Virtual Reality, Face-To-Face, and 2D Video Conferencing in Group Work: A Comparative Study on Fatigue, Creativity, Flow, and Interactions

doi:10.21203/rs.3.rs-3413959/v1

This study investigates the impact of different interaction settings— two-dimensional Video Conferencing (VC), Face-To-Face (FTF), and Virtual Reality (VR)—on group dynamics, cognitive performance, and aspects of well-being in a professional context. Utilizing a sample of 40 participants from a large Italian electricity transmission company, the study employs a within-subjects design to explore various metrics, including flow, creativity, fatigue and aspects of interaction. The results indicate that FTF interactions are optimal for idea generation and task absorption. VR, although initially more fatiguing for first-time users, fosters a more collaborative and peaceful environment, encouraging participants to engage more openly with each other. VC was found to be the least fatiguing, but also the least engaging in terms of task absorption and idea generation. Additionally, age-related differences were observed, particularly in the perception of motivational and emotional fatigue in the VR setting. The study provides empirical evidence supporting the integration of VR in professional settings for specific types of meetings, while also highlighting the limitations and areas for future research. These findings have implications for organizational well-being, cognitive ergonomics, and the evolving landscape of remote work technologies.

Biological sciences/Psychology/Human behaviour

Biological sciences/Psychology

The advent of technology has revolutionized the way we work, providing options beyond the traditional physical office space. While two-dimensional remote working Video Conferencing (VC) platforms are increasingly common, advancements in immersive environments in Virtual Reality (VR) are opening new dimensions for work settings. Despite growing research on VC remote working¹, literature comparing it with VR and Face-To-Face (FTF) settings is very limited, as we are describing in this introduction. This study aims to fill this gap by examining decision-making processes under these three conditions.

In the last years, VR technology has seen a significant expansion in its use not only by researchers^2–4, but also by organizations and private companies, for example for training⁵, thanks also to the increase in the accessibility of devices, in economic and diffusion terms. However, this type of technology has existed for many years in more rudimentary forms and its use is a source of interest for researchers in psychology and neuroscience, driven to learn about their potential and their influence on the user who uses them. As reported in Bohil et al.², the goal of the current form of VR is to stimulate brain and behavioral responses in the virtual environment that are analogous to those experienced in the real world. The user immerses himself/herself using hardware components that are usually composed of a specific viewer that allows stereoscopic views of the scene based on position and orientation⁶, obtaining the sensation of his/her point of view. Controllers are used to interact with the environment in which the user is immersed.

A more recent fundamental aspect of VR worlds is to give the possibility to interact with other people while connected from different parts of the physical world. To do this, users must create their own online identity. This is achieved through a three-dimensional representation of themselves, or of an alternative self, i.e. an avatar ⁷. This word comes from Sanskrit, that indicates the human manifestation of a deity on Earth ⁸. The sensation of representing ourselves in an avatar is important to nourish the sense of psychological immersion in what we do in the virtual environment, to support deep levels of collaboration, communication and relationship building ^9,10.

VR has found applications in diverse fields, ranging from gaming and education to military training and healthcare⁴. This technology is used in games and video games ^11,12, in education ¹³, for military exercises and training ^14,15, in architecture for the design of structures ¹⁶, to train and transmit social skills ¹⁷, in medicine to carry out simulations of the different innovative surgical procedures ¹⁸, and to assist the elderly and for different treatments in the psychological field ^19,20. It is increasingly being adopted in corporate settings, but research on its optimal use in such environments is still in its infancy⁵. This study, therefore, aims to contribute to this underexplored area by focusing on factors that significantly impact organizational performance.

A previous comparison between VR and VC in decision making showed that accuracy and time did not differ significantly²¹, while in another study the participants spent a longer average time for the decision making task in VR ²² (however, in this case, the researchers did not perform a statistical test for comparison). In the comparison of the three conditions (VR, VC and FTF) the quality of the discussion, the cognitive load and the probability of solving the task correctly did not differ between the VR and the other two. The social presence in VR was lower than that perceived FTF, but still evaluated with higher levels than in the VC mode ²³.

When it comes to examining the dimension of creativity across different interaction settings, existing literature predominantly focuses on specialized domains such as graphic design or artistic endeavors²⁴. However, our research aims to explore the facets of creativity that are integral to collaborative teamwork, particularly in conventional office settings. To this end, we have focussed on the process of idea generation, a cornerstone of brainstorming activities, and have undertaken a comprehensive analysis of extant studies that delve into this particular aspect.

Empirical evidence suggests that immersive VR offers a distinct advantage over traditional 2D platforms. Participants in VR settings report a heightened sense of playfulness and find the environment conducive to focused activity by minimizing distractions. VR has the potential to bring workers in environments that facilitate more uninhibited interactions and foster a relaxed atmosphere, thereby promoting responsive collaboration. However, it's crucial to note that VR is not without its limitations, such as the absence of some elements of interpersonal communication that are naturally present in FTF interactions²⁵.

In a preliminary study conducted by Petrykowski et al.²² the VR setting yielded a greater number of idea-laden post-its compared to the video call environment. It's important to qualify this finding by noting that the researchers did not perform a statistical comparison to substantiate the observed differences. Ide et al. ²⁶ conducted a nuanced investigation into the role of non-verbal cues—specifically gestures and facial expressions—in collaborative interactions. The study employed a controlled experimental design featuring three distinct conditions: VR with the inclusion of symbolic gestures, VR without symbolic gestures, and traditional FTF interaction. A range of metrics was employed to evaluate the outcomes, including the quantity of ideas generated during brainstorming sessions, the duration of each interaction, and the balance and frequency of participant contributions, which were recorded via microphones. Additionally, the study quantified the utilization of symbolic gestures and collected questionnaire data for further insights. Intriguingly, their findings revealed no statistically significant differences across the three conditions in terms of idea generation, interaction duration, or the balance of participation among team members.

Our research endeavors to fill a gap in the literature by comparing the efficacy of VC and in-person meetings, particularly in the context of creative group interactions. To inform our study design, we conducted an exhaustive review of existing literature, which yielded limited empirical data on this subject. One seminal review by Reiter-Palmon et al. ²⁷ stands out for its comprehensive examination of virtual teams, including an exploration of creativity within these settings. Despite the depth of their review, they highlighted a significant shortfall in empirical studies, noting that much of the existing discourse is theoretical and lacks rigorous evaluation of the factors that might either facilitate or inhibit creativity in virtual environments.

Among the scant empirical studies available, a 2011 paper by Han et al. offers some insights²⁸. This study compared various communication platforms— VC, voice-only calls, text-only interactions, and traditional FTF meetings—to assess their impact on the quality and creativity of ideas generated by groups. However, given that this research was conducted over a decade ago, it is important to consider the technological limitations of VC at that time. Han et al. concluded that there were no significant differences in the quality or creativity of ideas generated across the different platforms, but this finding may warrant re-examination in light of advancements in VC technology.

