Navigating Manual vs. Automated Coding: A Researcher's Dilemma

Manual Coding 

1. We engaged with the data: Listened to the initial interactions, made our notes 
2. We created a structure of analysis: Created codes map from the DG, notes and transcripts 
3. We identified themes: Grouped codes into themes, revealing underlying patterns. 
4. We iterated: Marked transcripts, and refined codes/themes we dug deeper into the data 
5. We sorted the data: Copied into Excel, created tables, and maybe did some math 
6. We summarized our findings: Various cuts by cohort, some numbers, and illustrative verbatims 

1. Hands-on and in-depth: You develop a connection with the data, and are able to explore the depth of the conversations, rather than just the volume. 
2. Contextual Nuances: By directly engaging with the data, you're better equipped to capture context, emotions, and subtle nuances. 
3. Adaptability: You can adapt your analysis to unexpected discoveries and findings quickly. 

1. Time-Consuming: It demands significant time and effort. 
2. Subjective: Researcher bias can inadvertently influence the coding process 
3. Scalability: For large datasets, manual coding becomes laborious and less efficient. 

Automated coding 

1. MS Office: Word and Excel were our go-to’s. Transcripts were digitized, and tools within these applications allowed us to annotate, tabulate, comment, track changes and the works. Even now, these are still commonly used for exchanging information and analysis at least. 
2. Open ended coding tools like Ascribe, Clarabridge: Though most industry standard tools have now been acquired by large enterprises, these went a more beyond Office. You could build and iterate codeframes more easily, and also bring in the collaboration aspects when many individuals are involved. Over time, they also gained the ability to automatically tag codes and text. 
3. CAQDAS’s like NVivo, Delve, Atlas.ti: Computer assisted qualitative data analysis software, which were able to do more complex activities and help reduce manual effort significantly. They did evolve over time and became more complex. 
4. Machine Learning and AI: This is where NLP, complex algorithms (LLM’s for one), self-learning systems, and hardware came together, at scale, to take on more research and reporting tasks. I’m sure everyone has heard of ChatGPT, DALL-e, Midjourney, which operate at unbelievable scale. Many CAQDAS’s now leverage these is various ways, and it promises a revolution.  

On the plus side – 

1. Speed and Efficiency: Automated coding significantly accelerates the process, ideal for time-sensitive research. 
2. Consistency: Automated methods reduce the potential for human bias, ensuring consistent results. 
3. Scalability: Handling large datasets becomes more manageable with automation. 

The downsides –  

1. Features: Some of these tools have truck load of features, many that make you change your established processes. And often you pay for things that you do not need. 
2. Ease of use: There is almost always a learning curve involved 
3. Flexibility: Once you start using a tool, you are stuck with its algorithms and shortcomings too.  
4. Contextual loss: While they are getting better, nuances and subtleties still pose challenges. Incorrect data can also filter in. 
5. Human intelligence: Machines cannot think like a researcher, at least, not yet.  

Machines also pose some interesting ethical conundrums 

1. Data privacy 
2. Developer bias 
3. Consent and transparency 
4. Environmental impact