Exploring Usability, Emotional Responses, Flow Experience, and Technology Acceptance in VR: A Comparative Analysis of Freeform Creativity and Goal-Directed Training

Abstract

This study compared two virtual reality (VR) interaction modes and assessed differences in characteristics, namely usability, emotional responses, flow experience, technology acceptance, activity effectiveness, preference, and satisfaction, aiming to gain insights for proposing design suggestions. The two types of VR interaction examined were freeform creativity, which enabled users to autonomously explore and create using the Gravity Sketch drawing program, and goal-directed training, which involved structured task completion by using a VR dumbbell exercise program developed with Unity. This study involved 33 participants and revealed three key findings. First, both VR modes exhibited excellent results in terms of flow experience, technology acceptance, preference, and satisfaction and evoked strong positive emotions. They also demonstrated shared VR advantages, including overcoming screen constraints and facilitating three-dimensional spatial activities. Second, compared with goal-directed training, freeform creativity elicited greater levels of pleasure, evoked more intense emotional responses, and demonstrated superior performance across related indicators. Third, in terms of usability, goal-directed training surpassed freeform creativity, particularly in overall responsiveness, simplicity, and clarity of information presentation, and learnability, underscoring the importance of enhancing usability for complex tasks in VR application design. These findings offer valuable insights for crafting more appealing, practical, and user-friendly VR systems in the future.

1. Introduction

In the contemporary age of continuous technological evolution and innovation, smart interface technology has made notable strides, fundamentally reshaping the interaction between humans and computers. Among these advancements, virtual reality (VR) technology has rapidly developed and is gradually being used more extensively due to its affordability and miniaturization. VR constructs three-dimensional (3D) virtual environments through computer simulation, offering users multiple sensory simulations that mimic real-world settings and objects in 3D space in real time [1,2]. With advancements in computing performance, VR has been applied across various industries and contexts, including gaming, healthcare, training, and design [3,4]. This study specifically investigated the potential of VR to motivate physical stretching and promote physical activity among users.

In common application scenarios, differences can be observed between two types of goal-related interactions: “exploration and browsing” and “goal-directed guidance”. A previous study concluded that applications often integrate elements of both interaction types [5]. For example, travel applications cater to diverse user preferences, with some allowing users to casually explore and browse destinations and others providing structured guidance for users following a step-by-step approach to achieve their travel goals.

Expanding on previous research, this study investigated the “exploration and browsing” type of interaction, which closely resembles the preliminary phase of exploration. VR interfaces provide users the freedom to navigate without specific predefined task objectives and do not urge users to complete tasks within strict timeframes. This allows users the opportunity to casually peruse and search for task objectives in a more relaxed interaction pattern. In this study, participants were provided a VR creation tool, namely Gravity Sketch, to facilitate autonomous exploration and foster creative expression, enabling them to freely manifest their imaginative ideas. Figure 1 illustrates the diverse, spontaneous, and unrestricted nature of user goals in this context.

The second interaction type is termed “goal-directed guidance”, which entails establishing distinct and explicit task objectives for users to accomplish methodically. In this study, a virtual dumbbell exercise game was employed to guide users to complete predetermined postures within a limited timeframe and reach various “achievement points” through a series of various arm movements. To continually motivate users to engage in appropriate stretching activities, the users were incentivized with visual and auditory rewards upon successful task completion. Figure 2 depicts the users’ objectives as specific, unambiguous, and singular in nature.

Despite the allure and entertainment value associated with VR, whether VR motivates users to engage in stretching exercises and achieve enhanced physical activity outcomes remains unclear. Moreover, whether the inherent differences between exploration and browsing and goal-directed guidance lead to variations in usability, emotional responses, flow experience, and technology acceptance requires further investigation. Insights from a comparison of these differences between interface types could help in proposing valuable design recommendations for promoting sustained and effective physical activities.

This study investigated the characteristic differences between the two primary interface interaction styles in a VR application environment, namely exploration and browsing and goal-directed guidance. Specifically, we compared various indicators and activity effectiveness and analyzed the results, which facilitated the identification of advantages and the proposal of design recommendations. The objectives of this study are outlined as follows:

• Comparison: To develop virtual motivational interface prototypes for freeform creativity and goal-directed training, aligning with different task goals of exploration and browsing and goal-directed guidance, and to compare the distinct characteristics of these two interaction types;
• Testing: To compare the usability, emotional responses, flow experience, technology acceptance, user satisfaction, and effectiveness in promoting physical activities between the two VR types;
• Recommendations: To evaluate user characteristics associated with the use of freeform creativity and goal-directed training VR interfaces and to propose practical design recommendations based on the findings. These findings will serve as a reference for both academic research in VR and the VR design industry.

2. Literature Review

2.1. VR Elements and Flow Experience

Krueger introduced the concept of VR as an interface consisting of three elements (3I): Immersion, Imagination, and Interactivity [6]. Immersion refers to the state where users become fully engrossed in a virtual scenario, completing tasks or achieving objectives, and experiencing pleasure and satisfaction while momentarily detaching from the real-world context. Imagination is fostered within VR environments through visual and sensory stimuli, creating a convincing illusion of presence within the virtual space. Interactivity allows users to freely navigate and interact with objects within the VR environment [7].