In a recent study²⁹, researchers conducted a series of experiments to scrutinize the differential impacts of VC and FTF interactions on creativity and decision-making. These experiments specifically focused on brainstorming activities undertaken by pairs of participants. Initial findings revealed that, while couples engaged in video calls generated a quantitatively reduced set of ideas compared to their in-person counterparts, the ideas they did produce were qualitatively superior, as evidenced by higher creativity scores and lower decision error rates. To delve deeper into these results, the researchers formulated a hypothesis positing that the constrained visual field in VC could be a limiting factor for idea generation. The hypothesis suggests that the bounded VR space shared by couples narrows the visual field, an this in turn restricts cognitive scope. Subsequent experiments validated this hypothesis. Couples participating via video calls were observed to focus more intently on their screens, thereby narrowing their visual field. This restricted gaze was not only limited to their partner on the screen but also led to reduced awareness of their physical surroundings. Quantitatively, this was reflected in lower recall scores for objects present in the room. Importantly, these room gaze and object recall metrics were found to be significantly correlated with the number of creative ideas generated, thereby substantiating the hypothesis that a narrowed visual focus could indeed be a cognitive bottleneck.

To bolster the robustness of these findings, the researchers replicated the experiment across five distinct locations with a larger sample size. The results were consistent: VC was associated with fewer generated ideas but higher decision quality. However, it's worth noting that the positive impact on decision quality was somewhat mitigated when controlling for the number of ideas generated, suggesting a nuanced relationship between the medium of interaction and cognitive outcomes.

Our research is designed to explore and compare three distinct interaction settings: VC, VR, and FTF interactions. In the first phase, participants are tasked with generating as many ideas as possible on an assigned problem, utilizing first divergent thinking. This 20-minute phase allowed us an evaluation of creativity, measured by the number of ideas generated and assessed across three dimensions: uncommonness, remoteness, and cleverness. Drawing on prior research, we hypothesize that VR immersion can lead to levels of creativity higher than the other two conditions ²², or comparable to those of FTF encounters²⁶, in terms of the number of ideas generated.

In the second phase, participants are instructed to collaboratively rank the top 5 ideas generated. This 20-minute, timed phase employs a methodology adapted from previous research²⁶ to assess group decision-making during convergent thinking. The level of decision-making is calculated by dividing the time spent by the number of ideas generated. Our hypothesis posits no significant differences in decision-making across the three settings²¹, although video calls may yield distinct outcomes compared to FTF interactions²⁹.

While interaction dynamics have yet to be fully explored, existing studies suggest that VR offers advantages such as enhanced social presence²³ and a more focused, relaxed environment²⁵. We hypothesize that these factors may positively influence interaction, leading to increased collaboration and reduced conflict. To assess this, all sessions were video recorded and later analyzed using the Bales Interaction Process Analysis evaluation grid³⁰.

Given that most participants in this study were using VR for the first time, we anticipated that the novelty may result in perceived difficulty and potentially greater fatigue. To gauge this, participants completed the Zoom Exhaustion & Fatigue Scale³¹.

Furthermore, given that VR elicits emotional responses and presence comparable to the physical world³², we expected that the immersive nature of VR might impact flow, a psychological state characterized by complete absorption and engagement in an activity, leading to enhanced performance and well-being²⁸. Studying flow in this context is particularly relevant as it can offer insights into how different interaction modalities may influence cognitive engagement and task performance. This is measured using the Flow Short Scale³³. Additional questionnaires assess participants' perceived performance using the Perceived Performance Scale³⁴ and their identification with their group via the Single-Item Social Identification³⁵.

In this study, we examined the impact of different interaction settings on decision-making and creativity. A total of 56 participants were randomly assigned to one of three conditions: FTF, VC, and VR. Each group engaged in a 20-minute brainstorming session to generate ideas on an assigned problem, followed by a 20-minute phase to collaboratively rank the top 5 ideas. Creativity was assessed using three dimensions: uncommonness, remoteness, and cleverness. Decision-making efficacy was calculated by dividing the time spent on ranking by the number of ideas generated. Participants also completed 4 questionnaires: the Zoom Exhaustion & Fatigue Scale, the Flow Short Scale, the Perceived Performance Scale, and the Single-Item Social Identification.

As depicted in Figure 1 and Figure 2, the graphs provide a comprehensive visualization of the outcomes across all investigated variables, which are denoted in the respective corners of the radar plot. To facilitate cross-variable comparisons, data normalization was employed, given that the constructs under investigation were assessed using disparate measurement scales. For variables lacking a predefined scale—such as the number of ideas generated—the normalization was executed based on the maximum observed value for that specific variable.

Figure 1 encapsulates the outcomes from questionnaires focused on Flow experience, perceived performance, social identification, and the Zoom Exhaustion & Fatigue Scale. Conversely, Figure 2 encompasses the analyses pertinent to creativity, decision-making, and the four dimensions scrutinized via the Bales Interaction Process Analysis grid³⁰.

Fig. 1: Multidimensional Analysis of Psychological and Cognitive Metrics Across Interaction Settings This figure graphically represents the mean scores for each of the three interaction conditions—face-to-face (FTF, represented by the green line), video call via Teams (represented by the red line), and virtual reality (VR, represented by the blue line)—across various psychological and cognitive constructs. The constructs included are Flow experience, Perceived Performance, Social Identification, and Zoom Exhaustion & Fatigue Scale. Data normalization was applied to account for different measurement scales across the constructs.

Fig. 2: Comparative Analysis of Creativity, Decision-Making, and Interaction Dynamic This figure illustrates the mean scores for each interaction condition—face-to-face (FTF, represented by the green line), video call via Teams (represented by the red line), and virtual reality (VR, represented by the blue line)—in relation to creativity, decision-making, and the four quadrants assessed by the Bales Interaction Process Analysis grid³⁰. Data normalization was performed to ensure comparability across variables that were measured on different scales.

Table 1 presents the means and standard deviations for each dimension across the different types of interaction settings. Tables 2 and 3 delineate the outcomes of the statistical analyses conducted. Specifically, Table 2 contains the results for data samples conforming to a normal distribution, for which Analysis of Variance (ANOVA) and t-tests were employed. Conversely, Table 3 features the results for non-parametric distributions, analyzed using Friedman and Wilcoxon tests. In both tables, the significance level of the p-value is denoted by asterisks.

