These diverse examples highlight that, in VR, augmentation beyond real-world capabilities can encompass a wide variety of forms. To better grasp this diversity, Raisamo and colleagues (2019) distinguish augmentations of perception, action, and cognition. In VR, this can translate into enhancing perception by allowing users to look in multiple directions simultaneously (Fan et al., 2014), expanding action by enabling them to fly (Rosenberg et al., 2013), or supporting cognition through additional visualizations during training (Khor et al., 2016). A complementary approach (Abtahi et al., 2022) focuses on how transformations of space, body, and time create beyond-real interactions. Space transformations reshape navigation, for instance by redirecting movement to accommodate limited physical space (Rietzler et al., 2020). Body transformations alter morphology or input mappings, such as extending one’s arms or mapping breathing to interaction (Sra et al., 2018). Time transformations impact temporal perception, for example by inducing slow-motion experiences (Rietzler et al., 2017). Taken together, these perspectives illustrate both the breadth of what can be augmented and the underlying transformations.

Building on this, we define virtual augmentation provisionally as interaction designed to extend user capabilities in VR beyond real-world physical constraints. Realizing the potential of “virtual unreality” requires a systematic understanding of how virtual augmentations affect users–what performance benefits they can provide, what trade-offs they entail, and what experiential consequences they yield. Although this remains an ongoing challenge, existing research has already begun to reveal both the promise and the limitations of virtual augmentations.

Several studies show that virtual augmentations can enhance usability and efficiency. For example, the HOMER technique (Bowman and Hodges, 1997) allows users to grab distant objects with a ray cast from their hand. This reduces the necessity to move through the environment. Other virtual augmentations, such as duplicating the user’s hand to enable parallel object selection (Schjerlund et al., 2021) or providing multiple simultaneous viewpoints (Schjerlund et al., 2022), can reduce effort by requiring less physical movement to reach targets and by speeding up the overall selection process.

Yet these advantages can come with trade-offs. As the number of virtual hands increases, interaction becomes harder to control, and users must spend time choosing which hand to employ (Schjerlund et al., 2021). Likewise, although multiple viewpoints can facilitate faster navigation, an excessive number can render the interaction incomprehensible (Schjerlund et al., 2022). More complex augmentations, such as echolocation (Andreasen et al., 2019) or biofeedback-controlled abilities (Järvelä et al., 2021), even require extended training. These examples illustrate that while virtual augmentations can extend capabilities, their effectiveness ultimately depends on the conditions under which they are adopted.

Users benefit most when new capabilities are absorbed into the body schema and experienced as part of themselves (Aymerich-Franch and Ganesh, 2016). Mashiyama and colleagues (2024) demonstrated that ownership illusions can extend even to supernumerary hands and duplicate bodies, if synchronicity and focused attention are maintained. Similarly, virtual augmentations work best when their control mappings preserve familiar sensorimotor regularities (Li and Kristensson, 2025), enabling users to adapt quickly, improve task performance, and experience their augmented abilities as natural extensions of the self.

In sum, prior work has begun to map the effects of virtual augmentations, highlighting both their potential for improving performance and the challenges they pose for usability and control. As the field grows, efforts have also been made to systematize augmentation-oriented designs and establish a corresponding research agenda. Based on their literature overview, Abtahi and colleagues (2022) formulated open questions, such as how input remappings affect sensorimotor control or how individual differences shape engagement with beyond-real interactions. While these questions extend our understanding of functional design challenges, they leave another critical dimension largely unexplored: the experiential side of virtual augmentation.

