According to Charmaz (2006, p. 43), coding is theprocess of “categorising segments of data with a short name that simultaneouslysummarises and accounts for each piece of data. Your codes show how you select,separate and sort data to begin an analytic accounting of them”. I will study thecollected data and begin to separate and sort these data through qualitativecoding using an open sourced software, TAMS Analyser. Charmaz (2006) recommendstwo types of coding: initial coding and focused coding. However, there is nofeature on coding software for these coding methods.Instead, in thesoftware, users “can import documents, field notes, or interview transcriptsand quickly code them by clicking on codes from your list or by creating newtags on the fly.
You can extract coded passages based on various criteria,recode found sets of data, and perform many other useful analytic andmanagement tasks. The program can be used simply to gather relevant text oraudio/video exemplars for use in writing up a project, or it can be used tooutput sophisticated concept maps exploring the relationship between differentnodes in your data” (Warters, 2005, p.321). This would helpcategorise the reviews based on best-fit keywords. These categories may includepositive or negative reviews, factors that customers are satisfied/dissatisfied with, etc. There is also a memo function to add comments toelaborate categories, define the relationship between them, etc. as user codes(Warters, 2005; Ong, 2012).
This allows the researcher to instantly note down newideas and insights and can easily refer to them again in future. Theoretical sampling and theory generation Next, review the data, elaborate and refine thecategories. With the better organisation of data and discovery of relationshipsbetween codes, I would be able to create new concepts, explanations, and providedetailed view of my study. In order to give a clearer view of the findings andresults, diagrams would be used. LimitationsThis study is subjected to a number of limitations.Limitation of using social media data is itssubjectivity. Information shared on social media may not be credible as thereare no metrics to measure emotions. People may exaggerate, suppress or playdown their expression of emotions (Guerrero and Floyd, 2006).
In addition,since all data are selected based on the researcher’s judgement, findings mayresult in some bias and are also often deemed as subjective (Goulding, 2002).Hence this threatens the validity of the research. As online information is constantly updated, data maybe outdatedResearch ethics”It is the responsibility of researchers to protectstudy participants from unintended harm resulting from the research” (Flickeret al.
, 2004, p.128). According to Silverman (2016), researchers conductingInternet-based research need to consider the following ethical issues: perceptions of privacy, informed consent and protection of personallyidentifiable information.
Although people acknowledge that information on theInternet is generally considered public, they maintain strong expectations ofprivacy (Silverman, 2016). For instance, he/ she would feel that his/ herprivacy is violated when their information is used inappropriately. However, itis sometimes impossible to obtain informed consent from people who contributedto this information, due to time constraints and lack of means of contact (Silverman,2016).Thus, this research will comply with ethnic guidelinesfor researching Internet communities developed by Flicker et al. (2004) as seenin Table.Additionally, this research will stay within the limits and comply withNewcastle University’s ethical guidelines to ensure that the student,participants in the project or the university’s reputation are not harmed.