	Video Conference		Virtual Reality		Face-to-face
Variables	M	SD	M	SD	M	SD
FSS
Flow experience	5.21	0.86	5.14	1.07	5.55	0.70
Perceived outcome importance	2.38	0.88	3.12	1.14	2.85	1.14
Fluency of performance	5.24	1.05	4.83	1.34	5.4	0.94
Absorption by activity	4.55	0.88	5.05	0.95	5.12	0.68
ZEFS
General	1.5	0.74	2.39	1.11	1.76	0.73
Social	1.97	0.96	2	1.05	2.09	1.09
Motivational	1.83	0.89	2.06	1.01	2.04	1.04
Emotional	1.35	0.76	1.59	0.75	1.64	0.82
Visual	1.49	0.65	2.33	1.02
PPS
Perceived performance	4.91	1.05	4.67	1.3	4.88	1.25
Different group	1.97	1.08	2.27	1.57	2.29	1.45
SISI
Single-item social identification	6.26	1.07	6	1.21	6.18	0.94
Number of ideas	15.6	5.1	16.8	7.77	16.2	5.03
Evaluation of ideas
Uncommon	3.44	0.21	3.35	0.3	3.44	0.28
Remote	3.05	0.36	2.99	0.26	3.06	0.31
Clever	3.66	0.25	3.55	0.14	3.57	0.13
Decision making	40.94	16.64	38.44	15.62	32.85	16.29
IPA
Social emotional Area: Positive Reactions	76.9	16.18	97.4	22.55	75	9.71
Task Area: Attempted Answers	82.1	12.82	101.3	15.55	113.6	14.78
Task Area: Questions	15	8.58	24	10.02	13.4	2.95
Social emotional Area: Negative Reactions	10.9	5.93	8	4.29	14.1	5.92

Table 1: Descriptive Statistics for Interaction Conditions

This table presents the means and standard deviations for all measured dimensions, segmented by each type of interaction condition: face-to-face, video conference and virtual reality. M stand for mean, and SD for standard deviation.

	VC x VR x FTF			VC x VR				VR x FTF					VC x FTF
Variables	F	P		M diff.	T	P		M diff.	T	P		M diff.		T	P
FSS
Perceived outcome importance	7.781	0.0008	***	-0.71	-4.343	0.0001	***	0.36	1.288	0.2055		-0.35		-2.458	0.0186	*
Fluency of performance	3.804	0.0266	*	0.48	1.551	0.1291		-0.60	-2.574	0.0141	*	-0.12		-1.213	0.2326
Absorption by activity	7.147	0.0014	**	-0.49	-2.480	0.0177	*	0.05	-0.432	0.6682		-0.44		-4.837	2.207e-05	***
PPS
Perceived performance	0.776	0.464		0.24	1.062	0.2951		-0.21	-1.108	0.2747		0.02		0.118	0.9070
Creativity
Number of ideas	0.123	0.885		-1.20	-0.513	0.6200		0.60	0.199	0.8463		-0.60		-0.347	0.7362
Decision making	1.038	0.375		2.51	0.344	0.7389		5.59	1.063	0.3156		8.10		1.882	0.0926	.
IPA
Social emotional Area: Positive Reactions	4.757	0.0219	*	-20.50	-2.148	0.0603	.	22.40	2.672	0.0255	*	1.90		0.330	0.7491
Task Area: Attempted Answers	18.380	4.49e-05	***	-19.20	-3.191	0.0110	*	-12.30	-2.317	0.0457	*	-31.50		-7.442	3.926e-05	***
Social emotional Area: Negative Reactions	3.116	0.0689	.	2.90	1.174	0.2704		-6.10	-2.690	0.0248	*	-3.20		-1.238	0.2471

Table 2: Parametric Statistical Analysis Results This table displays the outcomes of parametric statistical analyses, including ANOVA and t-tests, conducted on data samples with a normal distribution. Significance levels are indicated by asterisks next to the p-values. False Discovery Rate (FDR) correction was applied for multiple comparisons. VC stands for Video Conference, VR for Virtual Reality, and FTF for Face To Face interactions. M diff stands for Mean Difference.

	VC x VR x FTF			VC x VR				VR x FTF					VC x FTF
Variables	Chi-squared	P		M diff.	V	P		M diff.	V	P		M diff.		V	P
FSS
Flow experience	2.430	0.2968		0.13	391.5	0.7662		-0.38	203.5	0.0260	*	-0.25		168.5	0.0099	**
ZEFS
General	26.922	1.426e-06	***	-0.89	73.5	0.0002	***	0.63	444.0	0.0035	**	-0.26		70.5	0.0076	**
Social	2.346	0.3095		-0.03	149.0	0.9885		-0.09	165.0	0.5715		-0.12		152.0	0.5570
Motivational	1.863	0.3939		-0.23	160.0	0.2156		0.02	203.5	1.0000		-0.21		124.5	0.1201
Emotional	6.615	0.0366	*	-0.24	60.0	0.0955		-0.05	124.5	0.6906		-0.29		50.0	0.0132	*
Visual				-0.84	66.5	0.0002	***
PPS
Different group	0.640	0.7264		-0.30	141.5	0.5792		-0.02	123.0	0.9218		-0.32		99.5	0.2449
SISI
Single-item social identification	2.355	0.3081		0.26	98.0	0.1163		-0.18	46.5	0.4445		0.08		48.0	0.4644
Evaluation of ideas
Uncommon	1.800	0.4066		0.09	20.0	0.3525		-0.09	8.0	0.3525		0		16.0	0.8336
Remote	0.200	0.9048		0.06	16.0	0.7998		-0.07	10.0	0.5541		-0.01		17.0	0.9442
Clever	2.600	0.2725		0.11	21.0	0.2719		-0.02	13.0	0.9326		0.09		28.0	0.1834
IPA
Task Area: Questions	15.436	0.0004	***	-9.00	0	0.0058	**	10.60	55.0	0.0059	**	1.60		24.0	0.9055

Table 3: Non-Parametric Statistical Analysis Results This table showcases the results of non-parametric statistical analyses, specifically Friedman and Wilcoxon tests, conducted on data samples with non-normal distributions. Significance levels are denoted by asterisks adjacent to the p-values. False Discovery Rate (FDR) correction was applied for multiple comparisons. VC stands for Video Conference, VR for Virtual Reality, and FTF for Face To Face interactions. M diff stands for Mean Difference.

Flow Short Scale

The analysis yielded several noteworthy findings concerning the perception of the flow state across different interaction conditions.

Flow Experience

Statistically significant differences were observed in the perception of flow between VR and FTF conditions (p=0.026*; Mean Difference= -0.38), as well as between VC and FTF conditions (p=0.0099**; Mean Difference= -0.25). The FTF condition registered the highest mean score, while the VR condition recorded the lowest.

Perceived Outcome Importance

In terms of perceived outcome importance, the VR condition yielded the highest mean, whereas the VC condition had the lowest. A one-way ANOVA revealed a highly significant p-value of 0.0008***. Subsequent pairwise t-tests between VC and FTF (p=0.0186*; Mean Difference= -0.35) and VC and VR (p=0.0001***; Mean Difference= -0.71) were also significant.

Fluency of Performance

For the dimension of fluency of performance, a significant ANOVA p-value of 0.0266* was obtained. The lowest mean was associated with the VR condition, while the highest was observed in the FTF condition. A significant t-test result was found specifically for this pair (p=0.0141*; Mean Difference= -0.60).

Absorption by Activity

Regarding absorption by activity, the VC condition exhibited the lowest mean, while the FTF condition had the highest. The ANOVA yielded a significant p-value of 0.0014. Pairwise t-tests revealed significant p-values for VC vs. VR (p=0.0177*; Mean Difference= -0.49) and VC vs. FTF (p<0.0001***; Mean Difference= -0.44).

Zoom Exhaustion & Fatigue Scale (ZEF Scale)

Non-parametric statistical methods were employed for all dimensions of the ZEFS questionnaire.