The concept of flow experience, introduced by Csikszentmihalyi and Csikszentmihalyi [8], describes a state where individuals become fully immersed in an activity, displaying focused attention while filtering out irrelevant perceptions. The hallmark of VR is its immersive nature, wherein artificially created environments evoke a sense of reality [9]. This sense of presence, serving as a mediator of VR’s effectiveness, is as an objective metric for assessing the realism of a VR system [10].

Regarding the measurement of immersion experience, research suggests that it is characterized by a sense of playfulness [11], wherein users become deeply engrossed in an activity and enter a state of flow experience. In this scenario, time seems to stand still and external concerns become trivial, leading to a sensation of time distortion [12]. Novak et al. used control, focused attention, and time distortion as indicators to measure immersion experience [13]. Summarizing these studies, the present study adopted four widely used indicators, namely concentration, control, enjoyment, and a sense of time distortion [14,15], as survey variables. Concentration denotes focused engagement without distraction, whereas control indicates mastery and adaptability to change. Enjoyment encompasses the aspects of fun, entertainment, stimulation, and novelty. Sense of time refers to the perception of time passing quickly or the loss of track of time. A five-point Likert scale was employed in this study to measure these indicators.

2.2. VR Development Software (Unity) and VR Creation Software (Gravity Sketch)

Currently available VR platforms include HTC Vive, Meta Quest series, PlayStation VR, Samsung Gear VR, and Microsoft HoloLens. For development purposes, high-market-share software such as Unity (2023.2.0b17), Unreal (5.4), and CryEngine (5.7) are commonly used. VR and augmented reality experiences require different tools and development environments. For example, developing VR games with Unity often necessitates compatibility with specific head-mounted display brands such as HTC Vive (Taoyuan, Taiwan), Oculus (Menlo Park, CA, USA), and PlayStation VR (Tokyo, Japan). For the experiments conducted in this study, Unity was selected as the development platform due to its widespread adoption and versatility (Figure 3). Specifically, Unity was employed to design a dumbbell exercise challenge program focused on goal-directed training to be operated on HTC Vive Pro headsets. Known for its professional-grade 3D game engine, cross-platform capabilities, performance optimization, and advanced graphics rendering effects, Unity offers a broad application range spanning from mobile games to massive multiplayer online games and from serious games to e-commerce platforms. Moreover, games developed using Unity typically achieve rapid and real-time execution speeds.

Gravity Sketch (6.0), a 3D creation drawing software that employs polygon modeling composed of points, edges, surfaces, and whole elements, features 3D space brush settings, color selection systems, and VR painting controllers, enabling designers to craft their models using both hands (Figure 4). For the purposes of this study, Gravity Sketch was selected as the platform to implement the freeform creativity program.

2.3. VR Usability Testing and QUIS

Usability encompasses aspects of mental cognition, behavioral patterns, and cultural implications shared among humans, all of which can be empirically examined through experiments and tests. Some scholars have explored usability challenges inherent in various VR applications, emphasizing the importance of investigating the usability and ergonomic constraints associated with these technologies. Usability testing methods, including observation of users’ verbal and nonverbal behaviors, have been advocated to gain insights into user experiences [16]. Additionally, subjective ratings and questionnaires have been employed to evaluate dependent variables such as task completion time, usability issues, workload, and satisfaction levels [17,18].

Recent studies have developed evaluation frameworks for assessing the usability of VR interfaces and have devised scales to evaluate the usability and performance of VR applications [19]. The Questionnaire for User Interface Satisfaction (QUIS) is commonly used in usability testing protocols [5,20].

The QUIS was proposed by researchers in the Human–Computer Interaction Lab at the University of Maryland, United States, primarily for measuring users’ subjective satisfaction with human–computer interfaces [21]. The QUIS evaluates the system’s informativeness, visibility, learnability, and functionality. Moreover, the questionnaire and scale can be tailored to suit the specific needs of the research institution [22]. In this study, the QUIS was adapted into a five-point Likert scale for evaluating VR interfaces by using usability metrics such as overall response, interface representation, interface information, and interface learnability.

2.4. Emotional Responses and the Self-Assessment Manikin Emotion Scale

Emotional responses are typically measured using self-report methods due to limitations imposed by experimental equipment and time constraints. A commonly used instrument is the Self-Assessment Manikin (SAM) scale, which measures emotions across three dimensions: pleasure, arousal, and dominance [23]. Fang and Huang modified the SAM scale by removing the less relevant “dominance” dimension to better suit the measurement of emotions in computer interactive interfaces [3]. In the present study, a two-dimensional (2D) SAM scale, comprising a nine-point semantic differential scale, was adopted to measure emotional valence and emotional arousal. These dimensions are described as follows:

Emotional valence: This dimension reflects the degree of pleasure or displeasure experienced [24]. Bradley and Lang replaced pleasure with valence and used positive and negative endpoints [25]. The present study adopted this dimension to assess the positive and negative emotional responses of participants during interaction with a VR interface.

Emotional Arousal: In addition to the direction of emotion (positive or negative), the intensity of emotion also varies. Emotional arousal refers to the physiological and psychological changes induced by external stimuli, leading to changes in facial expressions and bodily reactions. This dimension ranges from calm to excited, depending on the stimulus [26].

