Scholarlink

<p>The proliferation of convergent social-media data, where the audio, visual, textual and other interactive elements from a YouTube deepfake, a TikTok reel, a Facebook post or an Instagram short do not arrive as a single stream but tightly coupled composite of moving image, sound, on-screen and embedded text, hashtags, captions, engagement metrics, and comment threads that talk back to the artifact in real time, produces artifacts that are dynamically integrated with audience commentary and algorithmic metadata, pose a fundamental methodological challenge for qualitative research. While the classical Grounded Theory Method (GTM) has been extended to visual and audiovisual data, it lacks explicit analytic procedures for systematically disaggregating, analyzing and reconverging these layered platform-centric multimodal artifacts. This paper addresses this gap by introducing the Multimodal Grounded Theory Method (MGTM), a formal methodological framework that operationalizes GTM’s inductive logic for the complexities of convergent digital data. MGTM makes three core analytical operations explicit: (1) modal differentiation, where artifacts are analytically segmented into visual, auditory, textual and receptional streams before coding; (2) convergent axial coding, which theorizes the relationships and tensions between codes across modes; and (3) the constitutive treatment of audience reception and platform metadata as primary data for theory generation. Developed through a double-track (Tracks A and B) systematic analysis, beyond eight major visual and audiovisual GTM adaptations, MGTM provides a structured yet flexible workflow, from data capture and multimodal open coding to iterative theoretical sampling. The extensive method is demonstrated first on a single test case, a YouTube deepfake with 258 viewer comments, and then on two further cases, from TikTok and Instagram, with respectively 448 and 737 viewer comments on each (totaling <i>n</i> = 1443 reception reactions overall) to generalize the apparatus. Across the three, the decisive variable demonstrates MGTM as a structured extension framework of doing grounded theory when the data are layered and algorithmically mediated.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multimodal grounded theory method: an extension framework to analyze layered social-media data in the current algorithmic culture

  • 期刊论文

  • 发表日期：

    2026年7月1日

查看全文
  • Shahnaz Bashir

摘要

The proliferation of convergent social-media data, where the audio, visual, textual and other interactive elements from a YouTube deepfake, a TikTok reel, a Facebook post or an Instagram short do not arrive as a single stream but tightly coupled composite of moving image, sound, on-screen and embedded text, hashtags, captions, engagement metrics, and comment threads that talk back to the artifact in real time, produces artifacts that are dynamically integrated with audience commentary and algorithmic metadata, pose a fundamental methodological challenge for qualitative research. While the classical Grounded Theory Method (GTM) has been extended to visual and audiovisual data, it lacks explicit analytic procedures for systematically disaggregating, analyzing and reconverging these layered platform-centric multimodal artifacts. This paper addresses this gap by introducing the Multimodal Grounded Theory Method (MGTM), a formal methodological framework that operationalizes GTM’s inductive logic for the complexities of convergent digital data. MGTM makes three core analytical operations explicit: (1) modal differentiation, where artifacts are analytically segmented into visual, auditory, textual and receptional streams before coding; (2) convergent axial coding, which theorizes the relationships and tensions between codes across modes; and (3) the constitutive treatment of audience reception and platform metadata as primary data for theory generation. Developed through a double-track (Tracks A and B) systematic analysis, beyond eight major visual and audiovisual GTM adaptations, MGTM provides a structured yet flexible workflow, from data capture and multimodal open coding to iterative theoretical sampling. The extensive method is demonstrated first on a single test case, a YouTube deepfake with 258 viewer comments, and then on two further cases, from TikTok and Instagram, with respectively 448 and 737 viewer comments on each (totaling n = 1443 reception reactions overall) to generalize the apparatus. Across the three, the decisive variable demonstrates MGTM as a structured extension framework of doing grounded theory when the data are layered and algorithmically mediated.