General Fatigue

The Friedman test yielded highly significant results for the dimension of General Fatigue (p<0.0001***). Subsequent Wilcoxon tests for pairwise comparisons revealed significant p-values across all three conditions, thereby establishing the means as significantly distinct. The ranking order of the means is as follows: VC (M=1.50), FTF (M=1.76), and VR (M=2.39). The pairwise results are: VC vs. VR (p=0.0002***; Mean Difference= -0.89), VR vs. FTF (p=0.0035**; Mean Difference= 0.63), and VC vs. FTF (p=0.0076**; Mean Difference= -0.26).

Emotional Fatigue

Another dimension that yielded a significant Friedman test p-value was Emotional Fatigue (p=0.0366*). In this dimension, the only pairwise comparison that reached statistical significance in the Wilcoxon test was between VC and FTF (p=0.0132*; Mean Difference= -0.29). The FTF condition registered the highest mean (M=1.64), whereas the VC condition had the lowest (M=1.35).

Visual Fatigue

The dimension of Visual Fatigue was assessed solely in the VC (M=1.49) and VR (M=2.33) conditions. A Wilcoxon test revealed a highly significant p-value for this pairwise comparison (p=0.0002***; Mean Difference= -0.84).

Perceived Performance Scale (PPS)

Perceived Performance

The initial dimension explored by the PPS was Perceived Performance. The ANOVA analysis yielded a non-significant p-value (p=0.464), suggesting that participants, on average, did not perceive any performance disparities across the three interaction conditions. Subsequent t-tests corroborated this finding, as they also produced non-significant p-values.

Different Group

The second dimension, termed Different Group, aimed to assess whether participants believed they would generate either a greater number or higher quality of ideas with alternative group members. For this dimension, a Friedman test was conducted, resulting in a non-significant p-value (p=0.7264). Pairwise comparisons using the Wilcoxon test further substantiated this outcome, as all returned non-significant p-values.

Single-Item Social Identification (SISI)

This questionnaire consists of a single item and investigates the perception of belonging to the group. The Friedman test conducted on the data of the three conditions obtained a non-significant p-value (p=0.3081). Further comparisons between pairs using the Wilcoxon test also all reported non-significant p-values.

Interaction Process Analysis (IPA)

This analytical tool was employed to scrutinize all interactions across the different conditions. Each interaction was categorized into distinct areas of interaction, and several significant findings were observed. These are elaborated upon below.

Social-emotional Area: Positive Reactions

ANOVA analysis for this area yielded a significant p-value (p=0.0219*). The condition with the highest mean frequency of positive interactions was VR (M=97.4), while the FTF condition registered the lowest mean (M=75). Pairwise t-tests revealed a trend towards significance for the VC x VR comparison (p=0.0603; M diff= -20.5) and a significant p-value for the VR x FTF comparison (p=0.0255*; M diff=22.40).

Task Area: Attempted Answers

This area encompasses all participant attempts at providing answers and explanations. The ANOVA analysis yielded a highly significant p-value (p=4.49e-05***). Subsequent t-tests for all pairwise comparisons were also significant: VC x VR (p=0.0110*; M diff= -19.20), VR x FTF (p=0.0457*; M diff= -12.30), and VC x FTF (p=3.926e-05***; M diff= -31.50). Given the significant differences across all pairs, the conditions can be ranked by their respective means: VC (M=82.1), VR (M=101.3), and FTF (M=113.6).

Task Area: Questions

This area exhibited a non-parametric distribution and was thus analyzed using Friedman and Wilcoxon tests. The overall comparison yielded a highly significant p-value (p=0.0004***). Pairwise comparisons revealed significant p-values for VC x VR (p=0.0058**; M diff= -9.00) and VR x FTF (p=0.0059**; M diff=10.60). Given these results and the observed means (VC M=15; VR M=24; FTF M=13.4), it can be concluded that VR significantly outpaces the other conditions in terms of the frequency of questions and requests for clarification.

Social-emotional Area: Negative Reactions

This area includes interactions that generate tension, attempts to dominate, and rejection of others' ideas. The ANOVA analysis produced a p-value approaching significance (p=0.0689). Subsequent t-tests revealed a significant p-value only for the VR x FTF pair (p=0.0248*; M diff= -6.10). VR registered the lowest mean frequency of negative interactions (M=8), while FTF had the highest (M=14.1).

Creativity

Preliminary Analysis: Topic Variability

Prior to assessing the influence of communication settings on creativity, an initial analysis was conducted to evaluate the potential variability in the number of ideas generated across three different thematic areas (see Methods) — namely Tourism (M=13.6), Restaurant (M=19.3), and Pollution (M=15.7). Shapiro's test confirmed the normality of the distributions for these samples. A repeated-measures ANOVA yielded a significant p-value (p=0.0344*). Pairwise t-tests revealed a significant difference between the Tourism and Restaurant topics (p=0.0301; M diff= -5.7). However, this variability was deemed inconsequential for the broader study, as each topic was employed in a balanced manner across all conditions.

Number of Ideas Generated

The mean number of ideas generated for each experimental condition were as follows: VC (M=15.6), VR (M=16.8), and FTF (M=16.2). Given the normal distribution of these data, an ANOVA test was employed for comparative analysis, resulting in a non-significant difference (p=0.885).

Qualitative Analysis: OSF Tool

Subsequently, two independent human raters evaluated each generated idea using the OSF tool, which employs three distinct criteria: Uncommon, Remote, and Clever. The mean ratings for each criterion across the conditions were: Uncommon (VC M=3.44; VR M=3.35; FTF M=3.44), Remote (VC M=3.05; VR M=2.99; FTF M=3.06), and Clever (VC M=3.66; VR M=3.55; FTF M=3.57).

Given the non-parametric nature of these samples, Friedman tests were conducted for comparative analysis. The results indicated non-significant p-values across all three dimensions: Uncommon (p=0.4066), Remote (p=0.9048), and Clever (p=0.2725).

Decision making

To assess the impact of the experimental settings on group decision-making efficacy, we computed a ratio representing the time required to reach a consensus on idea ranking relative to the number of ideas generated in each session. The mean ratios for the three experimental conditions were as follows: VC (M=40.94), VR (M=38.44), and FTF (M=32.85). An Analysis of Variance (ANOVA) was conducted to compare these means, yielding a non-significant p-value (p=0.375).

Correlation and linear regression

Table 4 presents the outcomes of correlation analyses, along with associated p-values, examining the relationship between participant age and responses across all questionnaires within the various experimental conditions. With respect to the questionnaire probing the state of flow, the correlation analyses between its multiple constructs and participant age yielded no statistically significant results.

In the analysis of responses to the Zoom Exhaustion & Fatigue Scale (ZEFS), several noteworthy correlations with age emerged. In the VC condition, three dimensions—General Fatigue (r= -0.2725; p=0.0933), Motivational Fatigue (r= -0.2812; p=0.0829), and Visual Fatigue (r= -0.2835; p=0.0891)—displayed p-values approaching significance, all indicating a negative correlation with age.

For the VR condition, significant p-values were observed in the dimensions of Motivational Fatigue (r= -0.3406; p=0.0339*) and Emotional Fatigue (r= -0.3525; p=0.0278*). Both dimensions exhibited a negative correlation, suggesting that as participant age increased, reported fatigue levels decreased.