2.5. Technology Acceptance Model

Before integrating VR technology devices into sports training, understanding user characteristics and analyzing their behavioral patterns regarding product adoption is crucial. The Theory of Reasoned Action (TRA), proposed by Ajzen and Fishbein, is a sociopsychological framework used for understanding and predicting individual behaviors. According to the TRA, behavioral intentions are influenced by an individual’s attitude toward the behavior and subjective norms [27]. Expanding on the TRA, Fred D. Davis developed the Technology Acceptance Model (TAM) to elucidate and predict individuals’ levels of acceptance of information technology [28].

The TAM is widely employed owing to its simplicity and clarity, which enable direct predictions of behavioral intentions based on users’ perceptions. Based on prior research, this study categorized the primary variables of TAM into five domains [29,30] and adapted them into a scale for the VR experiment by using a five-point Likert scale. The variables are the following:

1.

Perceived usefulness:

This refers to the extent to which a user subjectively believes that employing a particular system enhances job performance. When users perceive that the system is easy to use, the users may be inclined to accomplish more tasks with the same level of effort. Thus, this belief is simultaneously influenced by perceived ease of use and external variables;

2.

Perceived ease of use:

This indicates the degree to which a user believes that using a particular system would require minimal effort;

3.

Perceived enjoyment:

This refers to the extent to which the use of information technology is perceived as enjoyable and interesting;

4.

Attitude toward use:

User attitude toward the use of information technology is influenced by both perceived usefulness and perceived ease of use;

5.

Behavioral intention to use:

The decision to use an information system is determined by the behavioral intention, which is influenced by the user’s attitude toward technology use and perceived usefulness.

3. Research Methods

3.1. Research Design, Subjects, and Materials

In this study, two sets of experimental interface prototypes were designed on the basis of the two VR objectives, namely freeform creativity and goal-directed training. “Freeform Creativity” primarily allows users to express their creativity and explore freely; in this study, participants were asked to engage in freeform drawing. “Goal-Directed Training” allows participants to experience VR dumbbell challenge exercises, characterized by purpose and guidance. Participants underwent both experimental groups in sequence. The detailed experimental steps are described as follows:

• At the beginning, the instructor initiated the VR equipment and introduced the experiment’s precautions and tasks and participants filled out the consent form. This step took about 5 min. Subsequently, the instructor asked participants to put on the VR equipment, addressed any wearing issues, and ensured that participants correctly donned the VR equipment. Under the instructor’s guidance, participants began to familiarize themselves with the VR operations, gradually adapting to the VR equipment. The instructor answered questions and resolved issues during this step, which took about 10 min;
• Participants in the “Freeform Creativity” group used the Gravity Sketch program to freely create for 10 minutes, using a virtual brush to draw anything they preferred. The instructor guided participants in simulating a brush with the controller, using buttons to draw lines, change colors, and depict desired plants or animals;
• In the “Goal-Directed Training” experimental group, participants operated a dumbbell challenge exercise program designed with Unity. Participants were required to complete three specified tasks within 10 min. These tasks included the following: (a) picking up the controller (virtual dumbbell), (b) selecting a level, and (c) completing at least one exercise cycle within the allotted time. The exercise cycle comprised five levels, each involving simple dumbbell stretching exercises performed in five different postures;
• After completing the above steps, participants filled out a questionnaire and participated in a semi-structured interview.

The experiments were conducted on the SteamVR platform operated on Microsoft Windows 10 by using HTC Vive VR headsets, controllers, base stations, and a VR Ready laptop operated by the instructor to start the VR device and monitor the experiment.

Table 1. Experiment groups: freeform creativity and goal-directed training.

Interface Type	Freeform Creativity	Goal-Directed Training
Characteristic	Exploration and browsing	Goal-directed guidance
Content	Creative expression and free exploration	Learning operations, goal confirmation, and task completion
Tasks	Familiarize with the system with instructor assistance. Freely explore, experience, and create using a virtual brush for a duration of 10 min.	Familiarize with the system with instructor assistance. Complete the following three tasks: (a) pick up the controller (virtual dumbbell), (b) select a level, and (c) complete an exercise cycle. The exercise cycle features five levels of simple dumbbell stretching activities performed in different postures.
Operation interface
Software	Gravity Sketch (6.0)	VR dumbbell challenge exercise program designed using Unity (2023.2.0b17)
Execution platform	SteamVR platform operated on Microsoft Windows 10
On-site experimental setup and usage by participants
VR interaction equipment	HTC Vive Pro VR headset, controllers, and base stations
Display and operation devices	Asus ROG laptop operated by the tester with a 15.6-inch Full HD Display, Intel Core i7-7700HQ Processor, and NVidia GTX 1060-6G (Taipei, Taiwan).

lists the required tasks, interface operations, on-site experimental setup, and hardware configurations of the two VR experiment groups.

The study recruited 33 participants, with 12 men (36.4%) and 21 women (63.6%) and a mean age of 24 years. Among the participants, 10 (30.3%) had no prior experience with VR products and 3 (9.1%) had no experience with computer drawing software. This study is part of a 3-year project that underwent a research ethics review. All participants provided informed written consent forms prior to participation in the study.

3.2. Questionnaire Design

For data collection, this study employed a Google online questionnaire format. Prior to the formal experiment, a pilot test was conducted involving five participants. The pilot test participants were guided through the experimental procedure, which took about an hour, while also providing feedback to optimize the process. They then completed the initial version of the questionnaire and group discussions were held to review and refine the questionnaire design. Finally, the questionnaire details and experimental design were revised accordingly. Subsequently, the final formal experiment was conducted with 33 participants. The data collected were compiled and then analyzed using SPSS version 20.0 statistical software. The main statistical analyses comprised basic descriptive statistics and independent samples t-tests.