Studies of the effects of virtual augmentations still focus predominantly on objective outcomes such as efficiency or precision, whereas little attention has been paid to how augmented interactions are subjectively perceived and felt. To address this gap, Neuhaus et al. (2024a) introduced the notion of augmentation experience, defined as the subjective sense of being augmented compared to the real world. Augmentation experience is conceptually distinct from other VR constructs such as immersion (system properties that support natural sensorimotor contingencies; Slater and Sanchez-Vives, 2016), presence (the psychological state of “being there”; Slater, 2009), or embodiment (ownership, agency, and self-location with a virtual body; Kilteni, Groten, et al., 2012). While these measures capture the sense of being in a virtual environment, they do not consider whether this state is experienced as mirroring real-world capabilities or transcending real-world limitations. Neuhaus et al. (2024b) devised a reliable single scale subjective measures to capture augmentation experience consisting of five items such as “…I can do things that are impossible in the real world” or “…I can control things that I cannot control in the real world”. In two studies, they demonstrated that augmentation-oriented designs reliably elicited stronger augmentation experience compared to reality-oriented designs. This provides preliminary evidence for the construct validity of the augmentation experience. In addition, augmentation experience was positively related to other experiential aspects, such as affect and need fulfillment, product evaluation, and predicted usage intention. These findings support the notion of augmentation experience as a distinct dimension of the experience of VR.

While promising, these studies leave many questions open. One concern is the conditions under which virtual augmentation is experientially most beneficial. Prior research points to pragmatic and sensorimotor aspects–such as control mappings, learnability, and usability–as crucial for determining whether a virtual augmentation can successfully support a convincing augmentation experience. Beyond these functional considerations, however, task-related factors may also play a role in shaping whether users feel augmented.

Task-compatibility is the extent to which a capability available in VR directly facilitates the achievement of the user’s current task. This notion mirrors established perspectives on fit, such as Task-Technology Fit (Goodhue and Thompson, 1995) and the ISO 9241–110 principle of “suitability for the task” (International Organization for Standardization, 2020). In positive user experience research, pragmatic qualities such as task-compatibility have been shown to contribute not only to effectiveness but also to overall attractiveness. In the AttrakDiff framework (Hassenzahl et al., 2003), overall appeal is conceptualized as the joint result of both pragmatic and hedonic qualities. In VR research, by contrast, task-compatibility has primarily been examined in terms of performance outcomes, for example in testbed evaluations that compare travel and selection techniques across task parameters or in the design of reach extensions like Go-Go (Bowman et al., 1999; Poupyrev et al., 1996). Beyond performance, however, emerging evidence specific to virtual augmentations suggests that task-compatibility also shapes the experiential value of interaction. A comparison of augmented locomotion techniques revealed that their benefits varied with task demands (Abtahi et al., 2019). An analysis of concept designs suggested that aligning augmenting interactions with a specific task provided greater benefit than offering augmentations merely to make the experience more interesting (Neuhaus et al., 2024a). In the present study, we therefore focus on task-compatibility as a potential moderator of augmentation experience.

To examine whether virtual augmentations increase augmentation experience in general or whether feeling augmented depends on task-compatibility, we conducted an experimental vignette study (Aguinis and Bradley, 2014). Participants watched first-person videos of a VR user completing a course of four rooms, each requiring small obstacles to be overcome. We compared three interaction conditions: an augmentation-oriented, task-compatible design, where the user could grow or shrink to overcome obstacles specifically designed for this ability (e.g., fitting through a tiny passage or reaching a high switch), an augmentation-oriented, non-task-compatible design, where the user could use ray-casting to grab distant objects–a capability available but not useful for solving the course’s obstacles, and a reality-oriented design, where objects were manipulated by grabbing and moving them as in the physical world (see Figure 1). This design allows for a direct comparison between reality-oriented interaction, augmentation in general, and augmentation aligned versus misaligned with task demands.

Based on prior work, we formulated two hypotheses. First (H1), we expected augmentation experience to be most intense in the task-compatible augmentation condition, compared to both the non-task-compatible augmentation and the reality-oriented condition. Second (H2), we expected augmentation experience to be positively associated with key experiential outcomes (positive affect, need fulfillment) as well as product-related evaluations (hedonic quality, attractiveness).