In the FTF condition, a significant negative correlation was found for the dimension of General Fatigue (r= -0.4263; p=0.0068**).

Lastly, within the context of the Perceived Performance Scale (PPS), a significant p-value emerged in the linear regression analysis between age and the "Different Group" dimension, but only in the VR condition (r= -0.3766; p=0.0181*).

	VC				VR				FTF
Variables	r	F	P		r	F	P		r	F	P
ZEFS
General	-0.2725	2.9682	0.0933	.	-0.1871	1.3428	0.2540		-0.4263	8.2191	0.0068	**
Social	-0.2087	1.6850	0.2023		-0.1761	1.1844	0.2835		-0.0822	0.2517	0.6189
Motivational	-0.2812	3.1771	0.0829	.	-0.3406	4.8542	0.0339	*	0.0163	0.0099	0.9215
Emotional	-0.1447	0.7918	0.3793		-0.3525	5.2483	0.0278	*	-0.1672	1.0645	0.3089
Visual	-0.2835	3.0588	0.0891	.	-0.1301	0.6022	0.4430
PPS
Perceived performance	-0.1300	0.6356	0.4304		0.0398	0.0588	0.8098		-0.2174	1.8347	0.1838
Different group	-0.0773	0.2223	0.6401		-0.3766	6.1143	0.0181	*	-0.2819	3.1948	0.0821	.

Table 4 This table presents the outcomes of correlation analyses between participant age and responses to various questionnaires administered across different experimental conditions. Each cell contains the correlation coefficient and associated p-value, providing a statistical measure of the strength and direction of the relationship between age and questionnaire responses. False Discovery Rate (FDR) correction was applied for multiple comparisons. VC stands for Video Conference, VR for Virtual Reality, and FTF for Face To Face interactions, ZEFS stands for Zoom Exhaustion & Fatigue Scale, and PPS stands for Perceived Performance Scale.

In the present study we evaluated the differential impacts of three distinct modes of interaction—FTF, VC, and VR—on group work dynamics. Specifically, the study focused on variables such as idea generation, decision-making, and various psychological constructs.

Our data revealed that FTF interactions significantly outperformed the other two modes in enhancing components of the state of flow, including flow experience, fluency of performance, and activity absorption³⁶. Interestingly, despite VC being second in promoting flow experience, VR exhibited a lower average but a higher standard deviation across all constructs. This variability suggests that the VR experience is highly individualized, particularly among users who are unfamiliar with the technology.

The VR setting was associated with higher scores in the construct of perceived importance of the task outcome. This could be attributed to the novelty of the VR experience, which may have heightened participants' sense of commitment³⁷. In contrast, VC sessions scored lower in perceived importance. This could be attributed to the ubiquity of VC in contemporary work settings; indeed, 87.5% of participants reported engaging in more than eight video calls per week. Furthermore, the informal, often domestic settings in which VC sessions occurred may have attenuated the social pressures that typically enhance task commitment.

The environmental context had a significant impact on activity absorption, a key component of the Flow State Scale (FSS). Video recordings of VC sessions revealed frequent instances of participants engaging with their smartphones, thereby diluting their focus on the task at hand. Such distractions were notably less prevalent in FTF interactions and were obviously non-existent in VR settings due to the immersive nature of the technology. Consequently, VR and FTF modalities yielded comparable levels of activity absorption, underscoring the potential of VR to simulate the cognitive engagement typically associated with physical presence.

Drawing upon the findings from the Flow State Scale (FSS), FTF interactions appear to be the most effective modality for fostering both task importance and activity absorption³⁶. However, when such interactions are not feasible, VR emerges as a viable alternative, offering comparable levels of activity absorption and a heightened sense of task importance. Therefore, this data suggest that the modality of interaction has a nuanced impact on various psychological constructs, including perceived task importance and activity absorption. While FTF interactions remain the gold standard, VR offers unique advantages that merit further exploration, particularly as the technology continues to evolve and gain traction in professional settings.

Analysis of the ZEF (Zoom Exhaustion Fatigue) questionnaire revealed that VR was the most cognitively taxing of the three interaction modalities, while VC was the least fatiguing. Responses spanned the entire range of the Likert scale, underscoring the subjective nature of the experience across participants³⁸. Notably, both VR and FTF interactions necessitated physical presence in company offices, thereby imposing additional cognitive and physical demands on participants as in dual tasks³⁹.

Two additional dimensions of fatigue emerged from the ZEF questionnaire: visual fatigue and emotional fatigue. VR was associated with higher visual fatigue, likely exacerbated by participants' unfamiliarity with the headsets. On the other hand, FTF interactions were linked to elevated levels of emotional fatigue, potentially due to the direct social and emotional engagement required in such settings.

Bales Interaction Process Analysis (IPA) model yielded intriguing insights into the socio-emotional dynamics of group interactions. VR outperformed FTF settings in fostering positive socio-emotional behaviors, such as mutual esteem, support, and understanding. Moreover, VR was associated with fewer negative socio-emotional behaviors, suggesting that it may cultivate a more harmonious group environment³⁰.

In terms of task-oriented behaviors, FTF interactions yielded the highest frequency of response attempts, including proposed ideas and clarifications. VR followed closely, significantly outperforming VC in this regard. Furthermore, VR sessions were characterized by a higher frequency of questions, indicating a greater willingness among participants to seek clarification and share ideas³⁰.

Unexpected negative correlations emerged between age and various constructs, including motivational fatigue and emotional fatigue in VR, and general fatigue in FTF settings. These findings suggest that increased familiarity and experience with different interaction modalities may mitigate the perception of fatigue as one ages.

Our study elucidates the nuanced effects of different interaction modalities—FTF, VR, and VC—on group dynamics, cognitive load, and task performance. VR, although a nascent technology with which participants had limited familiarity, emerged as a compelling medium for fostering a peaceful and collaborative environment. This observation is substantiated by higher scores in the positive socio-emotional domain as per Bales Interaction Process Analysis (IPA) model³⁰. Interestingly, this modality also elicited a greater perception of the task's importance, potentially due to the immersive nature of the technology that commands heightened attention and commitment⁴⁰.

However, the VR condition was not without its drawbacks. Participants reported elevated levels of general and visual fatigue, likely exacerbated by their inexperience with the technology. This aligns with existing literature highlighting the cognitive and physiological demands of virtual environments². Conversely, FTF interactions were found to be most conducive for the generation of ideas, corroborated by higher scores in flow experience and activity absorption metrics. This is consistent with theories positing that physical presence enhances cognitive engagement and social cues, thereby facilitating a more effective collaborative process⁴¹.

VC, the most ubiquitous form of remote interaction, was the least cognitively taxing but also the least engaging, as evidenced by lower scores in activity absorption and the frequency of response attempts. This could be attributed to the familiarity and ubiquity of the technology. This familiarity may have attenuated the perceived importance of the task and the social pressures that might otherwise enhance engagement.