The questionnaire comprised nine sections: (a) users’ emotional responses, (b) QUIS, (c) flow experience, (d) technology acceptance, (e) self-assessed activity effectiveness, (f) preference, (g) overall satisfaction, (h) reasons for interface advantages and disadvantages, and (i) personal information.

Table 2. Questionnaire framework and content.

Section	Questionnaire Item	Number of Items	Scale Type	Questionnaire Content and Objective
1. Users’ emotional responses	Emotional valence	1	9-point semantic differential scale	To assess the level of pleasure experienced
	Emotional arousal	1	9-point semantic differential scale	To evaluate the intensity of emotion felt
2. QUIS	1. Overall response	4	5-point semantic differential scale	To evaluate the overall response to the interface and the intensity of feeling
	2. Interface representation	3	5-point semantic differential scale	To evaluate the readability, function framework, and continuity
	3. Interface information	3	5-point semantic differential scale	To assess the presentation and clarity of information and system prompts
	4. Interface learnability	3	5-point semantic differential scale	To determine the correct methods, remember names and functions, ease of task implementation
3. Flow experience	1. Concentration	4	5-point Likert scale	To measure the concentration, engagement, and the ability to remain undistracted
	2. Enjoyment	4	5-point Likert scale	To evaluate enjoyment, entertainment value, stimulation, and novelty
	3. Sense of time distortion	4	5-point Likert scale	To assess the sense of time passage and the ability to forget troubles and pending tasks
	4. Control	2	5-point Likert scale	To determine the sense of control and the ability to cope with sudden changes
4. Technology acceptance	1. Perceived usefulness	3	5-point Likert scale	To assess enjoyment in the process, perceived usefulness, and desire for the technology
	2. Perceived ease of use	2	5-point Likert scale	To evaluate the ease of understanding and use
	3. Perceived enjoyment	3	5-point Likert scale	To assess fun and pleasure experienced during use
	4. Attitude toward use	2	5-point Likert scale	To measure the level of enjoyment in engaging with the technology
	5. Behavioral intention to use	2	5-point Likert scale	To determine the perceived worthiness of engagement and the likelihood of future use
5. Self-assessed activity effectiveness		1	5-point semantic differential scale	To assess activity effectiveness based on experience
6. Preference		1	5-point semantic differential scale	To evaluate the level of preference
		2	Short-answer questions	To investigate the reasons for likes and dislikes
7. Overall satisfaction		1	5-point semantic differential scale	To determine the level of satisfaction
8. Reasons for interface advantages and disadvantages	Reasons and opinions on interface advantages	1	Brief-description questions	To conduct semistructured interviews inquiring about the reasons and opinions on the interface advantages
	Reasons and opinions on interface disadvantages	1	Brief-description questions	To conduct semistructured interviews inquiring about the reasons and opinions on the interface disadvantages
9. Personal information	Participant demographics	3	Short-answer questions	To gather information on age and gender
	Participant background	3	5-point Likert scale	To gather information on the exercise frequency and experience with computer drawing software and VR software use

details the components of the questionnaire.

The questionnaire’s first section assessed user emotions by using SAM to examine emotional valence and emotional arousal. Figure 5 provides an illustrative example, depicting question 1-1. Participants were instructed to follow the provided guidelines and select a checkbox item by item. Progression to subsequent questionnaire pages was permitted only after completion of the answers on the current page.

The second section comprised the QUIS, a standard test for usability that was modified to align with the objectives of this study. The third section assessed the flow experience and comprised four indicators evaluated on a 5-point Likert scale: concentration, enjoyment, sense of time distortion, and control. These indicators, respectively, examined the following: 1. focus, engagement, and resistance to external influences; 2. enjoyment, playfulness, excitement, and novelty; 3. perception of time passing quickly, loss of the sense of time, and the forgetting of troubles and other concerns; and 4. ability to manage and adapt to changes.

The fourth section of the questionnaire assessed users’ overall acceptance of VR technology. This section comprised five indicators: perceived usefulness, which evaluated whether users found the process enjoyable, perceived it to be useful, and desired the technology; perceived ease of use, which assessed whether the technology was easy to understand and use; perceived enjoyment, which evaluated users’ experience of fun, enjoyment, and entertainment throughout the process; attitude, which gauged whether users liked engaging with this technology; and behavioral intention, which considered whether users perceived it to be worthy of engagement and whether they intended to use this technology frequently in the future.

The fifth section comprised a scale for self-assessed activity effectiveness, which measured the effectiveness of the activities by using a 5-point semantic differential scale. This assessment focused on evaluating an individual’s perception of physical activity status, considering factors such as perceived effort. The VR dumbbell exercise and creative drawing activities in this study aimed to motivate participants to stretch their limbs and attain a specific level of activity intensity.

The sixth section explored participants’ preferences for the two VR types and asked participants to briefly describe their reasons for liking or disliking the specific VR type. The seventh section assessed participants’ satisfaction with use. The eighth section comprised semistructured interview questions that asked participants about the advantages and disadvantages of the interface and their rationale behind them. The final section gathered participants’ personal data, including age, gender, and VR experience, to investigate whether personal characteristics influenced perceptions regarding the different interface types.