For our study, we developed a dedicated VR application in Unity3D. Users must complete a course by completing several rooms. To complete a room, simple physical obstacles must be overcome, for which the users need their virtual abilities. At the end of each room, a switch must be pressed. The room is then complete and can be left. Completion is indicated by a light that turns green over the entrance of the room. The goal is to reach the final room, which only unlocks once all four rooms are completed.

We chose this simple, rather neutral VR environment to control for effects of scenario-specific content. Everything is kept in a light-colored lab aesthetic to not distract from the interaction. Interactable objects were kept light red, non-interactable objects white and walls light blue. Non-interactable everyday objects are included to provide users with a sense of scale. Figure 2 shows a first-person view when entering the course; Figure 3 shows a schematic map of the course. Before the task, the user is placed in a smaller room (not pictured in Figure 2 or Figure 3) to familiarize with the controls and interactions.

We provided users with different abilities and different obstacles to manipulate the dependent variables of interest. Apart from this, we matched the number of rooms, action steps, and video duration and followed a fixed script to keep the video as similar as possible.

Table 1 shows an overview of the differences per condition. In the augmentation-oriented, task-compatible condition, users can grow or shrink themselves at will. They can use this augmenting interaction technique to overcome particular obstacles and complete the course (e.g., get through a tiny door, reach a switch that is high on the wall). In the augmentation-oriented, non-task-compatible condition, users can grab objects from a distance through ray-casting. However, while this ability can be used on decorative objects in the virtual environment, it does not help to overcome obstacles and to solve the course. To solve the course, users can manipulate objects by grabbing and moving, for example, to overcome obstacles like doors or a box that can be pushed to the side. Finally, in the reality-oriented condition, users can manipulate objects by grabbing and moving them, just like in the non-task-compatible, augmentation-oriented condition.

We chose the two augmenting interactions because they are both simple to understand and easy for experienced VR users to imagine using. This is especially important as participants did not interact first-hand but instead watched first-person videos. Further, the interactions represent good examples of virtual augmentations because they leverage capabilities unique to VR, such as manipulating scale or interacting with distant objects, which are unattainable in the real world.

We recorded first-person videos of a successful run through each condition. We followed a fixed, predetermined script to make the videos as similar and therefore as comparable as possible. For instance, the order of how the rooms were solved and the overall length are almost identical (each video lasts just over 6 minutes). At the beginning of each video, a short text asks the viewer to imagine that they are the user from whose perspective the video was recorded and points out the capabilities available in the condition (e.g., “In this VR application, you can walk around and use your hands to grow and shrink yourself./First, you are in a room to familiarize with the controls and your abilities.”). The user is then in a test room to familiarize with the abilities and all interactive objects of the upcoming course. After going through all abilities, the user is teleported to the start of the course. The video ends when the user reaches the goal room, and a success message is displayed. To make the videos easier to understand, we included short text descriptions and instructions before sections of the videos. The entire vignette videos are included in the Supplementary Material.

The study was conducted online using SurveyMonkey ¹ . Participants were recruited via Prolific ² and received about €4.70 for their participation (equivalent to over €14 per hour, i.e., above local minimum wage). The study typically took between 12 and 15 min to complete. In total, 122 participants completed the study, although two participants were excluded due to failed attention checks and unrealistically short completion times. The final sample consisted of 120 participants (52 female, 66 male, 1 diverse, 1 not specified; age range: 18–63, M = 30.1, Median = 29). As the study was conducted in English, proficiency in English was a prerequisite for participation.

Participants were randomly assigned to one of three conditions: “augmented, task-compatible”, “augmented, non-task-compatible”, or “reality-oriented” (control). We employed a between-subject design, since a within-subject design implies direct comparison and is prone to various contrast and anchor effects. In this sense, a between-subject design is the more conservative approach. It keeps respondents naïve to the underlying structure of the experiment and related hypothesis (Aguinis and Bradley, 2014; Sniderman and Douglas, 1996). We obtained complete data from 40 participants in each condition. As it was vital to vividly imagine being the user in the video of the VR application, participants were selected in advance to be experienced VR users. Using Prolific’s screening options, we made sure that every participant owned a VR system and used VR frequently. This should further increase the imaginability of the interaction based on the video vignettes and reduce unwanted variance from unfamiliarity with basic VR interactions and principles.