Our analysis also revealed intriguing age-related trends. Younger participants reported higher levels of motivational and emotional fatigue in the VR condition, suggesting either heightened sensitivity or perhaps challenges in emotion regulation. In contrast, older participants, who presumably have more experience with FTF meetings, reported lower levels of general fatigue in that condition. These findings hint at the complex interplay between age, experience, and the cognitive and emotional demands of different interaction modalities.

In summary, our data suggest that the choice of interaction modality has significant implications for collaborative work. VR offers promising avenues for fostering positive group dynamics and task engagement but comes with the cost of increased cognitive load and fatigue. FTF interactions remain the gold standard for idea generation and cognitive engagement, while VC serves as a less demanding but also less engaging alternative.

While the present study garnered substantial data from a large participant pool, yielding significant insights into the dynamics of different interaction modalities, it is imperative to acknowledge its limitations and suggest avenues for future research.

One primary constraint pertains to the potential for experimenter bias in the analysis of video recordings. Despite the utilization of a comprehensive manual designed to objectify the categorization of various interventions, the fact that the recordings were evaluated by the study's designers introduces the possibility of confirmation bias. A methodological refinement to mitigate this limitation would be to employ multiple independent raters, although the time-intensive nature of this task rendered it impractical for the current study.

A second limitation arises from the inability to randomize the sequence of experimental conditions fully. The logistical necessity of conducting the VR and FTF sessions consecutively—aimed at optimizing the researchers' travel to corporate locations—may have introduced an order effect that could influence the results.

Additionally, the nascent state of VR technology itself serves as a limitation. While advancements in hardware are likely to yield more user-friendly and less fatiguing devices, the current study had to contend with the limitations of beta-phase software (Horizon Workrooms), which occasionally required participants to recalibrate settings, thereby potentially affecting their responses and task performance.

Lastly, the study was designed with a focus on non-expert users of VR technology. Future research could profitably extend this inquiry to include participants with varying degrees of familiarity with VR, thereby offering a more nuanced understanding of how prior experience with the technology might modulate the observed effects across different cognitive and emotional constructs.

In conclusion, the study provides an explorative initial framework for understanding the complex dynamics of group interactions across different modalities. While each modality presents its unique set of advantages and limitations, the findings offer valuable insights for organizations aiming to optimize collaborative efforts in an increasingly digitized work environment.

Participants

The study's participant pool was drawn from Terna S.p.A., a leading Italian electricity transmission company. Recruitment was facilitated by organizational managers to ensure the formation of heterogeneous yet familiar work groups, thereby isolating the influence of the experimental setting from the variable of unfamiliarity among participants. The initial cohort comprised 56 individuals, organized into 12 groups of four members each. During data collection some participants did not participate in all sessions, therefore their data and those of their groups were excluded. The final sample was made up of 40 people, divided into 10 work groups. The mean age of the participants was 38.03 years, with a standard deviation of 10.9 and an age range of 25 to 62 years. The gender distribution included 22 males and 18 females. A preliminary questionnaire assessed the frequency of video call usage for work-related communication, employing a Likert scale ranging from "0–1" to "8 or more" times per week. Notably, 87.5% of participants reported engaging in "8 or more" video calls weekly, while a mere 5% reported fewer than six.

It is pertinent to note that the participants predominantly qualified as first time users of VR technology. Those with prior VR experience reported limited exposure and unfamiliarity with the Horizon Workrooms software.

All methods were approved by the University of Trento, Human Research Ethics Committee (Protocol 2022-037). The whole procedure was realized in accordance with the Helsinki Declaration. Informed consent was obtained by all participants prior to the experiment. No financial incentives were offered for participation.

Experimental Design

The study employed a within-subjects design framework, wherein each participating group was exposed to three distinct interaction modalities: VC, FTF, and immersive experiences in VR environment. To account for potential order effects, the sequence of these conditions was randomized across groups. However, logistical constraints necessitated that the VR and FTF sessions be conducted consecutively on the same day, thereby precluding the possibility of interspersing the VC condition between them.

For the purpose of interaction analysis, comprehensive video recordings of all sessions were captured. This was facilitated through the utilization of Free Cam 8, a specialized software designed for screen recording, capable of capturing both visual and auditory data.

VC sessions were conducted remotely via Microsoft Teams. Participants had the latitude to join these sessions from locations of their preference, utilizing their personal computing devices. Conversely, FTF and VR sessions were predominantly orchestrated within the premises of Terna S.p.A.

During FTF interactions, the existing technological infrastructure within the meeting rooms was leveraged. This included a large display screen, a webcam, and an array of microphones, which not only facilitated task explanation by the experimenter but also enabled video capture via screen recording software. Participants were seated in a horseshoe-shaped arrangement around a table, within a room specifically designated for meetings by the organization.

For the VR sessions, Oculus Quest 2 headsets were employed, selected for their market prevalence and functional versatility. The VR meeting platform of choice was Horizon Workrooms, a software designed for corporate virtual interactions. Within this platform, a "mountain cabin" setting was selected for the study. Avatars, meticulously designed to resemble participants based on photographic references, were positioned in a manner analogous to their physical seating arrangement to minimize audio reverberation.

Prior to the commencement of each VR session, a 30-minute orientation was conducted by a designated company representative. This orientation was bifurcated into two segments: the first half elucidated the hardware functionalities in a real-world setting, while the latter half was conducted within the VR environment to familiarize participants with the software interface. This was particularly crucial for the effective utilization of Horizon Workrooms' interactive tools, such as the shared whiteboard. Although the primary responsibility for idea documentation was allocated to a single participant, all group members were trained in the tool's usage to mitigate potential technical disruptions, given the beta status of the software.

Three discussion topics were selected for the experimental tasks, each designed to be sufficiently generic to facilitate broad engagement and ideation. These topics were presented with a specific question to be addressed by the participants:

Tourism: “What could be done to improve and/or increase tourism in the Italian state?"
Restaurant: “A restaurant in the city center is experiencing a drop in customers, what can they do to increase them?”
Pollution: “Pollution is a global problem, what could be done to improve the situation?”

The sequence of these topics was also randomized to control for potential order effects, thereby ensuring a balanced distribution of interaction modalities across topics.

Procedure

Each experimental session adhered to a uniform structure, lasting approximately one hour and encompassing the following sequential phases:

Orientation and Topic Assignment: Participants were initially briefed on the specific activity they would engage in following a standard illustrative hands-on procedure, along with the thematic focus designated for that particular session.

Idea Generation: A 20-minute interval was allocated for the uninhibited generation of ideas pertaining to the assigned topic. Within each group, a designated individual was responsible for documenting the emergent ideas. The documentation medium varied depending on the interaction modality: traditional pen-and-paper for FTF sessions, a communal chat feature within several platforms for VC, and a virtual blackboard manipulated via handheld controllers in the VR environment. Importantly, participants were explicitly instructed to abstain from evaluative judgments during this ideation phase, focusing solely on the capture of emergent ideas.

Idea Prioritization: Subsequent to the ideation phase, participants collaboratively engaged in the formulation of a ranked list featuring the top five ideas. This ranking process was to be consensually agreed upon within a maximum time frame of 20 minutes. Upon completion, the group communicated their finalized list to the experimenter, who then terminated the timing mechanism and recorded the elapsed duration.