4. Results

4.1. Mean and Standard Deviation for Each Scale

In this experiment, the usability, emotional responses (emotional valence and arousal), flow experience, technology acceptance, activity performance, preference, and satisfaction were compared between participants engaged in freeform creativity and goal-directed training VR activities. Following the completion of both VR tasks, the participants used the SAM scale to rate their emotional valence and arousal levels. Emotional valence referred to the positive or negative feelings of pleasure, which were rated on a nine-point scale, where higher scores (closer to 9) indicated more pleasure and lower scores (closer to 1) indicated less pleasure. Emotional arousal referred to the intensity of emotion aroused in the participants, with higher scores (closer to 9) indicating more intense emotions and lower scores (closer to 1) indicating calmer emotions. The remaining scales were evaluated on a five-point scale, with higher scores (closer to 5) indicating better performance and lower scores (closer to 1) indicating poorer performance.

Finally, the study compiled data from the 33 participants and conducted statistical analyses. Figure 6 presents a line graph depicting the mean scores for each scale across the two VR types and

Table 3. Mean values and SDs for each scale across two VR types: freeform creativity and goal-directed training (Unit: five-point Likert scale).

Item	Goal-Directed Training		Freeform Creativity
	Mean	SD	Mean	SD
Emotional valence	3.92	0.96	4.31	0.92
Emotional arousal	3.18	1.25	4.14	0.86
QUIS	4.21	0.48	3.87	0.69
Flow experience	3.68	0.68	3.98	0.55
Technology acceptance	3.92	0.76	4.21	0.66
Self-assessed activity effectiveness	2.76	1.37	3.18	1.29
Preference	3.61	1.06	4.36	0.74
Overall satisfaction	3.73	0.91	4.24	0.75

presents the mean values and standard deviations (SD). Since emotional valence and emotional arousal were originally measured on a nine-point Likert scale, while the other scales were measured on a five-point scale, we converted the nine-point scales for emotional valence and emotional arousal to five-point scales to facilitate accurate comparisons and avoid misleading interpretations. A preliminary analysis of the descriptive statistics revealed that freeform creativity performed better on most indicators, including emotional valence (4.31, SD = 0.92) and emotional arousal (4.14, SD = 0.86). Regarding flow experience, technology acceptance, preference, and satisfaction, the mean values were 3.98 (SD = 0.55), 4.21 (SD = 0.66), 4.36 (SD = 0.74), and 4.24 (SD = 0.75), respectively. Notably, both VR types achieved a satisfactory mean score of 3.61 or higher on the five-point Likert scale in terms of flow experience, technology acceptance, preference, and satisfaction.

Interestingly, freeform creativity exhibited superior performance in activity effectiveness (mean = 3.18, SD = 1.29) compared with goal-directed training (mean = 2.76, SD = 1.37) on the self-assessed activity effectiveness scale. Although superior performance was observed for other indicators, the scores for usability for both VR types were closer to the median value, with goal-directed training (mean = 4.21, SD = 0.48) outperforming freeform creativity (mean = 3.87, SD = 0.69). Further detailed analysis entailed paired sample t-tests to investigate significant differences between indicators.

4.2. Analysis Results of Each Scale

Table 4. Results of the paired sample t-tests between the two VR types, freeform creativity and goal-directed training (Unit: points; values in parentheses indicate negative values. Emotion scale: nine-point Likert scale; other scale: five-point Likert scale; * p < 0.05).

Pairwise Comparison (Goal-Directed Training vs. Freeform Creativity)	Paired Variable Difference					T Value	Significance (Two-Tailed)
	Comparison of Mean	SD	Standard Error of the Mean	95% Confidence Interval of Difference
				Lower Bound	Upper Bound
Emotional responses
Emotional valence	(0.70)	1.93	0.34	(1.38)	(0.01)	(2.08)	0.046 *
Emotional arousal	(1.73)	2.11	0.37	(2.48)	(0.98)	(4.70)	0.000 *
Usability
Overall response	(0.29)	0.75	0.13	(0.55)	(0.02)	(2.20)	0.035 *
Representation	0.47	0.87	0.15	0.17	0.78	3.13	0.004 *
Information	0.06	0.58	0.10	(0.15)	0.27	0.60	0.553
Learnability	0.82	0.74	0.13	0.56	1.08	6.38	0.000 *
Total QUIS score	0.34	0.56	0.10	0.14	0.54	3.51	0.001 *
Flow experience
Concentration	(0.45)	0.67	0.12	(0.68)	(0.21)	(3.85)	0.001 *
Enjoyment	(0.59)	0.81	0.14	(0.88)	(0.30)	(4.21)	0.000 *
Sense of time distortion	(0.32)	0.64	0.11	(0.54)	(0.09)	(2.86)	0.007 *
Control	0.17	0.83	0.14	(0.13)	0.46	1.16	0.255
Total flow experience score	(0.30)	0.46	0.08	(0.46)	(0.13)	(3.71)	0.001 *
Technology acceptance
Perceived usefulness	(0.57)	1.15	0.20	(0.97)	(0.16)	(2.83)	0.008 *
Perceived ease of use	0.56	0.83	0.14	0.27	0.85	3.89	0.000 *
Perceived enjoyment	(0.43)	0.82	0.14	(0.73)	(0.14)	(3.03)	0.005 *
Attitude toward use	(0.35)	1.09	0.19	(0.73)	0.04	(1.84)	0.075
Behavioral intention to use	(0.65)	0.96	0.17	(0.99)	(0.31)	(3.88)	0.000 *
Total TAM score	(0.29)	0.76	0.13	(0.56)	(0.02)	(2.18)	0.036 *
Self-assessed activity effectiveness, preference, and satisfaction
Self-assessed activity effectiveness	(0.42)	1.44	0.25	(0.93)	0.09	(1.70)	0.100
Preference	(0.76)	1.23	0.21	(1.19)	(0.32)	(3.55)	0.001 *
Satisfaction	(0.52)	1.00	0.17	(0.87)	(0.16)	(2.95)	0.006 *

