New is not always better: Virtual reality does not necessarily enhance mnemonic processing

Abstract

Integrating new technologies such as Virtual Reality (VR) can contribute to increasing efficiency in several areas relevant to society. VR can be applied in various contexts and has the potential to improve mnemonic processes and memory performance. However, the specific conditions under which VR is more beneficial than conventional learning methods remain unclear. To further investigate the value of VR for mnemonic processing, participants performed a memory task under three different conditions. For that task, they were presented with rules regarding the spatial arrangement of building blocks with a written text or a video in 2D on a screen or in 3D/360° with a head-mounted display. Following the learning session, memory performance was measured by a recognition test involving a multiple-choice questionnaire, in which participants had to mark the correct arrangement of building blocks, and a construction test, in which they had to arrange five different building blocks according to the rules learned. Additionally, participants had to arrange 38 building blocks according to the rules in a free recall test the following day. Surprisingly, results revealed no superiority effect for learning in VR. Instead, learning the rules with the text yielded the best memory performance results, indicating that prior experience with conventional learning methods facilitates declarative knowledge acquisition. Considering previous findings regarding cognitive processing in VR, our results suggest that in passive learning, processing the more salient and personally relevant virtual stimuli in the surrounding VR environment requires more attentional resources. Therefore, VR impairs focusing on the relevant declarative information and impedes the transfer of the learned knowledge to different contexts. When considering to implement VR, the value to the particular domain and specific learning task should be taken into consideration: For learning basic declarative information without actively involving the students, conventional learning methods seem sufficient and more efficient for mnemonic processing compared to new technologies.