To contextualize participants’ responses and assess potential influences on their evaluations, we measured their frequency of VR use, attitudes toward VR, and previous experiences with VR. Thirty-eight participants (32%) used VR once or several times a week, 59 participants (49%) used VR once or several times a month, and 23 participants (19%) used VR less often than once a month. Further, we assessed participants’ attitudes towards VR with four questions. On an intensity scale from 1 (“not at all”) to 7 (“extremely”) we asked how much participants agree that VR is “…innovative” (M = 5.93, SD = 1.17), “…exciting” (M = 6.02, SD = 1.04), “…superfluous” (M = 3.89, SD = 1.92), and “…a waste of time” (M = 2.02, SD = 1.20). In addition, we asked the participants to rate their previous experiences with VR on a seven-point scale (“very negative” to “very positive”; additional option: “can’t say”). Forty-two participants (35%) rated their experiences as very positive, 46 (38%) as positive, 22 (18%) as rather positive, 6 (5%) as neutral, and 4 (3%) as negative. Overall, the sample reported a positive attitude towards VR and generally positive experiences. Several one-way analyses of variance (ANOVA) revealed no significant differences between the conditions in terms attitude and previous experience.

After a brief introduction, participants were asked to watch the 6-min video in full-screen mode at normal speed from start to finish (for more detail on the video, see Section 3.1). Participants were asked to imagine that they use the application themselves. We included two instructional manipulation checks (Oppenheimer et al., 2009) to ensure quality responses. One check was inserted immediately after the video (“How many rooms did the user solve to complete the course in the video?” with a choice of 1, 2, 3, or 4) and one towards the end of the survey (“The color of the walls in the video shown earlier is light blue. Please select ‘light blue’ in the following question. This is an attention check.” with a choice of ‘light green’, ‘light blue’, ‘dark brown’, and ‘dark red’). For both attention checks we followed guidelines for performing fair attention checks provided by the platform Prolific ³ and made sure that no memory recall was necessary to successfully complete the checks (for the first check, participants were able to watch the video again if they did not remember how many rooms were completed). One participant failed both attention checks, and another completed the survey faster than the duration of the video. Both were excluded from the analysis and replaced with new participants, leading to a final sample of N = 120 in the analysis. After the video, participants were asked to answer questions based on their expected experience.

To test our hypotheses, we measured the central construct of augmentation experience, several experiential outcomes, and imaginability as a control variable.

Because no a priori power analysis was conducted, a post hoc sensitivity analysis was performed using G*Power 3.1 (Faul et al., 2009). For a one-way ANCOVA with three groups (N = 120, α = 0.05, power = 0.80) and two covariates, the smallest detectable effect was f = 0.29 (η² = 0.08), which corresponds to a medium effect size (Cohen, 1988). This also holds true for an ANOVA with three groups. The study was therefore sufficiently sensitive to detect effects of this magnitude or larger with high probability.

A sensitivity analysis for bivariate correlations showed that, under the same parameters, the smallest detectable correlation was r = 0.23, also representing a medium effect (Cohen, 1988). This indicates that the study was adequately powered to detect correlations of medium size or larger.

We calculated an ANCOVA with interaction (augmented task-compatible, augmented non-task-compatible, reality-oriented) as between-subject factor, imaginability and novelty as covariates (controlling for their effects) and augmentation experience (AE) as measure to test H1.

The ANCOVA revealed a significant main effect of interaction on AE, F (2, 115) = 15.36, p <0 .001, partial η 2 = 0.21, ω 2 = 0.17. The augmented task-compatible interaction led to a more intense AE (M = 5.09, SE = 0.20, 95% CI [4.69, 5.49]) compared to the augmented non-task-compatible (M = 4.03, SE = 0.20, 95% CI [3.63, 4.42]) and the reality-oriented interaction (M = 3.52, SE = 0.20, 95% CI [3.12, 3.92]).

