to verbally anchor relevant knowledge of the key components of perception, beyond reciting the names of the subtopics. If John Green decided to end the video by asking questions to the audience that were relevant and meaningful to the material discussed in the video, he might be able to muster some generative processing through internal reflection from his audience, but unfortunately he makes no such effort.

Figure 1

Challenges

Bruning, et al. (2004) theorizes that the reflection principle will not apply to learners who already have the necessary skills of planning, monitoring, and evaluating their performance (self-regulated learners). In this case, they hypothesize that reflection prompts could actually get in the way of learning by providing an extra extraneous cognitive load. Further research on how self regulation, reflection, and the three types of cognitive loads work with each-other could provide fresh insight into this concern.

References