Step 1 – temporality of data |
- Online data provide traceable, searchable and easily accessible retrospective data - Online grounded theorists can resort on the temporality of data to make easier the process of constant comparison and theoretical sampling (the field is always open for the researcher to come back) - Real-time events in which the grounded theorists can have access to organizations and audiences’ instant reactions are abundant - Online data is a rich source of synchronous interactional data that would be stricter in traditional offline settings - Online settings provide opportunities to conduct longitudinal research in a short period of time |
- It is difficult to define the boundaries between real-time (synchronous) and retrospective (asynchronous) data - If the grounded theorist decides to work with synchronous data, the immersion in the field can become time-consuming or even exhausting because of the full-time availability and access to data - The sources of asynchronous or retrospective data are so abundant that they can become overwhelming for the researcher to deal within the first place |
- Massa (2017)Massa, F.G. (2017). “Guardians of the internet: building and sustaining the anonymous online community”. Organization Studies, 38, 959–988, 10.1177/0170840616670436 https://doi.org/10.1177/0170840616670436...
relies on retrospective (asynchronous) online data (previous computer-mediated communications such as forum postings, chat logs, memes, images and videos) to trace longitudinally the emergence of an online community (Anonymous) from 2008 to 2011 - Roberts and Zietsma (2018)Roberts, A., & Zietsma, C. (2018). “Working for an app: organizational boundaries, roles, and meaning of work in the “on-demand” economy”. In L. Ringel, P. Hiller, & C. Zietsma, (Eds), Toward permeable boundaries of organizations? (research in the sociology of organizations) (Vol. 57, pp. 195–225). Bigley, UK, Emerald. conducted an online ethnography with real-time (synchronous) online data (forum and social media interactions between app drivers) to theorize on how workers of online applications (e.g. Uber drivers) understand their roles in boundary defining activities of on-demand organizations |
Step 2 – platform identification |
- The pool of platforms available for online data collection is astonishing, varying from blog posts to podcasts, streaming, forums and online communication tools (see for a comprehensive list) - The combination of multiple platforms provides opportunities to adopt a multimodal approach to research (gathering textual and visual data simultaneously) - There is likely an adequate platform for collecting data for a wide range of research questions concerning organizational phenomena - Different platforms can provide access to real-time data without geographical or budgetary constraints (e.g. long-distance interviewing) |
- It is not easy to access the quality criteria of online platforms (relevance, accessibility, credibility and activeness) - Although online data tends to be traceable and searchable, some platforms have a limited timeframe to store their events (e.g. social media platforms such as Snapchat, Instagram or Facebook have “stories” that disappear 24 h after being posted) - Grounded theorists must avoid the trap of collecting data in platforms that are not suitable to answer their research questions (e.g. if social media data is not helpful, gathering data in such platforms will consume time and effort that could be invested in collecting data in other suitable platforms) |
- Vaast and Levina (2015)Vaast, E., & Levina, N. (2015). “Speaking as one, but not speaking up: Dealing with new moral taint in an occupational online community”. Information and Organization, 25, 73–98, 10.1016/j.infoandorg.2015.02.001 https://doi.org/10.1016/j.infoandorg.201...
show how grounded theorists can effectively identify if a platform attains the four quality criteria, given that the BankingOC online community was relevant for addressing their research question, it was accessible, it was credible for the banking community (over 23 thousand registered members) and it had an active and stable membership - Barberá-Tomás et al. (2019)Barberá-Tomás, D., Castellò, I., de Bakker, F. G. A., & Zietsma, C. (2019). “Energizing through visuals: how social entrepreneurs use emotion-symbolic work for social change”. Academy of Management Journal, 62, 1789–1817, 10.5465/amj.2017.1488 https://doi.org/10.5465/amj.2017.1488...
use social media data (Facebook, Twitter, Instagram, YouTube, etc.) because it can convey traces of emotional contagion and manipulation that are necessary to bring about social change in the case of commotion for fighting plastics pollution - Younger and Fisher (2020)Younger, S., & Fisher, G. (2020). “The exemplar enigma: New venture image formation in an emergent organizational category”. Journal of Business Venturing, 35, 10.1016/j.jbusvent.2018.09.002 https://doi.org/10.1016/j.jbusvent.2018....
combine blog posts, website texts and podcasts to get key organizational actors’ accounts (founders and investors) regarding organizational image formation of new ventures of an emerging organizational category (venture accelerators) |
Step 3 – data access and collection |
- Grounded theorists have at their disposal several tools that facilitate the access and the collection of online data (e.g. CAQDAS) - Researchers have at their disposal almost unlimited access to online kind of data, considering that much of the data available in the new online era are stored and can be accessed through public-domain servers and platforms - Access to research informants can become less challenging because of the fact that several online platforms (e.g. LinkedIn) allow contacting organizational actors without the burden of getting through gatekeepers like in offline traditional data collection procedures |
- Collecting online data may demand extra efforts from the researcher to familiarize with jargons that can be particular to each online community or even to understanding the codes to locate relevant data in each platform (mentions, hashtags, etc.) - Collecting observational data in online settings does not exempt the researchers of building high levels of trust within the community where he/she is engaging with. One might have in mind that those involved in interactional data collection should be aware that their interactions are being observed - It is difficult to assess whether people are being honest or not when collecting interactional data through online sources, what makes difficult for the researcher to assess whether he/she is dealing with relevant data or with noise |
- Massa (2017)Massa, F.G. (2017). “Guardians of the internet: building and sustaining the anonymous online community”. Organization Studies, 38, 959–988, 10.1177/0170840616670436 https://doi.org/10.1177/0170840616670436...
