Immersive Journalism and the Shift to the Metaverse: A Bibliometric and NLP Analysis with Projections to 2030

Abstract

Immersive journalism—a practice that situates audiences within news events through VR, AR, and 360-degree video—has developed steadily since the mid-2000s. Yet how metaverse discourse has reshaped the field’s research agenda remains unclear, as existing bibliometric studies remain descriptive, tracking publication counts without examining whether the metaverse label reflects a conceptual shift or merely a trend. This study addresses that gap using bibliometric and NLP analysis of 143 Scopus-indexed publications (2005–2026). First, the Metaverse Proximity Score, introduced here, shows metaverse-proximate terminology increased nearly fivefold in mean score between 2022 and 2025 (0.48 to 2.29). Second, although empathy-focused studies attract the highest mean citation count (37.4), their share peaked in 2023 (42.1%) before falling sharply to 21.4% by 2025; this “empathy paradox” shows that citation impact and thematic visibility diverge. Third, discussions of trust and disinformation rose markedly over the sample period, reaching their highest share by 2025, though this estimate rests on a small subsample (n = 14) and should be read as directional rather than precise. Projections to 2030 suggest a plausible scenario in which publication volume stabilizes while metaverse-related themes continue to gain influence, suggesting the field may be moving beyond its technology-driven origins toward an agenda shaped by ethical and normative concerns.

1. Introduction

Immersive journalism emerged as a distinct area of journalistic experimentation in the mid-2000s and was conceptually consolidated through research on virtual reality, 360-degree video, and interactive storytelling during the 2010s. Since the early 2020s, however, the field has increasingly been discussed alongside the metaverse and related digital environments, raising questions about whether this shift represents a substantive transformation in the research agenda or primarily a change in terminology. The concept was formally defined for the first time in an article published in 2010 by Nonny de la Peña and colleagues (De la Peña et al., 2010); de la Peña subsequently established herself as one of the academic pioneers of the field through her landmark contributions to immersive journalism experience, most notably “Hunger in Los Angeles” and “Project Syria.” Employing technologies such as VR, 360° video, and AR, immersive journalism denotes a journalistic tendency premised on enabling the first-hand experience of reality, fostering audience participation, and enhancing users’ empathic capacity. This study traces immersive journalism from the newspaper to the metaverse, seeking to conceptualize the “empathy paradox” as a structural tension—namely, the sharp decline of the empathy theme’s representational presence within the research agenda despite its demonstrably high academic impact. The study further aims, through the provision of data-driven near-term projection forecasts covering the 2026–2030 period, to transcend the function of retrospective mapping and to constitute an instance of more thematically oriented analysis that assumes the character of a forward-looking research agenda proposal, as well as to provide an analytical framework for discussing recent developments in immersive journalism research. In pursuit of these objectives, the study seeks to address four core research questions.

The first concerns the structural phases through which the immersive journalism literature has passed during the 2005–2026 period and inquires how the conceptual transformations characteristic of each phase may be defined. The second interrogates why the empathy theme—which has generated the highest citation impact in the literature—has sharply lost its representational standing within the research agenda, and aims to account for this tension, conceptualized in the present study as the “empathy paradox.” The third question is whether the concept of the metaverse functions in the immersive journalism literature as a superficial terminological adoption or as a framework that structurally reconfigures the research agenda. The fourth is directed toward elucidating the relationship between the rise in ethical concerns observed toward the end of the sample period and the metaverse-oriented conceptual transformation of the field.

The study is expected to make three original contributions to the literature. First, the “empathy paradox”, used here as an analytical label for the observed divergence between empathy’s high citation impact and its declining thematic visibility, is introduced for the first time in this study and substantiated with quantitative evidence. Second, the Metaverse Proximity Score, which assigns a numerical value to the metaverse-oriented conceptual disposition of each article, is introduced for the first time in the literature by this study and documents the paradigm shift from VR to the metaverse through a measurable and replicable indicator. Rather than serving as a predictive model, the Metaverse Proximity Score was designed as an exploratory indicator to capture the extent to which metaverse-related concepts have become embedded in immersive journalism research over time. Third, in a field where the overwhelming majority of existing bibliometric studies have fulfilled a retrospective mapping function, this study advances a forward-looking research agenda proposal by means of data-driven projection analyses pertaining to the 2026–2030 period. The study is grounded in a hybrid methodological approach that integrates bibliometric techniques with text-based content analysis; the dataset was retrieved on 1 March 2026 through a search of the Scopus database encompassing the title, abstract, and keyword fields. Three distinct research phases are identified in the study—2005–2017, 2018–2021, and post-2022—demarcated in accordance with observed shifts in publication density. The transformation of journalistic values is examined through a rule-based text-matching procedure drawing on pre-defined conceptual lexicons. A survey of the relevant literature reveals that, across the historical trajectory of immersive journalism, the foundational assumption has consistently centered on the capacity of new communication technologies to afford readers a first-hand experience of unfolding events. This trajectory extends from among the earliest empirical instances—Wilson and Tan’s (2005) study, which focused on the role of cultural differences in the online newspaper reading experiences of Chinese and Malaysian users and proposed “less tangible reading methods” in journalism-to Bonnie Brennen and Erika Dela Cerna’s article “Journalism in Second Life” (Brennen & Dela Cerna, 2010), which assessed three newspapers operating within Second Life—a computer-generated alternative reality platform—namely the Alphaville Herald, Metaverse Messenger, and Second Life Newspaper, from the perspective of next-generation journalistic practices; to Nonny de la Peña and colleagues’ foundational article “Immersive Journalism: Immersive Virtual Reality for the First-Person Experience of News” (De la Peña et al., 2010); to Eva Dominguez-Martín’s pioneering contribution to immersive journalism, “Immersive Journalism or the Influences of Virtual Reality and Video Games on the Interface and Interactivity of Current Affairs News” (Domínguez-Martín, 2015), which examined the effects of video games on non-fictional digital narratives and journalism while exploring the possibilities of sensory and interactive experience in reporting; and to Don Belt and Jeff South’s study “Slow Journalism and the ‘Out of Eden’ Walk” (Belt & South, 2016), which evaluated Pulitzer Prize-winning journalist Paul Salopek’s seven-year “Out of Eden Walk” journey through the lens of slow journalism. Moreover, existing bibliometric studies on immersive journalism have generally sought to characterize the structural properties of the field through descriptive indicators such as publication counts, journal distributions, and author productivity metrics. The present study departs from such research approaches at both methodological and conceptual levels. Rather than confining itself to standard bibliometric techniques, the study applies a mixed-methods framework that integrates Python-based NLP analysis, co-occurrence network modeling, and rule-based text matching, with the aim of measuring not only the extent of the field’s growth but also the nature of its conceptual transformation. In this regard, the study is of particular significance insofar as it develops a set of original indicators tested for the first time in the literature. Furthermore, the study possesses distinctive value both in its quantitative documentation of the paradigm shift from VR to the metaverse—achieved by assigning a numerical value to the conceptual orientation of each article through the Metaverse Proximity Score—and in its data-driven anticipation of the discipline’s prospective trajectories through near-term projection analyses.

2. Literature Review

2.1. Immersive Journalism as an Emerging Field of Research

First defined in 2010 by Nonny de la Peña and colleagues as “news production that enables audiences to experience first-hand the events or situations narrated in the news” (De la Peña et al., 2010), immersive journalism represents a development that far exceeds the status of a mere technological innovation, notwithstanding its reliance on the affordances offered by new communication technologies. Through the VR news experiences it provides, immersive journalism constitutes a reporting practice that resists audience passivity and disengagement, privileging active participation and experiential engagement over the modalities of conventional presentation (Jones, 2017; Kang et al., 2019; Alpsoy, 2026). By employing virtual reality technologies—including VR, 360° video, and AR-to afford readers a deeply immersive encounter with the events, circumstances, and spaces that constitute news content, and by enabling stories to be experienced from a first-person perspective, immersive journalism has introduced a set of novel concepts into the existing literature, among them presence, empathy, and narrative immersion (Sundar et al., 2017; Shin & Biocca, 2018). Some scholars have argued that immersive journalism, which draws upon augmented reality capabilities and commands a considerably more extensive digital technological infrastructure than earlier forms such as data journalism, mobile journalism, or online journalism, is a product of an increasingly competitive media market and represents an exciting development for audiences whose sense of reality has been dulled by the accelerated news cycle (Cheng & Verboord, 2025). According to certain analysts, interactivity—one of the defining characteristics of immersive journalism—has not merely added an objective dimension to reality but has also given rise to a personal narrative shaped by immediate affect and the sense of presence. From this perspective, bodily participation is itself a precondition of experience, insofar as “the body is where experience takes place; immersive journalism productions are therefore triggers or potential carriers of that experience” (Fonseca et al., 2026). Among the other salient features of immersive journalism are its capacity to generate in the reader an impression of transcending spatial and temporal distance, which produces an “illusion of virtual body ownership,” and its orientation toward the use of virtual reality as a medium for conveying real-world stories (Hendriks Vettehen et al., 2019; Barreda-Ángeles et al., 2021; Cheng & Verboord, 2025). It should further be noted that one of the significant factors driving the development of immersive journalism is the view that the gradual erosion of the traditional news audience, in the face of steadily rising content production costs, positions this emerging journalistic practice as a potential source of alternative solutions to existing industry challenges (Hendriks Vettehen et al., 2019). This is important in terms of explaining why major media organizations, including The New York Times, The Washington Post, The Guardian, USA Today, and the BBC, alongside a range of independent content producers, have increasingly compelled companies to invest heavily in 360-degree news video production. By adding a new platform to its news production and distribution processes in November 2015, distributing more than one million cardboard virtual reality (VR) headsets to its subscribers and launching an application that presents news content through 360-degree filming, the New York Times, in particular, moved decisively in this direction (Jones, 2017). A significant goal underlying the sense of reality intended to be produced by this technical apparatus in the user is the conviction that it may contribute to an effective empathic response to dramatic events and human suffering occurring at a distance (Van Damme et al., 2019; Sánchez Laws, 2020). The nature of such empathic engagement is, of course, closely connected to the experiential reality created by metaverse technology.

2.2. The Metaverse: The New Horizon of Immersive Media

Baía Reis and Coelho (2018) seek to reveal the conceptual framework of immersive journalism and analyze its applications and limitations, and they designate this emerging journalistic form as “being-there journalism”, which is a label that explicitly invokes its connection to virtual spatial experience and the metaverse. According to the authors, this reporting practice—characterized by processes of self-perception through virtual reality, the immersion of presence, visual representation, and participatory engagement—bears witness to the emergence of a new generation of journalism. The concept of the metaverse, first introduced by Neal Stephenson in his novel Snow Crash, was employed to denote “a persistent virtual world that reaches into, interacts with, and affects the lives of its participants” (Stephenson, 1992; Lucia et al., 2025). At one point in the novel, Stephenson writes: “On the back is gibberish explaining how to reach him: a phone number. Encrypted radio waves. And an address in the Metaverse” (Stephenson, 1992, p. 11). Elsewhere, he offers the following description: “He is in a computer-generated universe that his computer is drawing onto his goggles and pumping into his earphones. This imaginary place is known as the Metaverse” (Stephenson, 1992, p. 14). Matthew Ball emerges as the second major figure to engage with the concept in depth, doing so in his 2022 book The Metaverse: And How It Will Revolutionize Everything, in which he describes the metaverse as “a massively scaled and interoperable network of real-time rendered 3D virtual worlds” (Ball, 2022). Ball and Heath similarly address the diverse applications of the metaverse across “lifestyle-oriented businesses such as education, fitness, dating, and social networking,” noting that, given the cultivation of a sufficiently critical target audience, the technology could equally be applied to sectors such as gaming (Ball, 2022; Heath, 2023). While some scholars conceive of the metaverse as an evolved form of natural reality, others contend that it represents a transition from the current two-dimensional internet to a three-dimensional reality universe, and in this sense constitutes a digital replica of the real world in which activities such as work, education, entertainment, gaming, and communication are experienced (Koçak, 2023; Zeren & Güler, 2022; De la Casa & Bautista, 2023). When metaverse debates arise in the context of journalism and media more broadly, a number of concepts that are of critical importance to the experience of reality come immediately to the fore—most notably, virtual reality, spatial experience, and the avatar. In its most elemental formulation, “virtual reality describes the experience of people being present in an alternative environment” (Doyle et al., 2016). In the production of a digital replica of the physical world, both “augmented reality” and “360-degree video” technology—which enables viewers to capture an entire scene from multiple angles—may be said to play a significant role. According to Toursel and Useille (2019), however, the technology employed in virtual reality (VR) journalism is not, in fact, particularly new; as they note, “the first experiments in this domain were conducted in the 1960s by American computer scientist Ivan Sutherland.” Media and journalism studies, which had largely neglected the question of the interaction between journalism and the metaverse in earlier periods, began from 2020 onward—following the recognition of the remarkable potential of VR and AR (Augmented Reality) technologies for the production of immersive news, and particularly through research that combined the logics of news and gaming—to produce work focused on how users experience such instances of immersive journalism (Olausson & Engström, 2025). The relationship between the conjunction of immersive journalism with the concepts of virtual worlds, avatars, and spatial experience on the one hand, and metaverse technology on the other, is of considerable significance—particularly with respect to the metaverse’s capacity to elicit powerful emotional responses in users, such as heightened awareness and the motivation to act (Eskiadi & Panagiotou, 2026). The interaction-based content structure of immersive journalism and its narrative mode, which conveys to the user a sense of “being there,” inspire not only a deeply layered experience but also a heightened awareness of the real physical environment surrounding the user. Such awareness or empathy is facilitated through what may be termed immersive elements—including “Response-As-If-Real” (RAIR); hardware devices such as desktop computers, mobile phones, and head-mounted displays (HMDs); transmediality strategies employed in creative content production and particularly in interactive documentary storytelling; and systems such as the “Cave Automatic Virtual Environment” (CAVE) (Falandes & Renó, 2023; Zhou, 2024). The strategies employed to draw the user into the event through 360-degree video technology may be exemplified by the first episode of the BBC production “Damning the Nile in 360,” in which viewers find themselves seated around the same table in a restaurant alongside the presenter and other participants. The presenter abruptly turns to the audience and declares: “So you’re joining us for a traditional dinner in Addis Ababa” (Gallego Abellán et al., 2024). Certain researchers focused on experimental work in non-fictional VR storytelling—particularly in the areas of micro-narrative, abstraction, and interactive navigation—address the partial relationship between metaverse technology and immersive journalism through virtual reality-based documentary productions such as “In Pieces VR,” which is designed to explore contemporary political crises and the personal experiences of individuals imprisoned for political reasons (Soler-Adillon, 2022). It is evident that such experiences meet the audience’s need for both empathic engagement and the pursuit of emotional arousal. In this context, through 360° video news, users and audiences not only acquire a richer and more nuanced perception of events but also discover an effective means of entering into and seeking to comprehend the stories of others through the pursuit of excitement or experiential engagement (Yunju & Heejun, 2022). In brief, these and analogous technologies—which afford users spatial experience through virtual reality—may be said to constitute concrete instances of the participatory forms generated through immersive media. Indeed, research conducted by analysts such as Hannah Greber into the ways in which interactivity shapes participation in immersive journalism has demonstrated that inclusive technologies exert significant experience-shaping effects by adding new layers of meaning to reality (Greber, 2024).

Most studies mapping the immersive journalism literature have taken “immersive journalism” as their core term and used VOSviewer or similar software (R/RStudio/biblioshiny) (Arık et al., 2024; Alpsoy, 2026; Welskop, 2025). Sanchez-Acedo et al. (2023), by contrast, employ a qualitative coding framework (DESLOCIS). Arık et al. (2024), Alpsoy (2026), and Welskop (2025) identify Spain as one of the most productive centers in the field, whereas Sanchez-Acedo et al. (2023) do not report this type of country- or author-level productivity finding. Arık et al. and Alpsoy also identify Sara Pérez-Seijo by name as one of the most productive authors. The databases and corpus sizes used in these studies also differ. Arık et al. and Alpsoy use Web of Science without filters and obtain large corpora of n = 955 and n = 200, respectively. Sanchez-Acedo et al. narrow their search to three databases (WoS, Scopus, and Google Scholar) using PRISMA filters to n = 61, whereas Welskop restricts Scopus by article type, English language, and publication year and analyzes n = 135 research units. The time periods covered by the studies also vary. Arık et al. (1999–2023) and Alpsoy (1997–2025) use the broadest time windows, Welskop (2010–2024) uses a medium-length period, and Sanchez-Acedo et al. select the narrowest window (2017–2022). The scope of the search terms also differs. Arık et al. and Alpsoy use a narrow set of terms, whereas Welskop combines eight term variants, and Sanchez-Acedo et al. use seven separate Boolean operator sets in English and Spanish. In terms of methodological depth, Arık et al., Alpsoy, and Welskop conduct only quantitative bibliometric analyses (publication and citation counts, collaboration networks, and keyword maps), whereas Sanchez-Acedo et al. additionally code the content of each article using the DESLOCIS framework. Welskop and Alpsoy report similar findings: both identify “immersive journalism,” “virtual reality,” and “experience”/”interactivity” among the most frequent keywords and identify Spain and the United States as leading countries. Their thematic maps, however, differ. In Alpsoy’s map, “immersive journalism” and “virtual reality” appear as basic themes, while “journalism” and “news” are classified as motor themes. In Welskop’s analysis, “metaverse” appears in a separate cluster and is identified as a concept associated with the most recent period (2022–2023). The findings of Arık et al. have a different character. Their most highly cited articles were authored by researchers in virtual reality and neuroscience, and most of the ten most productive authors were affiliated with computer engineering or telecommunications; Pérez-Seijo was the only scholar from communication studies on the list. Each of the four studies positions its original contribution differently. Arık et al. (2024) state that they provide a “general and comprehensive intellectual map.” Sanchez-Acedo et al. (2023) show that the field is “still in an initial exploratory stage” and that the metaverse and immersive technologies are studied from “isolated and non-convergent perspectives.” Welskop (2025) identifies the complete absence of MR/XR concepts from the map, the disappearance of ethical issues after 2019, and the absence of artificial intelligence as research gaps. Alpsoy (2026) states that the study provides a comprehensive roadmap spanning the period 1997–2025, while also emphasizing that the field remains “open to further theoretical development.” Unlike these four studies, the present research uses the Metaverse Proximity Score, introduced in the literature for the first time, and demonstrates that this score increased approximately fivefold between 2022 and 2025. It tests the ethical-concern trend using the Cochran–Armitage trend test and a bootstrap confidence interval, confirming that this increase is statistically significant (p < 0.05). Citation impact is assessed using an age-normalized citation rate, revealing that internationally collaborative publications have the highest citation rate. The study also provides data-driven projections for 2026–2030. The introduction of the concept of the “empathy paradox” into the literature shows that, despite the high citation impact of empathy-themed publications (37.4 mean citations), their thematic visibility declined markedly from its peak in 2023 (42.1%) to 2025 (21.4%). Spain has emerged as one of the most productive national centers in this field, generating a range of contributions at both empirical and conceptual levels. Pérez-Seijo (2021) examined 360° video productions by digital-native media outlets in Spain and found that the vast majority had failed to develop a storytelling language specific to the 360° format. In other words, producers had not meaningfully incorporated the spatial and immersive affordances of the new format into their narratives. They continued to produce content largely according to the logic of conventional video. The author also noted that the widespread claim that VR functions as an “empathy machine” capable of eliciting empathy in audiences still lacks robust empirical support. In fact, Pérez-Seijo (2021) offers an early observation anticipating the “empathy paradox.” Based on focus groups with academics and practitioners, Herrera Damas and Benítez de Gracia (2022) identified a strong consensus that immersive journalism remains at an experimental stage of production. López-García et al. (2017) situated immersive journalism within the broader framework of “technological journalism,” arguing that the profession’s dependence on technology is irreversible and requires new hybrid competencies. Finally, Aitamurto et al. (2022) experimentally demonstrated that augmented reality visualizations generate a strong sense of physical presence but do not produce a corresponding increase in objective knowledge acquisition.

3. Materials and Methods

The present study is grounded in a hybrid methodological approach that integrates bibliometric techniques with text-based content analysis (Donthu et al., 2021). The dataset was obtained on 1 March 2026, using the following Scopus query applied to the article title, abstract, and keywords fields: TITLE-ABS-KEY ((“metaverse” OR “immersive journalism” OR “virtual journalism”) AND “journalism”) AND (LIMIT-TO(DOCTYPE,“ar”) OR LIMIT-TO(DOCTYPE,“re”)) AND (LIMIT-TO(SUBJAREA,“SOCI”) OR LIMIT-TO(SUBJAREA,”ARTS”)). No language restriction was applied to the query, and the resulting corpus was multilingual (100 records in English, 34 in Spanish, 5 in Portuguese, and 1 each in Russian, French, Italian, and Arabic). The query was restricted to articles and reviews within the Social Sciences and Arts and Humanities subject areas, yielding 144 records. Manual screening identified one duplicate record. The duplicate was excluded from the analysis, and the analysis proceeded with 143 units. To address the possibility that keyword-based retrieval might also capture publications only indirectly related to immersive journalism, all 143 abstracts were screened for substantive relevance to immersive-technology content (VR, AR, 360-degree video, the metaverse, or related concepts), rather than relying solely on keyword co-occurrence. This screening identified 10 articles (7.0% of the research corpus) in which terminology related to immersive journalism appeared only in the author keywords, while the abstracts themselves contained no corresponding discussion of immersive technology. These articles were retained in the research corpus because they reflect the field’s own keyword-assignment practices and Scopus indexing conventions, with their limited thematic centrality explicitly acknowledged. To test how the results would change if these ten articles were excluded, the analysis was rerun on the remaining 133 articles using the same formulas and threshold values defined above. The growth rate of the Metaverse Proximity Score increased from 4.80× to 5.17× (2022–2025), the ethical-concern trend became stronger (Cochran–Armitage z = 2.38, p = 0.017 to z = 2.66, p = 0.008), and the 2025 ethical-concern proportion changed only slightly (6/14, 42.9% to 6/13, 46.2%). The empathy citation premium, the early- and late-period shares of empathy, Spain’s share of the corpus, and the proportional distribution across the three research periods remained largely unchanged (mean citations from 37.4 to 35.9; early-period share from 10.3% to 10.7%; late-period share from 31.0% to 31.5%; Spain’s share from 42.0% to 41.4%; period distribution from 9/55/79 to 8/50/75). This robustness check confirms that excluding the ten articles did not substantially alter the study’s main temporal and thematic patterns and, in some respects, even strengthened them. In all the analyses, a Python 3.11 environment was used by employing the pandas and numpy libraries for data processing, matplotlib for visualization, networkx for network analysis, and scikit-learn for text analysis.

The gradual development of the literature was reviewed by means of annual publication counts and year-on-year growth rates. On the basis of the shifts in publication density, three distinct research phases were identified: 2005–2017, 2018–2021, and post-2022. In order to reduce low-sample bias, years with fewer than five annual observations were excluded from proportional analyses. Metaverse orientation was measured through a score that was calculated on the basis of a keyword list that was pre-defined. When a relevant term appeared in the author keywords field, a score of +3 was assigned, and when it appeared in the abstract, a score of +2 was assigned.

$$\mathrm{M}\mathrm{P}\mathrm{S}=\sum _{\mathrm{t}\in \mathrm{T}}\left[2\cdot 1\left(\mathrm{t}\in \mathrm{Abstract}\right)+3\cdot 1\left(\mathrm{t}\in \mathrm{Keywords}\right)\right]$$

By using the recurring terminology in the literature, the Metaverse Proximity Score was constructed. The terms used in the procedure to score were as follows: metaverse, virtual world, avatar, digital twin, spatial computing, extended reality, mixed reality, web3, and blockchain. To assess the robustness of the Metaverse Proximity Score to the specific weighting scheme, the index was recalculated under three alternative weightings (keyword:abstract ratios of 2:1, 5:3, and 1:1, compared against the original 3:2 specification). The 2022–2025 growth ratio remained stable across all four specifications (4.8–4.9×), indicating that the observed metaverse-oriented shift is not an artifact of the particular weights chosen. To assess the robustness of the Metaverse Proximity Score to word selection, the indicator was recalculated using a narrower, metaverse-specific set of terms (metaverse, virtual world, extended reality) rather than the full nine-term list. The resulting 2022–2025 growth rate (3.75×) remained directionally consistent with, though approximately 22% lower than, the estimate obtained with the full-term set (4.80×). This indicates that the observed structural shift is not an incidental result dependent on the selected vocabulary. In the analysis of the thematic structure, keyword frequencies and co-occurrence relationships were employed (Cobo et al., 2011). The network analysis incorporated keywords used a minimum of five times and word pairs sharing at least two co-occurrences.

$$\mathrm{density}\left(\mathrm{k}\right)=\sum _{\mathrm{j}\in {\mathrm{C}}_{\mathrm{k}}}\mathrm{c}\left(\mathrm{k},\mathrm{j}\right),\hspace{1em}\mathrm{c}\left(\mathrm{k},\mathrm{j}\right)\ge 2$$

$$\mathrm{relevance}\left(\mathrm{k}\right)=\mathrm{f}\left(\mathrm{k}\right)$$

where f(k) denotes the frequency of keyword k in the corpus.

$$\mathrm{d}=\ln \left(1+{\displaystyle \frac{10\cdot \mathrm{density}\left(\mathrm{k}\right)}{\max \_\mathrm{density}}}\right)$$

$$\mathrm{r}=\ln \left(1+{\displaystyle \frac{10\cdot \mathrm{relevance}\left(\mathrm{k}\right)}{\max \_\mathrm{relevance}}}\right)$$

where max_density and max_relevance denote the maximum density and relevance values observed across all keywords.

Themes were positioned into four quadrants (motor, niche, basic, emerging/declining) on the basis of their density and relevance scores relative to the sample median. In parallel, keyword growth between the pre-2022 and post-2022 periods was assessed, enabling the identification of rising and declining thematic emphases across the two periods.

$${\mathrm{rate}}_{\mathrm{pre}}={\displaystyle \frac{{\mathrm{count}}_{\mathrm{pre}}\left(\mathrm{k}\right)}{{\mathrm{n}}_{\mathrm{pre}}}}$$

$${\mathrm{rate}}_{\mathrm{post}}={\displaystyle \frac{{\mathrm{count}}_{\mathrm{post}}\left(\mathrm{k}\right)}{{\mathrm{n}}_{\mathrm{post}}}}$$

where ${\mathrm{n}}_{\mathrm{pre}}$ denotes the number of publications in the pre-2022 period (2005–2021) and ${\mathrm{n}}_{\mathrm{post}}$ the number in 2022–2025.

$$\mathrm{ratio}={\displaystyle \frac{{\mathrm{rate}}_{\mathrm{post}}-{\mathrm{rate}}_{\mathrm{pre}}}{{\mathrm{rate}}_{\mathrm{pre}}}},\hspace{1em}{\mathrm{rate}}_{\mathrm{pre}}>0$$

$$\mathrm{ratio}=\min \left({\mathrm{n}}_{\mathrm{pre}}\cdot {\mathrm{rate}}_{\mathrm{post}},25.0\right),\hspace{1em}{\mathrm{rate}}_{\mathrm{pre}}=0$$

where the ratio is capped at 25.0 for keywords absent in the pre-2022 period.

Within the scope of citation analysis, the performance of publications was compared according to the variables of author count and geographical collaboration. International collaboration was defined as the presence of institutional addresses from more than one country within a single publication. The article-presence method was adopted for country counting: the country of affiliation was extracted as the token following the final comma in each semicolon-delimited affiliation string, with five known exceptions corrected via a manual institution-to-country lookup table, whereby each country was counted once if at least one author affiliated with an institution in that country contributed to the relevant article.

Citation values were normalized by dividing by article age:

$$\mathrm{age}=2027-\mathrm{Year}$$

$$\mathrm{velocity}={\displaystyle \frac{\mathrm{Cited}\_\mathrm{by}}{\mathrm{age}}}$$

where a one-year buffer is added to the data-collection year to partially offset the incomplete citation accumulation of recent publications.

Beyond raw normalized values, articles published from 2023 onward were additionally compared within their own publication-year cohort, using the 90th-percentile velocity threshold within each cohort:

$${\mathrm{threshold}}_{\mathrm{yr}}={\mathrm{quantile}}_{90}\left({\mathrm{velocity}}_{\mathrm{yr}}\right)$$

where quantile₉₀ denotes the 90th percentile of citation velocity within each publication-year cohort (subject to a minimum of three publications per year), allowing recent high-impact studies to be identified independently of their shorter citation-accumulation window.

The transformation of journalistic values was analyzed through a rule-based text-matching procedure drawing on pre-defined conceptual lexicons, comprising four traditional categories (objectivity, credibility, ethics and norms, source and witness accountability) and four experiential categories (empathy, presence/immersion, interactivity, user experience):

$$\mathrm{annual}\_\mathrm{pct}\left(\mathrm{category},\mathrm{yr}\right)={\displaystyle \frac{\mathrm{count}\left(\mathrm{category},\mathrm{yr}\right)}{\mathrm{n}\left(\mathrm{yr}\right)}}\times 100$$

where n(yr) denotes the total number of publications in the given year. The traditional and experiential averages were then computed as the mean of the four categories’ annual percentages within each group.

A separate lexicon addressing ethical concerns (manipulation, deception, trust, misinformation, disinformation, among related terms) was applied independently to track the prevalence of normative debates over time. The lexicons were constructed on the basis of highly cited works in the field, and the content validity of each category was assessed by the researchers. Throughout the analytical process, all calculations were verified against the raw data, and no manual coding was performed. To further verify that the observed year-on-year variation in ethical-concern prevalence reflects a genuine trend rather than sampling noise, a Cochran–Armitage trend test was conducted across the 2018–2025 period (z = 2.38, p = 0.017); Fisher’s exact test was additionally used to compare the 2025 subsample (n = 14) with the pooled 2018–2024 baseline (p = 0.013), indicating a statistically significant difference between the two distributions, despite the wider confidence interval associated with the smaller final-year sample. A 95% confidence interval was calculated for the 2025 ethical-concern estimate (21.4–71.4%). This calculation was performed using bootstrap resampling (10,000 iterations). An additional robustness check was conducted for the small subsample (n = 14). The Wilson score interval was then calculated. This interval was comparable but narrower (21.4–67.4%). This result is important: the wide uncertainty is not attributable to a methodological error. Rather, it should be regarded as a natural consequence of the small sample size. Finally, forward-looking projections were calculated for publication volume and thematic indicators. The dataset was retrieved on 1 March 2026. The two records published in 2026 are retained in the overall corpus (n = 143) and in raw publication counts (e.g., Figure 1); however, because 2026 constitutes an incomplete calendar year with negligible accumulation of publications and citations, it was excluded from all proportional, rate-based, and forward-looking projection analyses. The results of this analysis should therefore be interpreted as exploratory scenarios rather than precise forecasts. Although 2025 technically constitutes a complete calendar year, the number of publications (n = 14) is markedly lower than that observed during 2022–2024 (19–23). For this reason, the publication-volume projection was based on the 2022–2024 moving average rather than the 2022–2025 average. The uncertainty band, by contrast, was deliberately computed over the full 2022–2025 window: including the anomalously low 2025 count widens the resulting standard deviation, yielding a more conservative projection interval than would be obtained from the shorter, pre-2025 window alone.

Four indicators were projected using distinct approaches selected in accordance with the statistical properties of each:

(1)

for publication volume, the 2022–2024 moving average (±1 SD uncertainty band, computed over 2022–2025):

$$\mathrm{proj}=\mathrm{mean}\left({\mathrm{pub}}_{2022},{\mathrm{pub}}_{2023},{\mathrm{pub}}_{2024}\right)$$

$$\mathrm{band}=\mathrm{proj}\pm \mathrm{SD}\left({\mathrm{pub}}_{2022:2025}\right)$$

where pub_t denotes the number of publications in year t, and SD denotes the standard deviation over 2022–2025.

(2)

for metaverse keyword share—measured independently of the Metaverse Proximity Score, solely through the percentage of articles containing metaverse-proximate terminology in author-assigned keywords—the linear regression slope for the 2022–2025 period (β = +11 pp/year, R² = 0.96), capped at a 70% keyword saturation threshold for 2030:

$$\widehat{\mathrm{y}}={\mathsf{\beta }}_0+{\mathsf{\beta }}_1\cdot \mathrm{x}$$

$$\mathrm{projection}=\min \left(\widehat{\mathrm{y}}\left(2026–2030\right),0.70\right)$$

where x denotes the observation years 2022–2025 and 70% is the analytical saturation ceiling applied to the extrapolation.

The linear trend was extrapolated to 2030, with a 70% ceiling applied as an analytical saturation threshold to avoid mechanically projecting the observed linear increase beyond a plausible upper bound; this threshold is not interpreted as an empirical estimate of a maximum attainable share. A sensitivity check comparing ceiling values of 60%, 70%, and 80% confirmed that, given the observed slope (β = 11.03 pp/year, R² = 0.956), the projection reaches whichever ceiling is applied by 2027–2028 and remains at that ceiling through 2030 under all three scenarios (60%, 70%, and 80%, respectively). The 2030 estimate is therefore best understood as a function of the chosen analytical cap rather than an independent extrapolation of the underlying trend, and this constraint is acknowledged as a limitation of the projection method.

(3)

For the prevalence of ethical concerns, a fixed band based on the 2022–2025 mean ±1 SD-given that the 2025 value (43%) constituted an outlier peak exceeding the upper boundary of the band (36%)—the projection was initiated from 2026 onward, with a mean-reversion expectation duly noted:

$$\mathrm{proj}=\mathrm{mean}\left({\mathrm{eth}}_{2022:2025}\right),\hspace{1em}\mathrm{band}=\mathrm{proj}\pm \mathrm{SD}\left({\mathrm{eth}}_{2022:2025}\right)$$

(4)

for the share of VR/360° technology, a fixed band based on the 2022–2025 mean ±1 SD:

$$\mathrm{proj}=\mathrm{mean}\left({\mathrm{vr}}_{2022:2025}\right),\hspace{1em}\mathrm{band}=\mathrm{proj}\pm \mathrm{SD}\left({\mathrm{vr}}_{2022:2025}\right)$$

where ${\mathrm{eth}}_{\mathrm{t}}$ and ${\mathrm{vr}}_{\mathrm{t}}$ denote, respectively, the annual share of publications addressing ethical concerns and the annual share featuring VR/360° technology.

The artificial intelligence and empathy indicators, which exhibited high variance, were excluded from the projection scope on the grounds that no reliable trend could be identified. All projections are inherently indicative and might be affected by technological, institutional, or societal disruptions occurring after 2025.

4. Results

4.1. Structural and Historical Dynamics of the Field: Growth, Geography, and Power

4.1.1. The Development of the Literature and the Transformation of the Field

As shown in Figure 1, based on the annual publication distribution, immersive journalism research between 2005 and 2026 can be classified into three separate periods. In the first period, which was named the pre-emergence period (2005–2017; n = 9), the field exhibited an extremely limited level of output, which is characterized by a dispersed and unstable developmental trajectory. The low publication volume during this period might be because of the immaturity of consumer-level immersive hardware and the absence of a consolidated research community. The period from 2018 onward, corresponding to the growth phase (n = 55), is characterized by a rapid increase in annual publication counts. While the number of publications was 10 in 2018, the number of publications reached a first peak of 23 in 2020, a clear indication of a pronounced acceleration. A decrease was observed in 2021 (n = 12), which is thought to be due to difficulties caused by the pandemic in empirical data collection. The metaverse period, which started in 2022 (n = 79, including the two records from the incomplete 2026 calendar year), experienced a fundamental transformation of the field not only in terms of volume but also in terms of conceptual orientation. Annual publication counts ranging from 19 to 23 during 2022–2024 declined to 14 in 2025. This finding might indicate the field’s transition from an emerging specialization to an established sub-discipline. The 2026 value was excluded from the projection reference period because it represents only a partial calendar year. The 2022–2025 period was therefore used as the latest complete observation window.

4.1.2. Geographic Concentration: Spanish Dominance, Global Distribution, and the Citation Advantage of International Collaboration

The geographical concentration of immersive journalism research, and its relationship with citation impact, is shown across three panels in Figure 2.

Panel A exhibits Spain’s annual share of publications in the period ranging from 2018 to 2025. Spain’s prominence within the field reached its peak during the period from 2021 to 2022, in which 58% and 62% of all publications, respectively, originated from Spain, a proportion indicating that the field exhibited a strong concentration in the Spanish research context during this period. However, this concentration has not remained constant over time. From 2023 onward, a steady decline was observed in Spain’s share, and it fell to 14% in 2025, a trend suggesting that the field is progressively assuming a more decentralized and internationally distributed character. Panel B displays publication output by country across the corpus, to which researchers from 32 countries have contributed. Spain leads by a considerable margin with 60 articles, followed by the United States (n = 20) and the Netherlands (n = 11). Brazil (n = 10) and Austria (n = 9) complete the top five. Bilateral co-authorship remains limited overall: 22 distinct country pairs were identified across 22 internationally co-authored publications, which corresponds to 15.4% of the dataset. This configuration indicates that geographical diversity has not yet been fully translated into strong collaborative networks. The comparison of mean citation counts by geographical origin is given in Panel C, and here a striking asymmetry is seen. Spain-based studies (n = 60, the aggregate of Spain-only and internationally co-authored Spain publications) occupy the most dominant position in terms of publication volume; among the three mutually exclusive categories shown in Panel C, Spain-only studies exhibit the lowest academic impact, with a mean of 11.5 citations. By contrast, internationally collaborative works co-authored by Spanish and foreign institutions (n = 6) generate by far the highest impact, with a mean of 39.7 citations-a value approximately 3.5 times that of Spain-only publications. Non-Spanish country contributions (n = 83) occupy an intermediate position with a mean of 25.0 citations. In general, these findings indicate that geographical dominance in terms of production volume does not translate into a corresponding advantage in terms of academic impact. By contrast, internationally collaborative publications were associated with substantially higher citation impact, even with a comparatively modest number of publications. It follows that quantitative productivity and qualitative impact are markedly dissociated in immersive journalism research. The findings suggest that international collaboration may be associated with greater citation visibility, though the small number of internationally co-authored articles (n = 6) warrants caution in generalizing this pattern. The citation-impact claim underlying the empathy paradox was also tested for age bias: empathy-themed articles retained a substantially higher mean citation rate (5.91 citations/year) than non-empathy articles (2.43 citations/year), confirming that the observed pattern is not an artifact of publication age. The geographic citation comparison in Figure 2C is based on raw citation counts, which favor earlier-published studies. To address this issue, the same comparison was recalculated using an age-normalized citation rate (citations per year since publication). Three mutually exclusive publication categories were used in the comparison: internationally coauthored articles with at least one Spanish institutional affiliation, articles authored exclusively by researchers affiliated with Spanish institutions (“Spain-only”), and articles with no Spanish institutional affiliation (“non-Spain”). The ranking observed with raw citation counts was preserved: internationally coauthored articles involving Spain retained the highest mean citation rate (6.78 citations/year), followed by non-Spain articles (3.96) and Spain-only articles (1.73). This finding confirms that the citation advantage stems specifically from international collaboration, rather than from Spanish institutional affiliation itself, and is not an incidental result of publication age. Spain’s central position in the corpus, as measured by publication volume, is therefore analytically distinct from its citation performance. Citation performance depends on whether Spanish research is conducted collaboratively or in isolation.

4.2. Mapping the Conceptual Terrain: A Literature Evolving Toward the Metaverse

4.2.1. The Metaverse Proximity Score: The Temporal Trajectory of Metaverse-Oriented Conceptual Change

The annual mean values of the Metaverse Proximity Score are presented in Figure 3 for the period ranging from 2018 to 2025. This is an original indicator developed within the scope of the current study. This score was calculated based on the density of metaverse-proximate terminology in the abstract and keyword fields of each article, and thus the transformation in the field’s conceptual orientation was rendered visible. The score was calculated at the article level. Each metaverse-proximate term was counted once within the author-keyword field and once within the abstract, irrespective of the number of times it occurred within a field. This procedure was adopted to prevent repeated mentions of the same term from disproportionately increasing an article’s score. Nine metaverse-proximate terms were included in the lexicon (metaverse, virtual world, avatar, digital twin, spatial computing, extended reality, mixed reality, web3, blockchain), yielding a theoretical maximum score of 45 per article (9 terms × [2 + 3]). The findings reveal that during the period ranging from 2018 to 2021, the score increased only incrementally, from 0.00 to 0.33. During this period, immersive journalism research was shaped primarily within the framework of the VR and 360-degree video paradigm, independently of metaverse discourse. This shift coincides with Meta’s corporate rebranding at the end of 2021. The score subsequently rose to 0.48 in 2022, 0.89 in 2023, 1.48 in 2024, and 2.29 in 2025. The fivefold increase indicates a substantial strengthening of metaverse-proximate terminology in the literature. While the score does not by itself establish a restructuring of research questions, the magnitude and persistence of the increase provide quantitative evidence of a changing conceptual orientation. The fact that this high mean score (2.29) is derived from as few as 14 articles in the 2025 data further indicates that metaverse-oriented studies, though still quantitatively limited, occupy a conceptually decisive position within the field.

4.2.2. Keyword Growth Rates: Rising and Declining Themes

In Figure 4, a comparative analysis of the concepts used more or less frequently by researchers before and after 2022 is presented. Among the rising themes, the metaverse (+10.5×) shows the strongest growth by a considerable margin, followed by gamification (+3.2×), storytelling (+3.1×), journalistic innovation (+2.2×), innovation (+1.4×), interactivity (+1.0×), and journalism (+0.9×). The notably high magnitude of the metaverse ratio is not incidental, as the concept had barely secured a place in the literature before 2022. When the declining themes are examined, it is seen that presence (−0.7×), empathy (−0.7×; based on a comparatively small total occurrence count), digital journalism (−0.6×), immersion (−0.4×) and virtual reality (−0.2×) have undergone a relative decline after 2022, even though several of these terms remain central to the field in absolute terms. Although empathy simultaneously remains one of the most highly cited themes in the literature, its use as a keyword declines, which is one of the first empirical indicators of the empathy paradox addressed in a subsequent section of the current study. The decline of digital journalism, in turn, may be seen as a strong indication that the field has begun to distance itself from the broader digital media framework and to develop a discourse specific to immersive journalism and the metaverse.

4.2.3. Thematic Map: The Strategic Positioning of Research Topics

Figure 5 shows a thematic map that positions research topics along two dimensions: internal density (x-axis: the degree of conceptual development of the theme) and relevance (y-axis: frequency of occurrence in the literature). Bubble size denotes keyword frequency. The motor themes which are located in the upper right quadrant demonstrate high density and high relevance, constituting the conceptual core of the field. The findings of the study indicate that this core consists of the following concepts: immersive journalism (n = 109), virtual reality (n = 77), 360-degree video (n = 33), augmented reality (n = 22), immersion (n = 16), metaverse (n = 13), digital journalism (n = 12) and presence (n = 11). In this configuration, it is demonstrated that the field has developed in an integrated manner around both its technological dimensions (VR/AR, metaverse) and its experiential dimensions, centered on the concepts of immersion and presence—the latter denoting the sensation of “being there” in an environment in which the user is not physically present. The concepts that are positioned in the lower left quadrant—including artificial intelligence, technology, immersive storytelling, innovation, interactivity, and journalism innovation—might be regarded as emerging or transitional themes, which are characterized by low density and low relevance. No keyword in this analysis meets the criteria for the niche-themes quadrant (high density, low relevance). Journalism (n = 10) instead occupies a borderline position at the edge of the basic-themes quadrant—its density score falls below the sample median, indicating that despite its high frequency of occurrence in the literature, this concept has not developed dense co-occurrence ties with the field’s other themes. When considered as a whole, the thematic map shows that while the field is structured around a robust core including immersive journalism and its associated technologies, new technological and narrative forms are still at an early stage of development.

4.3. Where Is Journalism Heading: Empathy, Ethics, and the Measurement of Impact

4.3.1. The Empathy Paradox and the Transformation of Journalistic Values

The transformation of values in immersive journalism research is illustrated in Figure 6 across three panels. In Panel A, experiential values—including empathy, presence, interactivity, and user experience—are represented at markedly higher rates than traditional journalistic values throughout the period examined. In 2018, experiential values accounted for 42% while traditional values remained at 10%. The gap reached its widest point in 2023, when experiential values peaked at 47% while traditional values stood at 18%. Following 2024, both value groups receded slightly, with experiential values settling at 38% and traditional values remaining at around 18% by 2025—a narrowing that is more pronounced on the experiential side.

In Panel B, the mean academic impact of the value categories is presented. Empathy (37.4 citations; n = 33) and user experience (36.1; n = 17) attract the highest mean citation counts, closely followed by credibility (33.5; n = 17)—a traditional value whose strong citation performance suggests that questions of truth—claims in immersive formats remain highly salient to researchers. Presence/immersion (22.2; n = 120), ethics and norms (22.1; n = 20), and interactivity (20.0; n = 40) occupy an intermediate tier, while objectivity (17.4; n = 16) and source/witness accountability (13.4; n = 39) yield the lowest returns. This distribution indicates that visibility within the literature and academic impact are not directly correlated—a pattern most conspicuous for empathy, which remains among the most cited themes despite its declining representational share, a tension we term the ‘empathy paradox.’

In Panel C, the trajectory of ethical concerns is presented. The proportion of articles addressing manipulation, trust, and disinformation rose from 0% in 2018 to 22% in 2020, then followed a volatile course (2021: 8%; 2022: 19%; 2023: 5%; 2024: 22%) before climbing to its highest recorded level, 43%, in 2025 (n = 14). Given the small subsample size for this final year, this estimate should be treated as indicative of a rising trend rather than a precise figure. Taken together, these findings suggest that while experiential values continue to dominate the research agenda, ethics- and accountability-based discussions have gained substantial momentum in the most recent period, particularly as the metaverse-oriented turn documented in Figure 3 has taken hold.

4.3.2. Citation Impact: Time-Adjusted Measurement

As a result of raw citation counts, a systematic disadvantage is introduced for recently published work, given the constrained accumulation window available to newer publications. In Figure 7, this bias is addressed by presenting citation values that are normalized per year and identifying, via two panels, both established and emerging high-impact studies. Panel A ranks the top 15 articles according to normalized values. Exploring Immersive Experience in Journalism (2018) leads the list with 27.8 citations per year and 250 total citations. Two articles share the second position with 21.5 citations per year: Being There in the Midst of the Story (2017), with 215 total citations accumulated over ten years, and Intimacy and Emotions in Podcast Journalism (2023), with 86 total citations accumulated over four years. The latter, despite occupying a mid-ranking position in the raw citation order, is elevated to the top of the normalized ranking—a finding that clearly demonstrates the analytical value of time-adjusted measurement. The fact that several of the top 15 articles comprise empathy- and immersion-focused studies is also remarkable, as it indicates that these themes are not only subject to frequent scholarly attention but also generate high academic impact. In Panel B, recent yet high-performing studies from the 2023–2026 period are listed. These studies rank within the top 10% of their respective publication year cohort. Behold the Metaverse: Facebook’s Meta Imaginary (2025; 9.5 citations per year; 19 total citations) and The Worthwhileness of Immersive Journalism (2025; 8.0 citations per year; 16 total citations) come to the fore in this group. The location of these studies at the intersection of metaverse critique and the interrogation of value in immersive journalism provides a strong indication of the direction in which the field’s emerging high-impact research trajectory is heading.

4.4. Forward-Looking Projections for Immersive Journalism Research, 2026–2030

The Near-Term Trajectory of the Research Agenda (2026–2030)

The projections toward 2030 suggest a possible scenario in which publication volume may stabilize while metaverse-related themes continue to grow in influence. Four indicators were projected for 2026–2030, based on observed data from 2018 to 2025, as shown in Figure 8. Each indicator has been modeled using a distinct method, which was selected in compliance with its statistical properties, with uncertainty bands set at ±1 standard deviation. Panel A presents a projection scenario in which publication volume remains within the range of 18–24 publications per year during 2026–2030. On the basis of the 2022–2024 moving average (21 publications per year), this projection shows that the field has gone through a transition out of the rapid growth phase of 2018–2021 and entered a phase of consolidation. Although 2025 constitutes a complete calendar year, its publication count was markedly lower than the 2022–2024 range and was therefore not used as a projection reference point (see Section 3). Panel B presents a scenario in which the share of the metaverse keyword follows the linear trend observed during 2022–2025 (β = +11 percentage points/year, R² = 0.96), approaching a saturation threshold of 70% by 2030. Metaverse terminology, which was entirely absent in 2021, reached 43% in 2025, and this momentum is assumed, under this scenario, to continue throughout the projection period, which is consistent with the concept becoming increasingly established in the literature. In Panel C, it is revealed that the prevalence of ethical concerns showed high volatility during the 2022–2025 period (σ = 13 pp). The 43% peak recorded in 2025—based on a small subsample of 14 articles—sits conspicuously above the projection band derived from the 2022–2025 mean (9–36%) and should be interpreted with corresponding caution. Therefore, the projection was started from 2026 onward, with a mean-reversion expectation duly noted. Under this scenario, ethical concerns are projected to remain within the 9–36% range during 2026–2030. Nevertheless, when the volatile nature of the series is taken into consideration, point estimates are not considered meaningful. Panel D presents a scenario in which the share of VR/360° technology remains at the stable level observed during 2022–2025 (mean = 59%, σ = 5 percentage points) through 2030. The narrow uncertainty band (54–64%) is an indication that the structural prominence of this technology within the literature is likely to persist in the near future. The projection scenarios suggest that the volume of immersive journalism research may reach quantitative stability during 2026–2030, while metaverse-oriented conceptual directions and ethical inquiry may become increasingly prominent on the research agenda.

5. Discussion

Taken together, the findings presented in Section 4.1, Section 4.2, Section 4.3 and Section 4.4 point to a field undergoing simultaneous quantitative consolidation and conceptual reorientation toward the metaverse. This section interprets these patterns in relation to the field’s normative trajectory and situates them within recent bibliometric studies of immersive journalism. The rise in ethical concerns to 43% in 2025 (n = 14, interpreted cautiously) suggests that immersive journalism may be shifting from an experience-oriented field of inquiry toward a discursive space in which normative interrogations occupy a more central position. This increase is not linear. It was 22% in 2020, then decreased to 8% in 2021, rose to 19% in 2022, declined to 5% in 2023, rose again to 22% in 2024, and finally reached its highest recorded value, 43%, in 2025—an estimate that warrants caution given the small subsample (n = 14) for that year. This fluctuating pattern indicates that ethical concerns are prone to rise reactively in response to specific developments in the field before subsiding, while the 2025 peak indicates increased recent visibility of ethical concerns, although the available data do not establish whether this increase represents a durable structural change. This pattern suggests that the metaverse-oriented turn documented above coincides with a growing emphasis on trust, manipulation, and accountability in the field’s normative debates, though the present design does not establish a causal link between the two. This resurgence stands in contrast to Welskop (2025), who identified the disappearance of ethical issues from the literature after 2019 as a research gap; the extended window adopted here suggests that the normative turn Welskop’s earlier cut-off could not capture has since intensified, consistent with Welskop’s own observation that ‘metaverse’ constitutes a distinct, recent cluster (2022–2023). More broadly, the empathy paradox identified in this study suggests that academic influence and thematic visibility do not necessarily evolve in parallel. This divergence in how prominence is measured is not confined to the present corpus: Alpsoy (2026), for instance, classifies ‘immersive journalism’ and ‘virtual reality’ as basic rather than motor themes, a divergence likely reflecting Alpsoy’s broader, unfiltered corpus (n = 200, 1997–2025) relative to the more narrowly defined corpus used here. Ideas that have shaped the intellectual foundations of a field may gradually lose prominence within the active research agenda as new technological priorities emerge. Finally, international collaboration, despite its limited prevalence (only six internationally co-authored articles, n = 6), creates a notable citation advantage, and this indicates that collaboration-based knowledge production assumes a more critical role in the development of the field than quantitative output alone. This pattern of consolidation aligns with, yet extends, Sanchez-Acedo et al. (2023), who—working with a narrower window (2017–2022)—describe the field as being in an ‘initial exploratory stage’; the present findings suggest this fragmentation has partially consolidated since 2022 around the metaverse-oriented paradigm identified here, even as the field’s normative vocabulary remains in flux.

Limitations

Several methodological constraints should be considered when interpreting these findings. First, three key findings rest on comparatively small subsamples: the 2025 ethical-concern estimate (43%, n = 14) carries a wide 95% confidence interval (21.4–71.4%); the citation advantage of international collaboration is based on only six co-authored articles (n = 6); and the declining growth ratio for the empathy keyword (−0.7×) is likewise calculated from a small total occurrence count. Second, the forward-looking projections rely on a short eight-year window (2018–2025) and simplifying assumptions—a fixed mean ± 1 SD band, and a linear trend capped at an analytically chosen 70% saturation ceiling—and should be read as exploratory scenarios rather than precise forecasts. Third, the corpus is drawn from a single database (Scopus), which may not capture all relevant literature indexed elsewhere. These constraints bound the confidence with which the magnitude, rather than the direction, of the findings can be generalized.

6. Conclusions

The current study demonstrates that the immersive journalism literature has not only experienced a quantitative expansion but also a significant conceptual and normative transformation. The findings indicate that the field has evolved from its dispersed configuration during the 2005–2017 period through a phase of rapid growth following 2018 and into a new metaverse-oriented research paradigm from 2022 onward. This transformation is quantitatively corroborated by the sharp increase in the Metaverse Proximity Score, confirming that the metaverse functions as a structural rather than a merely terminological shift in the field. The fact that this rupture coincides with Meta Platforms’ corporate rebranding in October 2021 provides a noteworthy contextual observation, without implying a causal relationship. Among the most striking findings of the study is the “empathy paradox”—the sharp decline in the representational presence of the empathy theme, calculated from a comparatively small total keyword count, despite its high citation impact in the literature. This finding demonstrates that there is no direct correspondence between academic impact and the research agenda. A related pattern emerged for international collaboration: despite involving only six co-authored articles (n = 6), internationally co-authored work was associated with a markedly higher citation impact than domestically authored studies, suggesting that collaborative knowledge production—though still limited in scale—plays an outsized role in shaping scholarly visibility in this field. Taken together, these findings extend previous bibliometric studies by demonstrating not only how immersive journalism has expanded, but also how its underlying conceptual priorities have been reconfigured over time. In general, these findings show that immersive journalism has become not merely a field of technological innovation but a developing sub-discipline located at the intersection of ethical, epistemological, and institutional transformations. In addition, the short-term projection scenario based on the 2018–2025 period suggests that publication volume may stabilize within the range of 18–24 publications per year during 2026–2030. Under this scenario, the share of the metaverse keyword continues its current trend (β = +11 percentage points/year). It approaches a saturation threshold of 70% by 2030, suggesting that the concept’s structural prominence in the literature may continue in the near future. Under this scenario, the prevalence of ethical concerns is projected to decline from the 2025 estimate of 43% (based on a small subsample) toward the 9–36% range. Under this scenario, VR/360° technology may maintain its structural dominance within the 54–64% range.

The artificial intelligence and empathy indicators, which exhibit high variance, were not included in the projection scope as no reliable trend could be identified for either. Taken together, the projections suggest that future developments in immersive journalism may be shaped not only by technological innovation but also by emerging normative and institutional concerns surrounding its use. Although the projections presented in this study should be interpreted with caution, they provide a useful framework for discussing possible directions in the evolution of immersive journalism. Future technological, institutional and societal developments may alter these trajectories, but the findings nevertheless indicate that debates surrounding ethics, trust and the metaverse are likely to remain central to the field.