presents the results of the paired sample t-tests conducted between the two VR types—freeform creativity and goal-directed training—across various scales. A test result was considered statistically significant when the p-value was less than the significance level (set at 0.05). The interpretation of the results for each indicator is detailed as follows.

4.2.1. Emotional Responses

The statistical analysis revealed significant differences between the two VR types in terms of both emotional valence and emotional arousal (p = 0.046 and 0.000, respectively), indicating that participants experienced greater pleasure and stronger emotions with freeform creativity than with goal-directed training. Although a significant difference in pleasure levels was observed between the two VR types, their mean scores exceeded 7 out of a maximum of 9 points, indicating that both VR types provided a positive sense of pleasure. Freeform creativity outperformed goal-directed training in emotional arousal, with a mean of 7.45 (SD = 1.54). Although the mean score for goal-directed training reached 5.73 (SD = 2.25) and exceeded the median value, it elicited lower emotional arousal.

4.2.2. Usability

The statistical analysis revealed a significant difference in usability between the two VR types (p = 0.001), indicating that goal-directed training exhibited superior usability compared with freeform creativity. Further investigation into the underlying indicators revealed that goal-directed training significantly outperformed freeform creativity in the dimensions of overall response, information presentation, and learnability (p = 0.035, 0.004, and 0.000, respectively). These results suggest that goal-directed training VR provided a simpler, more satisfying, and more engaging overall response, with clearer and simpler information presentation and greater learnability.

4.2.3. Flow Experience

The statistical analysis revealed a significant difference in flow experience between the two VR types (p = 0.001), indicating that freeform creativity provided a superior flow experience compared with goal-directed training. Further examination of its constituent indicators indicated that freeform creativity significantly outperformed goal-directed training in terms of concentration, enjoyment, and sense of time distortion (p = 0.001, 0.000, and 0.007, respectively). These results imply that users engaged in freeform creativity VR were able to concentrate better and experience more enjoyment and a greater sense of time distortion, thereby becoming more immersed in the VR activity.

4.2.4. TAM

The statistical analysis revealed significant differences in technology acceptance between the two VR types (p = 0.036), suggesting that freeform creativity exhibited superior overall technology acceptance compared with goal-directed training. Further examination of its constituent indicators suggested that users of freeform creativity VR perceived it to be more useful, easier to use, and more enjoyable and were more willing to continue using it (p = 0.008, 0.000, 0.005, and 0.000, respectively) than goal-directed training VR.

4.2.5. User Activity Effectiveness, Preference, and Satisfaction

The statistical analysis revealed no significant difference in self-assessed activity effectiveness between the two VR types (p = 0.100). Contrary to expectations, goal-directed training, which aimed to motivate users to exercise, did not demonstrate superior activity effectiveness compared with freeform creativity. This outcome prompts further exploration to identify potential influencing factors that were not accounted for in the study.

Regarding preference and satisfaction, the statistical analysis revealed significant differences between the two VR types (p = 0.001 and 0.006, respectively), indicating that users favored one type over the other and reported higher satisfaction levels after use. In the questionnaire, the participants were asked to briefly explain the reasons for their likes and dislikes regarding the two VR types—freeform creativity and goal-directed training. The results are compiled and presented in

Table 5. Participants’ reasons for liking or disliking the two VR interaction types.

VR Interaction Type	Reason for Liking	Reason for Disliking
Goal-directed training	Increases motivation to engage in stretching and physical activities. Enables completing various levels by following the interface requirements and performing the standard postures or movements. Has a novelty factor, is relaxing, and is easy to perform	The weight and grip of the handheld device do not match that of real dumbbells (the VR dumbbells are too light and are not realistic). Setup time is required for simple exercises. Tasks are monotonous and lack challenge.
Freeform creativity	Overcomes screen limitations or limitations related to 2D spaces, allowing freedom for ideation. Transforming freeform drawing into a 3D activity is interesting and more realistic. Using high-tech VR technology for drawing provides a sense of accomplishment, fun, and novelty. Enables drawing in a private space with no disturbances.	Requires continuous wearing of the VR headset. Fully mastering spatial positioning with current hand feel is challenging. Achieve drawing precision is difficult. Drawing functions are unclear and not easy to learn. Operating the drawing tools is complex (e.g., changing brushes and adjusting colors involve numerous steps).

.

5. Discussion

5.1. Main Findings

Both VR types yielded a satisfactory mean score of 3.61 on a five-point Likert scale for flow experience, technology acceptance, preference, and satisfaction. Moreover, results from the nine-point Likert scale indicated that both VR types evoked strong positive emotions. This suggests that both VR modalities offer shared benefits, including transcending the limitations of traditional screens or 2D constraints. They enable users to engage in 3D activities, fostering feelings of novelty, enjoyment, achievement, and immersion in virtual environments.

Freeform creativity outperformed goal-directed training across most indicators, including emotional responses (emotional valence and emotional arousal), flow experience, technology acceptance, preference, and satisfaction, with the differences reaching statistical significance. This implies that creative activities that allow free expression and exploration elicit greater pleasure and evoke stronger emotions. As a result, users perceive such technology as more useful, user-friendly, and enjoyable, fostering a willingness to continue its use. Moreover, freeform creativity enables users to concentrate more, enjoy the process, and experience a distortion of time, thereby facilitating deeper immersion in VR experiences.

Surprisingly, the results regarding activity effectiveness did not indicate a significant difference, although freeform creativity exhibited a slightly higher score than goal-directed training. This finding implies that despite the intention behind goal-directed training to achieve a sustained increase in exercise effectiveness, its inadequate design might have resulted in a failure to consistently meet the expected goals. Interviews conducted post-experimentation revealed the need for additional components, such as participant motivation, music design, incorporation of sound effects, virtual interaction of imagery, and the integration of context and narratives. These elements are crucial for fostering user willingness to engage in stretching activities with a sense of enjoyment.

The advantage of goal-directed training lies in its significantly higher usability compared with freeform creativity. Its underlying indicators illustrate that goal-directed training offers a more straightforward, satisfying, and engaging user experience and improved overall responses. Additionally, information is presented in a clearer and more concise manner, making it easier for users to comprehend and learn. This advantage can be attributed to several factors, such as Gravity Sketch, which features more complex functions beyond simple pen manipulation, such as color selection, brush changes, and even drawing complex surfaces. These features may pose learning challenges for users, potentially leading to frustration. By contrast, the dumbbell exercise task in goal-directed training involves simpler actions, such as picking up virtual dumbbells and repeating various postures, making it more straightforward and easier to learn. However, both VR types garnered mean scores of 3.87 or higher on the five-point Likert scale for usability, underscoring VR’s inherent characteristics of immersion and fun, which may increase QUIS scores, thereby influencing users’ perception of favorable usability.

This study analyzed two types of VR and identified common negative experiences and limitations associated with VR operations. Although VR interactions closely mimic natural human–computer interactions, our findings revealed several usability issues. The more complex the operation, the harder it is to learn. For instance, changing brushes, altering brush colors, or even deleting and editing one’s creative works were reported to cause frustration and confusion among users. Additionally, the differences between VR controllers and real-world operations require adaptation, memory, and learning, leading to further difficulties. While initial VR use may offer novelty, subsequent operations that are too simple, repetitive, or uninteresting can result in negative experiences and reduce the desire to continue. Lastly, despite advancements in technology reducing instances of dizziness, some participants still experienced slight discomfort.

This study suggests leveraging the unique advantages offered by both freeform creativity and goal-directed training to optimize VR user experiences. Understanding the common negative experiences and limitations of VR operations allows for the proposal of appropriate solutions. The following section elaborates on these suggestions in detail.

5.2. Practical Implications

Based on a comprehensive analysis of the statistical results and interview responses, we recommend the following practical strategies and design suggestions to enhance the effectiveness of both freeform creativity and goal-directed training in practical applications.

Exploration and goal: Designers can harness the inherent advantages of VR by transcending 2D limitations and offering users engaging 3D experiences. Leveraging the characteristics of freeform creativity, designers can create stress-free environments that encourage users to participate autonomously. Designers can evoke stronger feelings of pleasure, enhance perceived usefulness and enjoyment, and foster a desire for continued engagement with the VR platform among users. In the case of goal-directed training, the task design should be dynamic and engaging, avoiding monotony and a lack of challenge. By addressing these considerations, designers can prevent user disengagement and ensure that the desired outcomes of VR activities are effectively achieved.

Promoting physical activity: Although the impact of VR on physical fitness outcomes remains unclear, designers are encouraged to prioritize the development of engaging exercise scenarios that encourage sustained physical movement. By crafting engaging exercise experiences, designers can enhance user concentration, increase enjoyment levels, and deepen the sense of immersion in virtual environments. Additionally, designers should explore creative strategies to address the challenge of maintaining user engagement over extended periods. Incorporating elements, such as sound effects, music, and innovative interaction, can promote physical activity and optimize exercise effectiveness within VR environments.

Learning: Enhancing self-learning capabilities should be a central focus in future VR design. In real-world settings, users may not receive immediate assistance for adjusting headsets or navigating VR software. To address this, designers should prioritize the development of self-learning modules and incorporate exploratory activities to enable users to gradually acclimate to VR operations without feeling pressured by time constraints. During VR practice sessions, more cues and guidance should be provided to engage novice users. Additionally, considering the common practice of software updates, designers can leverage user feedback to update VR software features over time. To accommodate users with varying levels of expertise, designers could consider releasing different editions or versions of VR software, such as Lite or basic editions, tailored to different levels of complexity, ranging from novice users to professional designers.

VR equipment improvement: According to interview responses, participants noted a clear distinction between VR controllers and real-life counterparts such as dumbbells or paintbrushes. Developers are encouraged to consider designing customized VR controllers that closely mimic the shape, weight, and functionality of real dumbbells and paintbrushes to enhance the sense of immersion. Furthermore, as technology continues to advance, future developments may simplify complex positioning calibrations or setup procedures. VR equipment is becoming increasingly accessible due to the ongoing miniaturization and decreased cost of its components. Designers should capitalize on these advancements to create more user-friendly and accessible VR equipment, thereby enhancing the overall usability and learnability of VR systems.

Usability Enhancement of Freeform Creativity: The study results highlight the advantages of “ Freeform Creativity”, which provides an environment for users to autonomously participate in activities without pressure. However, as previously mentioned, autonomous exploration also presents usability challenges that need to be addressed. Based on the research results, this study proposes the following design principles: (a) leverage the sense of autonomous and free activity by designing intuitive operations that allow users to discover new functions independently and (b) simplify the user interface to make it more straightforward and intuitive, reducing the learning curve. This can be achieved by using fewer buttons or more intuitive gesture operations, making it easier for users to start, (c) introduce a detailed tutorial mode that enables novice users to gradually learn how to use the tools and functions through interactive step-by-step instructions, (d) provide more personalized setting options, allowing users to adjust tools and interfaces according to their needs and habits. For example, allowing users to customize shortcuts, brush sizes, and color selection methods, and (e) add user guides and tips within the interface, especially when users use certain functions for the first time. Guides should include tips, pop-ups, or messages during specific operations to help users understand how to use various tools and functions.

Engagement and sustainability with VR: This study focuses on immediate usability and emotional responses. However, it is crucial to also consider how to promote users’ long-term engagement and sustainability with VR. According to the results of this study, the design of VR tasks can affect user engagement, particularly if the tasks are too monotonous or lack sufficient challenge. Based on these observations, this study proposes the following design recommendations: (a) provide innovative storylines and engaging activities to motivate users to continue their participation, (b) implement a gradual increase in task complexity to enhance user retention, and (c) utilize appropriate real-time audio feedback, visual stimuli, and innovative interactive methods to strengthen the sense of immersion.

Dizziness problems: Some participants have reported experiencing mild dizziness following the experiment. To prevent this, users are advised to avoid continuously moving their viewpoint, particularly during repetitive goal-directed training tasks, as this can lead to abrupt screen changes. To mitigate dizziness, designers can create interfaces with stable viewpoints and fixed spaces. For tasks requiring consecutive performance, the design of static, gentle, and fun game levels that allow users to pause their viewpoint or a scene is suggested. This approach provides users with the opportunity to rest and alleviate feelings of dizziness.

5.3. Conclusions

This study investigated the characteristics of different VR types and analyzed their advantages to derive design recommendations. Through a comparative analysis of freeform creativity and goal-directed training VR across various metrics, including emotional responses, usability, flow experience, technology acceptance, preferences, and satisfaction, the present study revealed significant differences. The findings underscore the potential for crafting activities that foster autonomous user engagement in activities and a beneficial overall VR experience. The conclusions of this study are as follows.

Shared advantages: Both VR interaction types performed favorably in terms of flow experience, technology acceptance, preference, and satisfaction and evoked strong positive emotions. They share a common advantage of overcoming screen limitations and transforming into 3D spatial activities.

Freeform creativity outperformed goal-directed training: Freeform creativity VR provided more pleasure and evoked stronger emotional responses than goal-directed training. It also demonstrated superior performance in terms of indicators such as flow experience, technology acceptance, preference, and satisfaction.

Usability: Goal-directed training exhibited superior usability compared with freeform creativity, particularly in terms of overall response, clarity of information presentation, and learnability. This study contends that categorizing and optimizing complex tasks or leveraging intuitive operations to enhance VR usability is crucial in designing practical VR applications.

Physical activity promotion: In this study, the goal-directed learning VR design failed to achieve activity effectiveness in physical stretching. This underscores the need for designers to incorporate more motivational elements, such as music, sound effects, and innovative virtual interactions into VR designs to evoke enjoyment, positive emotions, and immersion, thereby motivating users to continue using the system to achieve exercise goals.

Although this study provided valuable insights, it has limitations that should be considered. Future research endeavors could explore the following avenues:

Task segmentation and optimization: Both types of VR can facilitate diverse tasks. For example, freeform creativity can enable users to draw single lines using a brush or even create complex surfaces with both hands, whereas goal-directed training can engage users in a range of movements, from simple exercises to dynamic interactions like those seen in VR rhythm games featuring lightsaber interactions or boxing aerobics. Future studies could investigate the performance and efficiency differences across such diverse tasks.

User Gender and Experience: Designing different VR tasks for males and females, as well as tasks of varying complexity for participants with and without VR experience, are important considerations. Future researchers are encouraged to study the differences in score distributions based on gender and experience, investigate the underlying reasons, and propose design recommendations accordingly.

Sample Size and Diversity: This study’s sample size of 33 participants did not cover all demographics and user backgrounds. We suggest that future researchers consider increasing the sample size and including more diverse participants in subsequent studies to enhance the generalizability of the results.

User feedback and design trends: Future researchers could further investigate user feedback to optimize VR software. With ongoing advancements in VR technology, future studies can aim to improve usability and provide more learning resources and training content, thereby extending the results and contributions of this study.