shows that accessing and collecting online data can also be time-consuming. The author spent 10 h per week online following Anonymous forum threads for a 38-mo period - Barberá-Tomás et al. (2019)Barberá-Tomás, D., Castellò, I., de Bakker, F. G. A., & Zietsma, C. (2019). “Energizing through visuals: how social entrepreneurs use emotion-symbolic work for social change”. Academy of Management Journal, 62, 1789–1817, 10.5465/amj.2017.1488 https://doi.org/10.5465/amj.2017.1488...
used as a tool for data collection a software package that fetches and prepares data for analysis (Social Data Analytics Tools, http://cssl.cbs.dk/software/sodato). Data collection was complemented by “hashtag” searches and monitoring and platforms such as Twitter and Instagram and by manual data retrieving (Facebook and YouTube comments) - Kataria et al. (2018)Kataria, N., Kreiner, G., Hollensbe, E., Sheep, M.L., & Stambaugh, J. (2018). “The catalytic role of emotions in sensemaking: Evidence from the blogosphere”. Australian Journal of Management, 43, 456–475, 10.1177/0312896217734589 https://doi.org/10.1177/0312896217734589...
show that even in open online spaces, there is a need to get informed consent from research participants. The authors requested permission and obtained permission from blog owners to read and use their blog posts as online data for their research |
Step 4 – data treatment and storage |
- There is a wide range of options available to the researcher to store and protect data on encrypted servers - Most CAQDAS provide the opportunity of treating data to preserve nuanced meaning and emotions in virtual written communication, such as emoticons, emojis, acronyms and abbreviations |
- Non-efficient treatment of data can be harmful in terms of losing deepness or thickness of data - Anonymizing and protecting data can be a challenging task in the new online era considering that simple online searches can reveal the identity of informants - The use of slangs and profanity language is not unusual. Thus, the grounded theorist must decide if he/she is cleaning them out knowing that it may lead to losing some deepness of data - Data must be stored in highly secured servers and protected by strong cryptographic systems, which can lead to additional costs for the research team |
- Massa (2017)Massa, F.G. (2017). “Guardians of the internet: building and sustaining the anonymous online community”. Organization Studies, 38, 959–988, 10.1177/0170840616670436 https://doi.org/10.1177/0170840616670436...
performed data treatment to avoid false positives in his research. For example, when searching the “Anonymous” in several online data sources, the author got 1,683 articles, but only 178 were not false positives - Roberts and Zietsma (2018)Roberts, A., & Zietsma, C. (2018). “Working for an app: organizational boundaries, roles, and meaning of work in the “on-demand” economy”. In L. Ringel, P. Hiller, & C. Zietsma, (Eds), Toward permeable boundaries of organizations? (research in the sociology of organizations) (Vol. 57, pp. 195–225). Bigley, UK, Emerald. were zealous to protect their online data sources to the point of anonymizing both platforms adopted by them in the research by using pseudonyms (i.e. social media and standalone). Moreover, the authors had also decided to not clean up all the profanity language used by the Uber drivers for the sake of preserving the thickness of data |
Step 5 – data analysis |
- Incorporate elements that may denote emotions and meaning, for example, sarcasm, anger or joy, such as the use of capitalization of words (all caps) - The openness of the field provides the opportunity to adhere to the systematic tenets of grounded theorizing (e.g. theoretical sampling and constant comparison) - The new online era increases the chances of getting to unexpected pathways during the iterative process of data collection and analysis, providing novel and creative insights to the theory emerging from data and theoretical sensitivity |
- The large volume of online data available is a double-edged sword. Although the process of getting lost in the data before getting found is commonplace in offline grounded theory, the risk gets amplified when dealing with data from the new online era - The history behind data is neither logically constructed nor the data is directed by the researcher as it would be in traditional offline ground theory (for example, it can become challenging to analyze and make sense of small chunks of unorderly data collected in platforms such as Twitter, Instagram or Snapchat) - As the field is always open, it is more difficult for the researcher to know when to stop data collection analysis (getting out of the field) |
- Massa (2017)Massa, F.G. (2017). “Guardians of the internet: building and sustaining the anonymous online community”. Organization Studies, 38, 959–988, 10.1177/0170840616670436 https://doi.org/10.1177/0170840616670436...
tackles the lack of order of online data by creating what he calls time codes. This analytical procedure was useful for the author to put data into chronological order, making possible to establish distinct phases in which events took place - Barberá-Tomás et al. (2019)Barberá-Tomás, D., Castellò, I., de Bakker, F. G. A., & Zietsma, C. (2019). “Energizing through visuals: how social entrepreneurs use emotion-symbolic work for social change”. Academy of Management Journal, 62, 1789–1817, 10.5465/amj.2017.1488 https://doi.org/10.5465/amj.2017.1488...
take advantage of the elements denoting emotions during their data analysis by coding and storing emojis, emoticons and capitalized words in social media comments - Younger and Fisher (2020)Younger, S., & Fisher, G. (2020). “The exemplar enigma: New venture image formation in an emergent organizational category”. Journal of Business Venturing, 35, 10.1016/j.jbusvent.2018.09.002 https://doi.org/10.1016/j.jbusvent.2018....
adopt constant comparison techniques both in first- and second-order coding. Roberts and Zietsma (2018)Roberts, A., & Zietsma, C. (2018). “Working for an app: organizational boundaries, roles, and meaning of work in the “on-demand” economy”. In L. Ringel, P. Hiller, & C. Zietsma, (Eds), Toward permeable boundaries of organizations? (research in the sociology of organizations) (Vol. 57, pp. 195–225). Bigley, UK, Emerald. provide an exemplar of the use of constant comparison and theoretical sampling to decide whether new data collection is needed to confront the emerging theory grounded upon online data |