First pass coding becomes more efficient with practice, but it is the discourse, feedback, and comparison process of coding with peers that has been beneficial to my understanding of the procedure. I have discovered that my low inference coding has evolved. During my initial attempts at this first phase of coding, keywords were used. These keywords lacked any “descriptive power.” Advice was given in class to accept that low inference codes do not need to be one word; they can be a phrase or even a sentence, they can be longer. When you code in a low inference manner, you translate what is being said without providing interpretation or adding judgment (S. J. Warren, course lecture, 7/21/14). Low inference does not mean “no” inference and simple keywords may lack description.
Three rounds of coding were discussed during our last online class session. The first round of coding is a low inference coding, or first pass coding. During second pass coding, categories are developed and codes are placed into these broad categories. The lead researcher usually guides this process, but the categories are developed as a team with inter-rater agreement to increase consistency and reliability. The third round will be establishing themes and bringing the developed categories into themes. The themes will develop and tell the story of the data.
The similar level of coding output during class this week surprised me. When my small group came back together to receive feedback and discuss the codes developed individually, our low inference codes were comparable. Even though we each coded different pages of a transcript, we each went beyond keywords into a more descriptive, yet succinct coding style. Typically, three to four word phrases. Interestingly, we also discussed the need to categorize certain phrases and codes, the next round of coding. This is progress.