Questionnaire Administration: Participants proceeded to complete a series of questionnaires, accessed via a hyperlink disseminated through email by the researcher. The estimated time for questionnaire completion did not exceed 10 minutes.

Measures

Below we list all the dimensions evaluated in this research with the related investigation tools used. All the questionnaires used were translated into Italian and completed at the end of each session by the participants.

Flow Short Scale (FSS)

The Flow Short Scale³³ is a self-assessment test that investigates the state of flow. It consists of 13 items, to be answered as much as you agree with a 7-point Likert scale, with the labels corresponding to 1 "Not at all", 4 "Partially" and 7 "Very much". The 4 subscales of the test measure the variables of: flow experience, perceived outcome importance, fluency of performance, absorption by activity.

Zoom Exhaustion & Fatigue Scale (ZEF Scale)

The Zoom Exhaustion & Fatigue Scale³¹ is a self-assessment test used to evaluate the perception of fatigue during the use of digital communication tools, in particular in a videocall. The test consists of 15 items, which investigate 5 dimensions of fatigue: general, social, emotional, visual, and motivational fatigue. The respondent indicates how much he feels the sensation indicated on a 5-point Likert scale, indicated with the labels: "Not at all", "Slightly", "Moderately", "Very" and "Extremely".

The questionnaire was adapted for each type of session. For the FTF session it was decided to exclude the 3 items investigating visual fatigue, as no device with a screen was used for the FTF activity. Then the visual fatigue variable was only compared between the VR and VC sessions.

Perceived Performance Scale (PPS)

The Perceived Performance Scale³⁴ is a self-assessment test, it is used to assess the participants' perception of their performance during the activity. The questionnaire consists of 5 items. Participants rate how much they agree with the statements on a 7-point Likert scale. In the extremes are the labels “Strongly disagree” and “Strongly agree”. The first three items specifically investigate how much the participants perceive they were productive during the activity. The last two items instead investigate the weight that the participants attribute to the other group members on their group performance, asking them if they think they would have generated more ideas or if these would have been of higher quality with different group members.

Single-Item Social Identification (SISI)

The Single-Item Social Identification³⁵ is a self-assessment test. The item reports "I felt part of the group", to which the participants respond by indicating the value on a 7-point Likert scale, with 1 "Not at all" and 7 "Completely" indicated at the extremes.

Interaction Process Analysis (IPA)

To investigate the dynamics of group interaction, the Bales Interaction Process Analysis evaluation grid was used³⁰. The researcher observed the video recordings of all the sessions and whenever any form of interaction occurred between the members of the group this was reported in one of the 12 categories that make up the grid. In this way we obtain the frequency of implementation of that behavior. The categories are then grouped into 4 survey areas: "Positive Socio-Emotional Area", "Negative Socio-Emotional Area", "Task Orientation Area: response attempts", and "Task Orientation Area: questions".

Creativity

The creativity component was evaluated by comparing the average number of ideas generated for each type of session. A second analysis was then done using the OSF tool for the assessment of divergent thinking. Two researchers assigned a score to each idea generated for the three variables:

Uncommon: indicates how unique the idea is compared to the other ideas generated by the sample. For example, a score of 5 indicates that no one else has reported the same thought
Remote: refers to how much that idea deviates from common thought in general, how remote it is
Clever: the qualitative level of the idea is evaluated, in terms of complexity, formulation and intelligence

Decision making

Decision making within the groups was evaluated calculating the ratio between the time required to draw up the ranking of the 5 best ideas and the number of ideas produced in that session. This allowed us to obtain the average time required by the group to evaluate each idea that emerged, then for the comparison we considered the means of all the sessions for each condition.

Data analysis

All statistical computations were executed utilizing R software (version 4.2.3) and Microsoft Excel.

Prior to conducting comparative analyses across experimental conditions, an initial assessment of data normality was performed. Specifically, the Shapiro-Wilk Normality Test was applied to each data subset to ascertain the distributional characteristics and identify any non-parametric distributions. This preliminary step informed the selection of appropriate statistical tests for subsequent analyses.

For data subsets conforming to parametric distributional assumptions, a repeated-measures Analysis of Variance (ANOVA) was employed. This was justified by the repeated-measures design, wherein each participant—and by extension, each group—engaged in identical activities across the three conditions (VC, VR, and FTF). A p-value threshold of less than 0.05 was considered indicative of statistical significance, thereby warranting the rejection of the null hypothesis positing equivalence across conditions. To control for Type I error arising from multiple comparisons, False Discovery Rate (FDR) correction was applied to all resulting p-values. Post-hoc pairwise comparisons were conducted using t-tests for dependent samples, focusing on the following condition pairs: VC-VR, VR-FTF, and VC-FTF.

Conversely, for data subsets exhibiting non-parametric distributional characteristics, the Friedman test was utilized as a non-parametric alternative to ANOVA for repeated measures. Subsequent pairwise comparisons were conducted using the Wilcoxon Signed-Rank Test for paired samples, with the resulting p-values serving as indicators of statistical significance. Test for paired samples, and these p-values were likewise adjusted using FDR correction to ascertain statistical significance.

Additionally, Pearson correlational analyses were conducted to explore the relationship between participants' age and their questionnaire responses across different session types. Regression analyses were performed using Microsoft Excel, with p-values less than 0.05 deemed statistically significant.

Author contributions

G.M. and N.D. designed the study. G.M and N.D. carried out the experiment. G.M. performed the analysis. G.M and N.D. wrote the manuscript. N.D. supervised the project.

Data availability

The datasets generated and analysed during the current study are not publicly available since participants did not provide explicit written consent regarding the sharing of their data on public repositories, but are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

We would like to express our gratitude to TERNA S.p.A. for their invaluable support and collaboration in facilitating this research with their workers. Their willingness to provide access to their facilities and assist in participant recruitment has been instrumental to the successful execution of this study. We also wish to express our sincere appreciation to MIDA SB S.p.A for their indispensable role in the organizational aspects of the experiment.

