Tweet Portraiture: Understanding the Gist of Electronic Identities through Microblogged Messaging in Social Media Using NCapture and NVivo 10

Introduction

Microblogging sites are those that aggregate and distribute short text messages between its users in private, semi-public, and public venues. If people are a sum of their actions (and habituated behaviors and thinking), then can it be said that an online account identity on a microblogging site (like Twitter) is a sum of its expressed ideas? Would 140-character microblogging messages—expressed as fleeting asides, often without the benefit of forethought or critical mind oversight—capture an impulse-based aspect of identity that would not be seeable otherwise? Said another way, does this microblogging medium elicit communication from people that they “are less likely to express using existing technologies” (Grace, Zhao, & boyd, 2010, p. 4517). In other words, does the speed and riposte of this technology erase people’s own personal filters on communications? After all, the sine qua non of microblogging is real-time and interactive short-form communications to capture top-of-mind mental chirps. This communications medium is all about speed, brevity, and dynamic Live Web interactivity among people connecting from anywhere in the electronic environment. Would a corpus of short-message Tweets linked to an entity be able to capture a gist (or the substantive core or essence) of that entity?

This chapter explores the use of NCapture (a web browser extension) and NVivo (a qualitative and mixed methods data analysis tool) to extract the cumulative messages of a Twitter account to see what may be learned from that endeavor. While microblogs are very time-based, particularly for breaking events, this research approach will involve the extraction of cumulative messaging linked to the lifespan of the account (in an approach called “author-wise pooling” in some contexts) without an analysis of the time element except for the beginning and current date of the account (or an end date if relevant). The data are reduced from an n = all of the messaging into word clouds, text-search-based word trees, tree maps, and other data visualizations, for study. Summative content analysis “begins from existing words or phrases in the text itself (the raw data) and counts these; then the researcher extends his or her exploration to include latent meanings and themes that are apparent in the data…” (Berg & Lune, 2012, p. 352). Latent analysis refers to the “interpretive reading of the symbolism underlying the physical data” (Berg & Lune, 2012, p. 355). This approach will not include the analysis of inter-relationships between various accounts—although clearly such interactions may have sparked certain messaging. All these other elements—time, interrelationships, memberships, synchronous events, and linked multimedia (linked photos and videos)—are backgrounded to the focus on the text message extractions.

This approach of conducting a micro-ethnographic study (vs. macro) using a mcroblogging site may be understood as part of online ethnography (Berg & Lune, 2012, pp. 238 – 240), social-cultural research which taps into online worlds, social media, massive online games, wikis, blogs, and discussion boards. In this field research approach, collecting online information and interacting with others online are common research tools in the natural setting of immersive online spaces. These comprise online ethnography fieldwork. Intercommunications to the public are a social performance. Identifying topics of interest in Twitter users’ accounts to learn about the account holders’ interests is not new (Michelson & Mackassy, 2010); microblogging content may be used to profile users “according to substance, style, status, and social tendency” (Tsur & Rappoport, 2012, p. 644). What may be somewhat original is the sequencing and the application of various publicly available technologies for the analysis. This is a micro study because its unit of social analysis is an individual or group (as expressed in an account). To provide a sense of this level of analysis. Figure 1, “A Private Company’s Twitter History Expressed as a Word Cloud” shows a real-world microblogging extraction depicted as a word cloud, with main words expressed by size and centrality.

Figure 1.

A private company’s Twitter history expressed as a word cloud