Manual data collection & qualitative analysis for social media data – “luddite” meme researcher insecurities in the age of AI
Keywords: mental health, memes, researcher insecurities
TL;DR: This paper discusses the insecurities of being a qualitative social media researcher who collects data manually in the age of automation and AI
Abstract: Manual data collection & qualitative analysis for social media data – “luddite” meme researcher insecurities in the age of AI
Digital data and digital research methods have undoubtedly opened a whole new world for researchers. However, they have also raised fears and insecurities about research ethics and the researcher’s role. For some, AI and other innovative digital tools seem too scary to learn or take control of. This paper focuses these personal and professional insecurities about digital, automated research methods as researcher on memes who utilizes manual data collection and qualitative analysis in the age of AI, alongside with presenting the overall PhD project about representations of mental health and mental illness in Internet memes.
The data of this paper consists of approximately 900 manually collected mental health themed memes collected under two social media and Internet platforms: Instagram and Imgur.com. On Instagram, the data is collected manually by taking screenshots of memes under hashtags #mentalhealthmeme and #mentalillnessmemes, as well as three accounts focusing on mental health related themes. On Imgur.com, the data is detected and collected on the Most Viral page, especially under larger image cluster posts called Meme Dumps. The data is collected gradually throughout the research process, with more intensive data collection periods implemented. During the time of writing, the latest more intensive data collection phase was conducted in January 2024. Moreover, reverse snowball method (Särmä, 2014: 99) will be utilized in the data collection throughout the whole research process. The aim of this study is to examine different semiotic ways these memes represent mental health issues by analysing the data with multimodal critical discourse analysis. In this paper I present the insecurities about utilizing digital data collection and analysis tools in meme research, as well as the arguments for not utilizing them in relation to the overall PhD project. Fears of not learning AI or other data scraping tools include being left behinf in digital development, cherry-picking research data, and being labelled as a luddite not willing or not being able to learn new methods. However, arguments for more manual, researcher-oriented methods in collecting social media data, will also be discussed. Overall, the aim of this paper is to seek encouragement from peers, as well as new ideas and support in utilizing digital research methods, and most importantly, how to connect those methods to qualitative and critical research.
References
Särmä, S. (2014). Junk feminism and nuclear wannabes: Collaging parodies of Iran and North Korea. Tampere: Tampere University Press.Cambridge, MA: MIT Press.
Submission Number: 31
Loading