Yang, L. et al. The effects of remote work on collaboration among information workers. Nat. Hum. Behav. 6, 43–54 (2022).
Bohil, C. J., Alicea, B. & Biocca, F. A. Virtual reality in neuroscience research and therapy. Nat. Rev. Neurosci. 12, 752–762 (2011).
Parsons, T. D., Gaggioli, A. & Riva, G. Virtual Reality for Research in Social Neuroscience. Brain Sci. 7, 42 (2017).
Cipresso, P., Giglioli, I. A. C., Raya, M. A. & Riva, G. The Past, Present, and Future of Virtual and Augmented Reality Research: A Network and Cluster Analysis of the Literature. Front. Psychol. 9, 2086 (2018).
Naranjo, J. E. et al. A Scoping Review on Virtual Reality-Based Industrial Training. Appl. Sci. 10, 8224 (2020).
Georgiev, D., Georgieva, I., Gong, Z., Nanjappan, V. & Georgiev, G. Virtual Reality for Neurorehabilitation and Cognitive Enhancement. Brain Sci. 11, 221 (2021).
Dalgarno, B. & Lee, M. J. What are the learning affordances of 3-D virtual environments? Br. J. Educ. Technol. 41, 10–32 (2010).
Mystakidis, S. Metaverse. Encyclopedia 2, 486–497 (2022).
De Freitas, S. Learning in immersive worlds: A review of game-based learning. (2006).
De Freitas, S. Serious virtual worlds: A scoping study. (2008).
Meldrum, D., Glennon, A., Herdman, S., Murray, D. & McConn-Walsh, R. Virtual reality rehabilitation of balance: assessment of the usability of the Nintendo Wii® Fit Plus. Disabil. Rehabil. Assist. Technol. 7, 205–210 (2012).
Zyda, M. From visual simulation to virtual reality to games. Computer 38, 25–32 (2005).
Englund, C., Olofsson, A. D. & Price, L. Teaching with technology in higher education: understanding conceptual change and development in practice. High. Educ. Res. Dev. 36, 73–87 (2017).
Alexander, T., Westhoven, M. & Conradi, J. Virtual environments for competency-oriented education and training. in Advances in Human Factors, Business Management, Training and Education 23–29 (Springer, 2017).
Liu, X., Zhang, J., Hou, G. & Wang, Z. Virtual Reality and Its Application in Military. IOP Conf. Ser. Earth Environ. Sci. 170, 032155 (2018).
Song, H., Chen, F., Peng, Q., Zhang, J. & Gu, P. Improvement of user experience using virtual reality in open-architecture product design. Proc. Inst. Mech. Eng. Part B J. Eng. Manuf. 232, 2264–2275 (2018).
Schmidt, M., Beck, D., Glaser, N. & Schmidt, C. A prototype immersive, multi-user 3D virtual learning environment for individuals with autism to learn social and life skills: a virtuoso DBR update. in International Conference on Immersive Learning 185–188 (Springer, 2017).
Gallagher, A. G. et al. Virtual reality simulation for the operating room: proficiency-based training as a paradigm shift in surgical skills training. Ann. Surg. 241, 364 (2005).
Freeman, D. et al. Virtual reality in the assessment, understanding, and treatment of mental health disorders. Psychol. Med. 47, 2393–2400 (2017).
Neri, S. G. et al. Do virtual reality games improve mobility skills and balance measurements in community-dwelling older adults? Systematic review and meta-analysis. Clin. Rehabil. 31, 1292–1304 (2017).
Moore, S. M. & Geuss, M. N. Familiarity with teammate’s attitudes improves team performance in virtual reality. PloS One 15, e0241011 (2020).
Petrykowski, M., Berger, P., Hennig, P. & Meinel, C. Digital Collaboration with a Whiteboard in Virtual Reality. in 962–981 (Springer, 2018).
Moser, I., Chiquet, S., Strahm, S. K., Mast, F. W. & Bergamin, P. Group Decision-Making in Multi-User Immersive Virtual Reality. Cyberpsychology Behav. Soc. Netw. 23, 846–853 (2020).
Obeid, S. & Demirkan, H. The influence of virtual reality on design process creativity in basic design studios. Interact. Learn. Environ. 31, 1841–1859 (2023).
Tenorio Morales, M. C., Yamamoto, K. & Tsujino, Y. SyncMeet: Virtual Work Environment for Collaborative Manga Creation. in International Conference on Human-Computer Interaction 518–532 (Springer, 2020).
Ide, M. et al. Effects of Avatar’s Symbolic Gesture in Virtual Reality Brainstorming. in 32nd Australian Conference on Human-Computer Interaction 170–177 (2020).
Reiter-Palmon, R., Kramer, W., Allen, J. A., Murugavel, V. R. & Leone, S. A. Creativity in Virtual Teams: A Review and Agenda for Future Research. Creat. Theor. – Res. - Appl. 8, 165–188 (2021).
Han, H.-J., Hiltz, S. R., Fjermestad, J. & Wang, Y. Does Medium Matter? A Comparison of Initial Meeting Modes for Virtual Teams. IEEE Trans. Prof. Commun. 54, 376–391 (2011).
Brucks, M. S. & Levav, J. Virtual communication curbs creative idea generation. Nature 605, 108–112 (2022).
Bales, R. F. Interaction process analysis; a method for the study of small groups. xi, 203 (Addison-Wesley, 1950).
Fauville, G., Luo, M., Queiroz, A. C. M., Bailenson, J. N. & Hancock, J. Zoom Exhaustion & Fatigue Scale. Comput. Hum. Behav. Rep. 4, 100119 (2021).
Chirico, A. & Gaggioli, A. When Virtual Feels Real: Comparing Emotional Responses and Presence in Virtual and Natural Environments. Cyberpsychology Behav. Soc. Netw. 22, 220–226 (2019).
Landhäußer, A. & Keller, J. Flow and its affective, cognitive, and performance-related consequences. Adv. Flow Res. 65–85 (2012).
Spruyt, J. Gamified Individual Brainstorming: Reaching for the stars. (2017).
Postmes, T., Haslam, S. A. & Jans, L. A single-item measure of social identification: Reliability, validity, and utility. Br. J. Soc. Psychol. 52, 597–617 (2013).
Snyder, C. R. & Lopez, S. J. Handbook of Positive Psychology. (Oxford University Press, 2001).
Bailenson, J. N., Blascovich, J., Beall, A. C. & Loomis, J. M. Interpersonal Distance in Immersive Virtual Environments. Pers. Soc. Psychol. Bull. 29, 819–833 (2003).
Hancock, P. & Warm, J. A Dynamic Model of Stress and Sustained Attention. J. Hum. Perform. Extreme Environ. 7, (2003).
Wickens, C. D. Multiple resources and performance prediction. Theor. Issues Ergon. Sci. 3, 159–177 (2002).
Slater, M. & Wilbur, S. A Framework for Immersive Virtual Environments (FIVE): Speculations on the Role of Presence in Virtual Environments. Presence Teleoperators Virtual Environ. 6, 603–616 (1997).
Hinds, P. J. & Mortensen, M. Understanding Conflict in Geographically Distributed Teams: The Moderating Effects of Shared Identity, Shared Context, and Spontaneous Communication. Organ. Sci. 16, 290–307 (2005).

No competing interests reported.

Virtual Reality, Face-To-Face, and 2D Video Conferencing in Group Work: A Comparative Study on Fatigue, Creativity, Flow, and Interactions

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Flow Short Scale

Zoom Exhaustion & Fatigue Scale (ZEF Scale)

Perceived Performance Scale (PPS)

Single-Item Social Identification (SISI)

Interaction Process Analysis (IPA)

Creativity

Decision making

Correlation and linear regression

Discussion

Methods

Participants

Experimental Design

Procedure

Measures

Flow Short Scale (FSS)

Zoom Exhaustion & Fatigue Scale (ZEF Scale)

Perceived Performance Scale (PPS)

Single-Item Social Identification (SISI)

Interaction Process Analysis (IPA)

Creativity

Decision making

Data analysis

Declarations

References

Additional Declarations

Status:

Journal Publication

Version 1