The effect was the most pronounced when comparing the augmented task-compatible condition to the reality-oriented condition (Mean diff = 1.57, p_tukey <0.001, B = −1.57, SE = 0.29, t (115) = -5.43). We found no significant difference between the augmented non-task-compatible condition and the reality-oriented condition (Mean diff = −0.51, p_tukey = 0.18). In fact, the augmented task-compatible condition also differed significantly from the augmented non-task-compatible condition (Mean diff = 1.07, p_tukey <0.001, B = −1.07, SE = 0.28, t (115) = -3.74). In other words, the effect of augmentation-oriented design on AE is primarily due to the task-compatible design. Effect sizes indicated that interaction as between-subject factor accounted for ∼17–21% of AE variance after covariate adjustment (partial η² = 0.21, ω² = 0.17; large by conventional benchmarks) (Cohen, 1988, 284–88), with novelty contributing ∼14% (partial η² = 0.14; medium–large) and Imaginability ∼3–4% (partial η² = 0.04; small). The task-compatible vs. reality-oriented contrast was 1.57 points on a 1–7 scale (≈26% of the scale range), underscoring practical relevance.

Imaginability and novelty had a significant effect on AE (Imaginability: F (1, 115) = 4.19, p = 0.04, B = 0.18, SE = 0.09, partial η 2 = 0.04, ω 2 = 0.02; novelty, F (1, 115) = 18.55, p <0 .001, B = 0.30, SE = 0.07, partial η 2 = 0.14, ω 2 = 0.10). Both, imaginability (r = 0.31, p <0 .001, N = 120) and novelty (r = 0.46, p <0 .001, N = 120) were positively related to AE. The better participants could put themselves into the situation and the more stimulating they found the scenario and interaction itself, the higher augmentation experience. However, the ANCOVA shows that even after controlling for these effects, the difference between conditions remains highly significant.

To assess whether the results hold without covariates, we ran an ANOVA not controlling for imaginability and novelty, which led to similar findings: F (2, 117) = 19.9, p <0 .001, ω 2 = 0.24, with the augmented task-compatible condition leading to higher augmentation experience (M = 5.28, SE = 0.22) compared to the reality-oriented condition (M = 3.35, SE = 0.22) and the augmented non-task-compatible (M = 4.01, SE = 0.22).

As expected (H1), the virtual augmentation led to more augmentation experience, but only if the augmentation is task-compatible, i.e., related to the given task. We controlled for imaginability to ensure that the results are not biased by general differences in quality or understandability of the vignette videos. We further controlled for novelty to ensure that the results are not due to systematic differences in the perceived novelty of the interactions. All in all, to feel augmented by virtual abilities that go beyond what is possible in the real world, it is not only beneficial but seems necessary that these new abilities are compatible with the given task. Otherwise, the augmentation does not create an augmentation experience.

Table 2 shows the bivariate correlations between augmentation experience (AE) and the experience-oriented aspects positive (PA) and negative affect (NA), need fulfillment (NEED), the product-oriented aspects hedonic quality (HQ) and general attractiveness (ATT).

H2 assumed a positive correlation of AE with key experiential aspects and product attributes. As expected, a more intense augmentation experience was accompanied by more positive affect (PA, r = 0.55) as well as higher need fulfillment (NEED, r = 0.52), more perceived hedonic quality (HQ, r = 0.45) and general attractiveness (ATT, r = 0.43). Conversely, we find no correlation at all for negative affect (r = −0.02). Consistent with prior research, the experiential aspect need fulfillment highly overlaps with positive affect (Hassenzahl et al., 2010; Sheldon et al., 2001). On the product level, hedonic quality is strongly related to general attractiveness (e.g., Hassenzahl et al., 2000).

As expected (H2), augmentation experience is positively related to crucial experiential aspects and product attributes. Participants, who had a more intense augmentation experience, felt better and liked the VR application more.

The study provides the first empirical evidence that the effectiveness of virtual augmentations in eliciting augmentation experience depends on their task-compatibility. Earlier research hinted at this connection (Abtahi et al., 2019; Neuhaus et al., 2024a) but did not explicitly test whether alignment with the task influences the subjective sense of being augmented. By operationalizing task-compatibility as a direct functional fit between an augmenting capability and task requirements, the present study demonstrates this relationship experimentally.

This finding extends traditional notions of fit–such as Task-Technology Fit (Goodhue and Thompson, 1995) or the ISO 9241–110 principle of “suitability for the task” (International Organization for Standardization, 2020) – which typically link compatibility to performance, usability, or acceptance. In the context of VR and virtual augmentations, our results suggest that compatibility also shapes how augmented interactions are experienced: when an augmenting capability transparently supports a task, users perceive it as an extension of their own abilities rather than an arbitrary or decorative modification. While prior HCI research has conceptually connected meaningful interaction to task relevance and pragmatic quality (e.g., Hassenzahl, 2004), the present study provides direct empirical evidence for this relationship in the specific case of virtual augmentations.

In this way, our study adds an experiential dimension to prior research on virtual augmentation. While earlier work focused on the functional and sensorimotor conditions under which augmentations operate effectively–such as remapping, control plausibility, and learnability (e.g., Abtahi et al., 2022; Li and Kristensson, 2025) – we show that alignment with the task is critical for augmentations to feel experientially meaningful. Future research in interactive VR should integrate experiential measures alongside performance-based measures to further clarify how functional fit, performance and the experience of being augmented jointly shape user acceptance.

While the study supports our interpretation, alternative explanations should be considered. One possibility is that the ray-casting interaction in the non-task-compatible condition was more familiar to participants, given their prior VR experience. From this perspective, the stronger augmentation experience in the task-compatible condition could reflect novelty rather than compatibility. However, this explanation seems unlikely: when controlling for perceived novelty (“creative,” “captivating”), the effect of task-compatibility on AE remained robust. The results therefore support the interpretation that functional alignment, rather than mere novelty, drives the stronger sense of augmentation.

Beyond establishing the role of task-compatibility, our results replicate previous findings that augmentation experience is positively associated with positive affect, need fulfillment, and hedonic quality (Neuhaus et al., 2024b). This replication is noteworthy because compared to earlier textual vignette studies, the present experiment employed first-person video vignettes depicting fully realized interactions. This reduced the need for imagination and provided more concrete visual cues, thereby increasing ecological validity while maintaining the methodological control typical for vignette-based studies. The consistency of results across these different modes of presentation supports the robustness of augmentation experience as a viable construct and its relevance for understanding how virtual augmentations contribute to positive user experience.

Finally, the low internal consistency of the pragmatic quality scale should be interpreted with caution. Given the non-interactive setup, participants observed rather than enacted the interactions, which limits reliable judgments of usability or efficiency. In contrast, hedonic impressions such as stimulation and attractiveness form rapidly and are less dependent on direct engagement (Lindgaard et al., 2006; Karapanos et al., 2009). Accordingly, the strong and consistent relationships between augmentation experience and hedonic outcomes in our study likely reflect stable, experiential appraisals, whereas pragmatic aspects remain an open question for future interactive replications.

Moving from theoretical interpretation to practical design implications, our findings offer guidance for how virtual augmentations should be applied in VR. In the present study, we tested virtual augmentations in a neutral environment to isolate the effect of task-compatibility without interference from domain-specific context or content. Such controlled testbeds are standard for understanding fundamental design principles before applying them to particular use cases (Bowman et al., 1999). At the same time, applying these insights in practice will require domain-specific adaptation.

In design practice, virtual augmentations should not be added arbitrarily. Instead, designers should begin with a clear analysis of the user’s tasks, then select or shape augmentations whose affordances directly support those tasks. For instance, when a user must reach, manipulate, or navigate, augmentations should extend these exact actions in meaningful ways. When augmentation capabilities serve immediate user tasks, they stand to contribute not just to efficiency, but to a more meaningful subjective experience.

Applying this principle will look different across domains. In training and education, augmentations that slow down or magnify fast or complex processes can make them easier to grasp. In industrial or professional contexts, extending reach or perspective can reduce effort or improve spatial awareness without compromising control. In rehabilitation, body-scaling or extended manipulation may help users explore motion ranges safely. In entertainment and creative applications, augmentations can intentionally deviate from realism to enrich engagement or provide novel forms of agency–yet should still align with the logic of the task to remain coherent and satisfying. Designers should take task-compatibility into account when developing augmentation concepts, for instance by using structured ideation tools such as the design cards for virtual augmentations (Neuhaus et al., 2024a). While these cards were developed to support exploration of “unreal” augmentations more broadly, they could be extended to explicitly address task-compatibility as a design dimension.

In summary, this study advances the understanding of virtual augmentations by identifying task-compatibility as a key factor that shapes whether users feel augmented. Rather than treating augmentations solely as tools for improving performance, our findings position them as experiential design elements whose value depends on their alignment with user tasks. These findings form the basis for future research investigating task-compatible augmentation in interactive and domain-specific contexts.

While our findings showed that task-compatible virtual augmentations increase augmentation experience, which is associated with positive experiential outcomes, several methodological aspects limit their generalizability. Using video-based experimental vignette methodology allowed for strong experimental control but necessarily captures ratings based on imagined rather than enacted interaction. That means presence, embodiment, and sensorimotor contingencies cannot be reproduced, and in turn, effect sizes may differ from interactive VR. Therefore, we consider our results a first step, which should be complemented by replications in fully interactive VR, where embodied interaction may amplify or further qualify the observed effects.

In addition, the vignette videos depicted a flawless, fixed-time walkthrough. While this design minimized noise from performance differences (Hyman and Steiner, 1996; Matza et al., 2021), it also precluded the collection of objective measures such as completion time or error rate. Future interactive studies should therefore integrate behavioral and physiological indicators (e.g., kinematics, eye movement, effort) alongside subjective responses to enable explicit cost-benefit analyses of virtual augmentations.

Our implementation of task-compatibility required adapting the environment to the available capabilities, which may have partly linked compatibility to the perceived usefulness of the interaction. Future research could employ a factorial design that systematically varies both augmentation type (e.g., scaling vs. ray-casting) and task demands, holding difficulty constant and allowing for intermediate levels of compatibility rather than contrasting extremes.

We used a between-subject design to keep participants unaware of other conditions and avoid contrast effects that could bias subjective judgments. While a within-subject design could increase statistical power, it would risk demand and carryover effects once participants recognize the experimental conditions. Future work might test mixed or counterbalanced designs, possibly with rest periods or filler tasks, to balance these trade-offs.

We recruited experienced VR users to ensure that participants could vividly imagine the interactions shown in the vignettes. Although prior work found little influence of experience level on augmentation experience (Neuhaus et al., 2024b), generalization to novices remains to be tested. Replications could include participants with varying VR familiarity, short training sessions, or familiarity covariates to better separate task-fit effects from novelty effects.

Furthermore, while the augmentation experience scale has shown strong internal consistency and sensitivity in multiple studies, its construct validity must be further established. Future work should include other measures (e.g., presence, agency, embodiment) to map how augmentation experience overlaps or differs from these measures.

Finally, this study focused on augmentations of action (i.e., extending what users can physically do in VR). Future research should examine perceptual and cognitive augmentations, which may rely on distinct experiential mechanisms. Extending this framework to domain-specific contexts–for instance, training and education, rehabilitation, or industrial teleoperation–can test how task-compatibility and augmentation experience function under real-world constraints. Such studies should combine controlled testbed experiments with field and longitudinal designs to observe adaptation, habituation, and long-term experiential change.