Nidhi Krishna Edakkuni Warriem
I’m a UX Researcher and Designer driven by understanding user behaviour and creating intuitive digital experiences.
I’m a UX Researcher and Designer driven by understanding user behaviour and creating intuitive digital experiences.
The study examined how first person motor mental imagery in IVR improves STEM learning. Through a mixed method study design, 38 students tested two IVR prototypes of an abstract science concept, one with mental imagery prompts and one without. Results showed imagery prompts improved immediate understanding but not long term recall. This research fills gaps in applying mental imagery in IVR for higher education and provides design guidelines for using imagery prompts in learning environment.
STEM (Science, Technology, Engineering, Maths) subjects are vital for innovation and future, yet most higher education students struggle with learning abstract concepts because they hold insufficient mental models, which often leads to persistent misconceptions (Belland, 2017; Chabay & Sherwood, 2006; Georgiou & Sharma, 2020; Törnkvist et al., 1993). Traditional teaching methods often lack the depth needed to fully support the conceptual change. Computer assisted teaching methods can support in creating mental models(Belland, 2017). Building on this, immersive virtual reality (IVR) is seen as an effective way as it enables students to interact with intangible scientific phenomena in an embodied manner (Radianti et al., 2020). Despite the potential, IVR’s effectiveness in higher education is mixed, many studies focussing more on user perceptions rather than learning outcomes (Coban et al., 2022). At the same time, first person motor mental imagery which is the ability to generate mental representations has shown to support memory and understanding for abstract concepts. However, the role of it in IVR based STEM learning remains underexplored.
The main aim of the study is to design and evaluate the effectiveness of conceptual understanding of an abstract STEM concept through an IVR based STEM simulation using first person motor imagery prompts to one that doesn’t.
RQ1) Does embedding first person motor imagery prompts in a IVR based STEM simulation improve learners conceptual understanding compared to IVR without such prompts?
RQ2) To what extent does first person motor imagery in IVR enhance learners ability to recall and apply STEM concepts after a delay (1 week)?
RQ3) How does learners self reported imagery ability relate to their conceptual learning and delayed recall in IVR based STEM tasks?
Insights from earlier research guided the choice of embodied cognition and first person motor imagery as the theoretical framework for developing the prototype. Two VR prototypes were developed in Unity (C#) with the XR Interaction Toolkit. ChatGPT was used to generate and refine code, while ensuring the final implementation reflected authors design choices.
The design was based on embodied cognition, which links learning to sensory and motor engagement. It was applied through three embodied design principles(Abrahamson & Lindgren, 2014; Acevedo et al., 2024):
Activities : learners moved objects to explore concepts.
Materials : manipulations produced immediate feedback.
Facilitation : audio, visual, and haptic cues guided learning.
Representation: The background of a virtual scientific lab is designed to increase the presence. Positive charged particle and negative charged particle (cp’s) are represented through red and blue spheres. Electric field lines (EF) are denoted through 3d black lines with arrows indicating the direction of force at the moment. The EF lines are connected to the center of each CP and they change dynamically as participants manipulate the charge or distance of the CP.
Onboarding phase: The aim of this phase is to familiarize participants with the 3D environment and introduce them to basic IVR interactions. Specifically, participants were introduced to basic navigation, grabbing, clicking on UI elements, haptic and audio feedback through simple steps.
Explanation phase: The aim of this phase is to introduce Coulomb’s law to the participants, clear theoretical definition, explanation and formula was introduced with the help of audio and animations. Participants could not interact with the elements in this phase.
Exploration phase: The aim of this phase is to help participants understand the concept further using the principle of embodied cognition. The experience consisted of three sequential scenarios. Participants were tasked to manipulate certain aspect of the simulation for example changing the distance between particles or magnitude of the particle. The objective was to convey the resulting change in force on the test particle, which was communicated through a multi modal feedback system, that included written value, audio and haptic feedback.
The difference between the two prototypes is in the exploration phase. The intervention group’s prototype include first person motor imagery prompts before they proceed to a task. It was implemented as follows :
Mixed-methods, between-groups study with 38 STEM students.
RQ1 Does embedding first person motor imagery prompts in a IVR based STEM simulation improve learners conceptual understanding compared to VR without such prompts?
Result: The intervention group scored higher on the immediate post-test (M = 6.0) compared to the control group (M = 4.65). A Mann Whitney U test confirmed this was statistically significant (p = 0.035). Power 61%.
Discussion: Imagery prompts supported initial conceptual understanding, suggesting that rehearsing actions mentally before execution can enhance learning.
“The pause helped me think ahead about what would happen before I did it (P7).”
RQ2 To what extent does first person motor imagery in IVR enhance learners ability to recall and apply STEM concepts after a delay (1 week)?
Result: The intervention group retained a slightly higher mean score (M = 3.44) than the control group (M = 2.81) after one week, but this was not statistically significant (p = 0.059).
Discussion: While imagery prompts boosted immediate learning, they had no effect on long term retention.
“I remembered the concept during the test, but some details were harder to recall (P12).”
RQ3 How does learners self reported imagery ability relate to their conceptual learning and delayed recall in IVR based STEM tasks?
Result: In the general group, there was no significant correlation between VVIQ and either the immediate test (p = −.187, p = .429) or the delayed test (p = −.342, p = .194). In the intervention group, VVIQ showed a significant positive correlation with the immediate post-test (p = .629, p = .005; post-hoc power = .89), but no significant correlation with the delayed test (p = .456, p = .057).
Discussion: The effect was condition dependent, with imagery prompts acting as a scaffold for learners with weaker imagery ability.
“I used the same visuals provided; the red and blue spheres (P11).”
Mental load
Test Used: Independent sample t-test
Result : An independent sample t-test was conducted to compare the Raw NASA TLX score for both general and intervention group. There was no significant difference in scores for general (M = 35.88, SD = 9.95) and intervention (M = 33.40, SD = 15.40 ; t(36) = -0.59 , p = 0.555, two tailed).
Discussion:
