Multimodal digital ethnography in platform cultures: a systematic review of integrating text, images, audiovisual content, and interaction data

INTRODUCTION

Everyday sociality is now conducted through platform interfaces that privilege images, short video, sound, live interaction and algorithmic circulation over the written post that once anchored online ethnography. This shift is not merely stylistic. Studies of platform cultures report that meaning is distributed across a video's caption, its visible hashtags, its engagement counters, its soundtrack, its comment thread and the emojis and symbolic responses that carry affective weight within that thread.(1)

Digital ethnographic work on heritage circulation shows that platform infrastructures, their persistence rules, recommendation systems and metadata regimes, shape what becomes visible and what can be studied at all.(2) Where multimodal resources are systematically combined, they do not simply decorate a message: in Colombian vaccine-sceptic content they functioned as the operative mechanism through which false context, manipulated material and fabricated material were assembled into emotionally saturated narratives,(3) and in Brazilian Instagram memes scientific content was entangled with humour, irony and political criticism in ways that no single-mode reading could recover.(4)

Netnography and its neighbouring traditions have long claimed to study culture online, but the labels have multiplied faster than their procedures have been specified. Recent methodological reviews distinguish virtual ethnography, digital ethnography, netnography and post-application programming interface ethnography along two axes: the degree to which the object is a bounded group or culture, and the degree to which the platform's own agency and affordances are treated as analytic material.(5)

A scoping review of the field similarly found that digital transposition is not a homogeneous approach, with differences in object, field, access strategy and analytic orientation, and warned against the tendency to decontextualise culture when fields are assembled from tags and hashtags instead of communities.(6)

Within the included literature, this distinction is drawn explicitly: multimodal ethnography, which follows a community across online and offline settings, is contrasted with digital ethnography focused on interaction within a platform, with the former argued to produce deeper immersion and the latter better suited to assessing platform inclusiveness and participant roles.(7) The methodological literature also warns that existing guidance rarely specifies reporting practices for multimodal online data, including consent, researcher positioning and protection of participant identities.(8)

Against this background, a specific problem recurs. Multimodality is frequently invoked as a rationale while the actual analysis remains textual. Several included studies are candid about this. A multi-sited netnography of assisted reproductive technology communities in Italy acknowledges that its textual focus left images, videos, voice messages and memes unanalysed.(9) A large computational digital ethnography of a Chinese women's fitness community states that its pipeline prioritised textual data and excluded visual and multimodal elements such as photographs, videos, memes and emojis.(10) A netnography of YouTube discourse on a political inauguration notes that multimodal elements were not integrated.(11)

Such admissions are analytically productive: they indicate that the field possesses a vocabulary for multimodality but an uneven repertoire of procedures for handling it. Where multimodal integration does occur, it tends to be achieved either by treating one mode as context for another, as when video serves as frame and comment as unit of analysis,(1) or by treating a single post as a composite semiotic unit in which verbal, visual and typographic resources are coded together.(12) The methodological question is therefore not whether digital evidence is multimodal, but how integration is designed, justified, executed and reported.

The analysis examines how netnographic and digital ethnographic studies published between 2020 and 2026 collect, combine, interpret and report multiple forms of digital evidence from platform cultures, and uses those practices to derive a methodological framework for multimodal netnography. The choice of window reflects the migration of platform communication towards visual, audiovisual, ephemeral and interactive formats, which makes pre-2020 procedural guidance partially obsolete; seminal older work is therefore treated only as conceptual background and is not counted in the primary evidence base.

The study is designed to build a framework, not to test a hypothesis, and it is bounded by the availability of retrievable full texts: the studies available for synthesis form a subset of the eligible literature and should not be read as a representative sample. We first map what the included studies actually did across sampling, collection, archiving, transcription, annotation, cross-modal analysis, ethics and reporting; we then translate recurring patterns into an explicitly bounded framework whose elements are labelled as supported, partially supported or proposed for testing. Throughout, we keep interpretive claims proportional to a body of evidence that is qualitative, heterogeneous and constrained by substantial full-text retrieval losses.

METHODS

This systematic review examined how netnographic and digital ethnographic research integrates multiple forms of digital evidence in platform cultures. The analysis was organised around six domains: platform affordances and cultural context; multimodal sampling and case selection; collection and archiving; transcription, translation and annotation; cross-modal analysis and triangulation; and ethics, reflexivity and reporting.

Study selection and analysis

Retrieval yielded 456 records. Removing 141 duplicates left 315 records for title-and-abstract screening. Full texts were sought for potentially relevant reports, 48 were assessed for eligibility, and 31 studies met the inclusion criteria. The identification and selection process is presented in Figure 1.

Data extraction covered study design, population, setting, platform, sampling strategy, modalities, collection and archiving procedures, transcription and translation, annotation, analytical integration, ethics, reflexivity, and reported limitations. Appraisal was domain-based and focused on reporting transparency and methodological coherence without reducing appraisal to a single numerical risk-of-bias score.

Synthesis proceeded through descriptive mapping of study characteristics followed by thematic comparison across the six analytical dimensions. The framework presented in the review was derived from recurring practices and gaps in the included literature; contextual sources were used only to support interpretation and were not treated as included result studies.

RESULTS

DISCUSSION

All work cited in this section that lies outside the core set is contextual or comparative. It is adjacent literature used to situate, contrast and extend the synthesis, it is not part of the included evidence base, and where it suggests extensions, these are presented as proposals from adjacent literature, not findings from this review.

The framework derived above converges with, and in several respects extends, methodological work outside the included evidence base. The distinction between contextual fields such as communities, blogs and forums and decontextualised narratives aggregated through tags has been identified as a major axis of divergence in digital ethnography, together with differences in access strategy and in whether analysis proceeds through text processing or hermeneutic interpretation.(6) Our included studies sit predominantly on the contextual side: the strongest multimodal designs observed named communities, media channels, campaigns, platforms or a single curated archive, and several explicitly refused aggregation across tags in favour of sustained engagement with a bounded field.(9,10,25)

Where hashtag or keyword aggregation was used as the primary field construction device, studies acknowledged that a single keyword may exclude synonymous or locally denominated discourse,(2) which matches the broader warning in the adjacent literature that decontextualised fields risk detaching meaning from its conditions of production. This comparison is contextual: it interprets a pattern already present in the reviewed study set; it does not supply evidence for the pattern.

The tension between group-centred and platform-centred designs is likewise well established in adjacent work, with existing typologies positioned along the boundaries of the culture investigated and the agency and affordances of the platform.(5) Our synthesis suggests that multimodal netnography requires both poles simultaneously: the included studies that integrated the most modalities were also those that theorised platform infrastructure explicitly, treating recommendation, persistence, editability and association as analytic objects (2) and treating affordances as agents that shape what can be said and by whom.(9)

Adjacent work argues that hashtags, platform recommendations, symbolic codes and cross-platform pathways should be treated as field signs, not merely recruitment tools, and that multi-platform protocols should be staged according to intrusiveness, moving from low-intrusion public mapping through ethically managed engagement in semi-private spaces towards direct interaction and triangulation where appropriate.(35) Several included studies effectively implemented such staging without naming it, as when a study moved from legal and policy analysis, through structured audit of publicly accessible content, to platform policy review and media investigation while deliberately excluding encrypted spaces.(24) Naming that progression explicitly would improve comparability across studies; the recommendation is ours, and the staging logic is imported from adjacent literature and is not an observed convention in the included studies.

Coding and annotation are the area where the wider methodological literature offers the most immediately transferable guidance. A dedicated study of coding social media data identified five recurring challenges: posts function as coherent multifaceted units of meaning, not discrete sentences; a single post frequently requires multiple simultaneous codes; reliable codebooks require several rounds of iteration; inter-rater agreement is lower than in conventional coding, with a 75% agreement rate requiring extensive discussion and refinement; and engaging more than two coders materially improves rigour.(36) Our synthesis finds partial resonance.

The multi-coding imperative is clearly visible where verbal and symbolic comments were coded under a shared framework with distinct analytical levels (1) and where multimodal narratives were coded simultaneously for deceptive technique, narrative orientation, form and meaning.(3) Iterative codebook development, by contrast, was rarely described, and reliability reporting was uneven: one study reported coefficients including 0.752 between human coders and lower values between model and human coders,(10) another relied on consensus validation by four researchers,(4) and others reported no coefficient while acknowledging the absence as a limitation.(16) The adjacent coding literature also notes that its own framework was developed for textual data and that visual and audiovisual elements warrant further methodological development.(36) Our evidence base suggests that this gap persists and that it is aggravated by the modal heterogeneity of platform content: the annotation problems posed by a gestural video vignette,(16) a meme whose meaning depends on intertextual reference (23) and a comment thread in which a flag emoji carries positional meaning (1) are not variants of one coding problem.

Emoji and reaction data illustrate the point. Within our core evidence, emojis were folded into the surrounding thread as affective and functional markers (1) and appeared as one resource among many in multimodal misinformation narratives.(3) Outside the core set, a study of Jordanian–Palestinian Facebook discourse treats emojis as one layer in a three-scale protocol, micro linguistic strategies, meso semiotic affordances and macro temporal and network patterns, and pairs that protocol with reproducible detection scripts while explicitly labelling repetition, clustered timestamps and repeated links as candidate indicators of amplification, not proof of orchestration.(37)

That is precisely the kind of explicit evidential status our core studies rarely provide for non-verbal material, and it suggests, as a proposal drawn from adjacent literature, not a finding of this review, a tractable extension of dimension five: code emojis both functionally, as thread-level affective markers, and distributionally, as candidate indicators whose inferential weight is declared. Related work on livestreaming shows what sustained analysis of platform-specific pictorial tokens can achieve, tracing how emotes circulate and are repurposed across streaming communities as a mode of interaction through which power relations are reproduced and contested.(38) Since our core evidence base contains no sustained livestreaming ethnography, this body of work supplies the missing comparator, not a result. Work on short-form video similarly treats coded multimodal strategies, orthographic manipulation, emoji encoding, euphemistic reformulation and multimodal masking, as coordinated negotiations with algorithmic governance instead of isolated lexical innovations (39); this is adjacent context for our finding that non-verbal resources are constitutive, not decorative, not independent corroboration of it.

Ephemerality and audiovisual multimodality constitute the second clear divergence between our findings and adjacent literature. Where our included studies largely worked with persistent, capturable material and acknowledged deletion, sorting and encryption as threats to the corpus,(1,24) methodological work on Snapchat has developed an appnography that adapts netnographic procedure to multimodal, user-generated, self-destructing data, arguing that smartphone applications present distinct data-access conditions that require purpose-built designs instead of inherited ones, and reporting both empirical and conceptual insights derived from that adaptation.(40)

A participatory alternative is the digital diary, in which participants document their own media practices over time through text, images, photographs, voice recordings and video, retaining control over what is shared and how it is interpreted, with the resulting material reviewed collaboratively with researchers.(41) Our core evidence contains no comparable participatory multimodal method: the closest instances are trace interviews used to strengthen observational data (32) and life stories whose voluntary, non-anonymous publication was a selection criterion, not a collaborative production.(15) Given that participants' own audiovisual elaboration has been argued in adjacent work to move analysis beyond representation and to give voice to participants when text, sound and image interplay,(42) the absence of participatory multimodal designs is better interpreted as a gap in the netnographic evidence base and does not show that such designs are unavailable.

Generative and synthetic media introduce a third divergence. Core studies already document artificial-intelligence-fabricated images as one route through which non-consensual material is produced,(26) and a digital ethnography of synthetic portraits used as reverse image search queries shows that search engines do not discriminate between authentic and realistic synthetic images, producing a dialectic between increasing precision and increasing scepticism and generating bureaucratic apparatuses of verification.(14) Outside the core set, virtual ethnography has been applied to generative-AI-assisted digital multimodal composing by platform language entrepreneurs, combining analysis of produced videos with viewer comments to argue that translanguaging spaces built with generative tools may serve attention capture as much as pedagogy.(43)

This compounds the annotation problem posed by our framework's fourth dimension: if part of a corpus may be synthetically generated, then mode-specific transcription and annotation must be accompanied by provenance assessment, and the practice of reading viewer comments as validation of production intent (32,43) becomes more fragile. This implication is ours; it is consistent with the reviewed evidence on synthetic imagery but is not demonstrated by it.

Comparison with the online ethnographic protocol literature clarifies where our framework is under-specified. Work on hidden and stigmatised populations develops a staged, ethics-attentive, multi-platform protocol that treats cross-platform pathways as field signs, operationalises ethical governance and data security measures, and explicitly declines to claim universal applicability, offering instead a structured, reflexive starting point requiring adaptation to population, platform, legal context and risk level.(35) Passive virtual ethnography conducted comparatively across two language communities has shown how offline parenting cultures shape practices of sharing children's textual and audiovisual content, and how the tension between privacy and openness is negotiated by parents on behalf of people who cannot consent.(44)

Our core evidence reproduces this privacy–openness tension in a different key: privacy concerns drove women on marriage platforms to avoid real photographs and to expect the same of prospective partners,(27) and a couple reconfigured their entire platform presence by deleting videos and reframing profiles, changing what the researcher could observe.(31) Where the adjacent literature proposes staged consent and governance mechanisms, our framework currently offers only a decision point, which we regard as its principal specification gap.

Finally, the wider literature usefully restates the reporting problem. Guidance for conducting netnography in second language acquisition research identifies recruitment and consent, researcher positioning including emic perspective and ethics, the analysis of online multimodal data, and general reporting practices such as protecting participant identities as the central considerations.(8) Multimodal analysis embedded in virtual ethnography of crisis-related videos has shown that analysing a small number of artefacts closely can yield claims about participatory practice that a broader textual corpus would miss,(45) which mirrors the logic of included studies that retained twelve memes from an initial universe of 83 (4) or five purposively selected videos from a channel archive.(21)

Hashtag-based sampling of top-ranked videos under a movement hashtag, combined with multimodal rhetorical analysis focused on the mediality of the body, demonstrates the same trade-off between breadth and semiotic depth (46) and parallels the core finding that embodied resources perform rhetorical work in short video.(16) What the adjacent literature adds is the insistence that such trade-offs be stated as design choices instead of being reported as incidental properties of the dataset.

A research agenda.

Table 2 sets out the gaps that follow most directly from the synthesis, ordered by the strength of the underlying evidence gap, not by novelty, together with the kind of design that could close each one. In each case the illustrative question is phrased so that it can be answered with multimodal evidence, not merely multimodal ambition. Each row distinguishes anchors that are included core studies, part of the reviewed evidence base and in some cases the source of the limitation motivating the question, from anchors that are adjacent literature cited only as contextual exemplars of feasibility. Adjacent anchors are not evidence that the proposed design has been tested within the netnographic corpus reviewed here.

Table 2. Research agenda for multimodal netnography.

Note. "Core" designates a study included in the reviewed evidence base; "core, conceptual synthesis" designates an included contribution that is not an empirical case; "external" designates adjacent literature used only as contextual exemplar, which is outside the included study set and does not constitute evidence of feasibility from this review.

CONCLUSIONS

Multimodal digital ethnography is strongest when different forms of evidence are analytically connected instead of simply being collected together. Across the reviewed studies, contextualisation, explicit sampling decisions, documented translation and annotation, and cross-modal comparison were central to preserving meaning across text, image, video, and interaction data.

Important gaps remain in the treatment of audio, live interaction, ephemeral content, and the ethical consequences of preserving or reproducing multimodal material. Greater transparency about what is captured, transformed, omitted, or translated would improve the interpretability and reproducibility of future studies.

The six-dimension framework proposed here provides a structured basis for designing and reporting multimodal digital ethnography. It should be treated as a synthesis of current practice and a guide for empirical testing, not as a validated methodological standard.

References

1. 1 Rudianto R, Margolang AH, Khairina SL. Digital discourse of Indonesian netizens on the US–Israel–Iran war on TikTok: a virtual netnography study. Frontiers in Communication. 2026;11. doi: 10.3389/fcomm.2026.1904436..
2. 2 de la Fuente Prieto J, Pérez Herranz C. Comparative analysis of digital practices related to cultural heritage on TikTok and Nostr. Frontiers in Computer Science. 2026;8. doi: 10.3389/fcomp.2026.1748070..
3. 3 Lombana-Bermudez A, García Ramirez D, Perez Avendaño D. Desorden informativo sobre las vacunas contra el COVID-19: narrativas multimodales difundidas en redes sociales por escépticos en Colombia. Contratexto. 2025. doi: 10.26439/contratexto2025.n44.7887..
4. 4 Francisco Junior WE, Albuquerque TCCD, Costa BLDS, Lima Gomes R. Science by means of memes? Meanings of Covid-19 in Brazil based on Instagram posts. Journal of Science Communication. 2023;22(04). doi: 10.22323/2.22040203..
5. 5 Nascimento T, Suarez MC, Campos RD. An integrative review on online ethnography methods: differentiating theoretical bases, potentialities and limitations. Qualitative Market Research: An International Journal. 2022;25(4):492-510. doi: 10.1108/qmr-07-2021-0086..
6. 6 Delli Paoli A, D’Auria V. The Digitalization of Ethnography: A Scoping Review of Methods in Netnography. Journal of Contemporary Ethnography. 2025;54(4):559-587. doi: 10.1177/08912416251342795..
7. 7 Бархатова. Обзор онлайн-этнографических исследований сообществ людей с инвалидностью. The Journal of Social Policy Studies. 2021;19(4):715-726. doi: 10.17323/727-0634-2021-19-4-715-726..
8. 8 Kessler M, De Costa P, Isbell DR, Gajasinghe K. Conducting a Netnography in Second Language Acquisition Research. Language Learning. 2021;71(4):1122-1148. doi: 10.1111/lang.12456..
9. 9 Decataldo A, Andreoni E. Being (a) patient—navigating desire through online communities in assisted reproductive technology. Frontiers in Sociology. 2026;10. doi: 10.3389/fsoc.2025.1683860..
10. 10 Fan S, Kawashima K. Becomings beyond the ideal: women's fitness and digital subjectivities in the Douban women's fitness community. Frontiers in Sports and Active Living. 2026;8. doi: 10.3389/fspor.2026.1786979..
11. 11 Saragih N, Ganiem LM, HI AR, Erlita N, Astuti SW, Umarella FH, et al. Beyond Consensus: Symbolic Convergence and Affective Polarization in Algorithmically Mediated YouTube Discourse on Trump’s 2025 Inauguration. Studies in Media and Communication. 2026;14(2):311. doi: 10.11114/smc.v14i2.7998..
12. 12 Guerrero-Sierra H, Wilches Tinjacá JA, Franco Correa A. Digital narratives and the legitimation of narcotourism: a critical perspective from Latin American thanatourism. Frontiers in Political Science. 2025;7. doi: 10.3389/fpos.2025.1601301..
13. 13 Ita Musarrofa. <b><i>Transforming Islamic Family Law Research Methodology in the Digital Era: A Netnographic Framework </i></b><b><i>within</i></b><b><i> </i></b><b><i></i></b> <b><i>Socio-Legal and Living Law Studies</i></b>. Syakhshiyyah Jurnal Hukum Keluarga Islam. 2026;6(1):201-235. doi: 10.32332/hk5fkp76..
14. 14 Ridgway R, Malevé N. Synthetic Data and Reverse Image Search: Constructing New Surveillant Indexicalities. Surveillance & Society. 2024;22(4). doi: 10.24908/ss.v22i4.18332..
15. 15 Vilchis Díaz R. Historias de vida del sujeto Dreamer desde la etnografía digital, una revisión metodológica. Mediaciones Sociales. 2025;24:e96406. doi: 10.5209/meso.96406..
16. 16 Sigurdardottir H, Rautajoki H. Embodying scenes of moral disorder: Bodily gestures as a site of signification in feminist TikTok activism. Nordic Journal of Media Studies. 2024;6(1):11-35. doi: 10.2478/njms-2024-0002..
17. 17 Yoo J. A Raciosemiotics of Appropriation: Transnational Performance of Raciogender among Mexican K-Pop Fans. Signs and Society. 2023;11(1):68-92. doi: 10.1086/722810..
18. 18 Zhang LT, Vázquez-Calvo B, Cassany D. The emerging phenomenon of L2 vlogging on Bilibili: characteristics, engagement, and informal language learning. El Profesional de la información. 2023. doi: 10.3145/epi.2023.may.01..
19. 19 Gazali, Gazali, Dewi, Yelfi S., Rehayati, Rina, Hadi, Rahmad Tri. Digitalization of the Teachings of the Naqsyabandhi Haqqani Tariqah in Indonesian Urban Society. International Journal of Islamic Thought. 2026;29. doi: 10.24035/ijit.29.2026.352..
20. 20 Khemies I, Ouissi MN, Hamma W. Netnography as a methodological lens for uncovering heritage ambience: A case study of the Tlemcen medina. Urbani izziv. 2025;36(1):148-159. doi: 10.5379/urbani-izziv-en-2025-36-01-05..
21. 21 Harliantara H, Widiarto DS, Ozturk B. Visual accountability and digital populism: the “digital ombudsman” model in hypermedia-based governance. Frontiers in Political Science. 2026;8. doi: 10.3389/fpos.2026.1816617..
22. 22 Hickey-Moody AC, Willcox MG. Feminist fabulation as refusal: Christine Yahya’s @pink_bits illustrating ‘bodies we are told to hide’. Frontiers in Communication. 2025;10. doi: 10.3389/fcomm.2025.1542825..
23. 23 Sola-Morales S, Hernández-Conde M, Arencón-Beltrán S, Sierra Caballero F. Mitos e imaginarios del activismo digital feminista. Análisis de memes de la cibercampaña #FuckGenderRoles. Teknokultura. Revista de Cultura Digital y Movimientos Sociales. 2021;19(1):43-54. doi: 10.5209/tekn.76690..
24. 24 Dwivedi AV. Jarawa as Spectacle 2.0: Tropical Tourism’s Algorithmic “Human Safari” in the Andaman Islands. eTropic: electronic journal of studies in the Tropics. 2026;25(2):65-96. doi: 10.25120/etropic.25.2.2026.4267..
25. 25 Ostrovskaya EA. POSTSECULAR IDENTITY NARRATIVES IN THE MIRROR OF BUDDHIST MEDIA. Ural Historical Journal. 2025;89(4):25-34. doi: 10.30759/1728-9718-2025-4(89)-25-34..
26. 26 Lorca JG, García-Mingo E. Busca, busca, perrita: comunidades digitales misóginas de difusión de imágenes sexuales sin consentimiento. ex aequo - Revista da Associação Portuguesa de Estudos sobre as Mulheres. 2023. doi: 10.22355/exaequo.2023.48.03..
27. 27 AKGÜL M. SİMÜLASYON, GERÇEKLİK VE HİPERGERÇEKLİK KAVRAMLARINDAN İSLAMİ EVLİLİK SİTELERİNİ OKUMAK. Bilimname. 2020;2020(41):413-442. doi: 10.28949/bilimname.673105..
28. 28 Akib M, Fauziyah NN, Hasanah I. DIGITAL AUTHORITY UNDER ALGORITHMIC CONTROL. TAJDID: Jurnal Ilmu Ushuluddin. 2025;24(2):705-738. doi: 10.30631/tjd.v24i2.5245..
29. 29 Nazário ME, Santos WD, Ferreira Neto A. Netnography of curricular reform for Brazilian high school. Revista Brasileira de Educação. 2021;26. doi: 10.1590/s1413-24782021260003..
30. 30 Lacriola M. Un cambiamento mediatico. Un approccio storico a YouTube Italia. Umanistica Digitale. 2026. doi: 10.60923/issn.2532-8816/22846..
31. 31 COTOC A, RADU A. THE IDENTITY REPERTOIRES OF ROMANIAN MICROCELEBRITIES. A DIGITAL ETHNOGRAPHY APPROACH TO TELLING A LIFE STORY ON SOCIAL MEDIA. Studia Universitatis Babeș-Bolyai Philologia. 2024. doi: 10.24193/subbphilo.2024.4.16..
32. 32 Gunasekara H, Brown C, O'Steen B. Promoting teacherpreneurship in creative education through Instagram. ASCILITE Publications. 2023. doi: 10.14742/apubs.2023.532..
33. 33 García-Vega A, De la Fuente Prieto J. La creación de contenidos de ficción juveniles en redes sociales. Revista ICONO 14. Revista científica de Comunicación y Tecnologías emergentes. 2022;20(1). doi: 10.7195/ri14.v20i1.1764..
34. 34 Sancho Guinda C. From isolation to constellation: Narrative focalisation in current technology disclosures. Ibérica. 2022. doi: 10.17398/2340-2784.44.17..
35. 35 Chan GH, Lee GKW, Lo TW. Online Ethnography: A Research Methodology for Reaching Hard-to-Reach Groups in the Contemporary Era. International Journal of Qualitative Methods. 2026;25. doi: 10.1177/16094069261473438..
36. 36 Kuźniarska A, Czakon W, Jedynak M, Mania K, Woszczyna K. Advancing netnography – hybdrid ways of coding social media content. Journal of Organizational Change Management. 2025;38(8):220-233. doi: 10.1108/jocm-06-2025-0491..
37. 37 Banikalef AA, Mehawesh MI. Performing Identity: Repetition, Emoji, and Candidate Indicators of Amplified Visibility in Jordanian–Palestinian Facebook Discourse. Social Media + Society. 2026;12(2). doi: 10.1177/20563051261454774..
38. 38 Rivero E, Jacobo S. Examining the linguistic power of emotes: multimodal race discourse on Twitch. Information, Communication & Society. 2026. doi: 10.1080/1369118x.2026.2648699..
39. 39 Elyamany N, Al Agroudi R, Hasib J, Eladl A. Algospeak as algorithmic negotiation: multimodal strategies, platform governance, and the politics of visibility on TikTok. Language and Semiotic Studies. 2026. doi: 10.1515/lass-2026-0031..
40. 40 Downing J. Methodological insights into digital vernacular (in)security: Developing from netnography to appnography on Snapchat in Marseille, France. Security Dialogue. 2025;56(5):519-535. doi: 10.1177/09670106251362733..
41. 41 Palmberger M. The Digital Diary: A Mobile, Multimodal, and Participatory Method and Part of Digital Ethnography. International Journal of Qualitative Methods. 2025;24. doi: 10.1177/16094069251329262..
42. 42 Pera R, Quinton S, Baima G. The use of embedded visual methods in unveiling consumers' values: A digital ethnography perspective. Psychology & Marketing. 2021;39(2):309-319. doi: 10.1002/mar.21596..
43. 43 Zhang Y, Jiang L. Constructing translanguaging spaces through GenAI-assisted digital multimodal composing: a virtual ethnography on platform language teacherpreneurs. Journal of Multilingual and Multicultural Development. 2026. doi: 10.1080/01434632.2026.2670679..
44. 44 Tartari M, Lavorgna A, Ugwudike P. Share with care: negotiating children’s health and safety in sharenting practices. Media, Culture & Society. 2023;45(7):1453-1470. doi: 10.1177/01634437231182002..
45. 45 Jiang L, Gu MM. Understanding youths’ civic participation online: a digital multimodal composing perspective. Learning, Media and Technology. 2022;47(4):537-556. doi: 10.1080/17439884.2022.2044849..
46. 46 Sigurdardottir H, Imani M, Edalati Z. Asking for solidarity: embodied feminist practices in digital space. Gender & Development. 2024;32(1-2):523-545. doi: 10.1080/13552074.2024.2365058..