Month: March 2015

The reflection site I chose is AllDeaf.com. This is an online forum for Deaf and Hard of hearing population. The qualitative content analysis for the forum is about the information flow within discussion on Cochlear Implant. The disciplinary context for this part is about audiology, CI and Deaf Culture. The social context for this part is about dispute between medical and social model of deafness: should deafness be cured? The conceptual context for that part is about CI usage and acceptance. There are 4273 posts from 2003 to 2015. The themes can be derived from all posts using content analysis methods.

At first glance, it does not seem as though conducting a content analysis seems very daunting. On the other hand, after getting some class time to actually perform a small content analysis, I found that the process is much more involved than the readings made it seem.

I chose an article from the Tallahassee Democrat about the proposed closure of Apalachicola Bay to oyster harvesting. I started by examining the framework the author used, trying to determine if the story was framed using an environmental perspective, an economic perspective, a social perspective, etc. After I determined which framework was used to write the article, I started pulling statements either written by the author or from interviews with users of the Bay to support the framework and began to write research questions from there.

A proper content analysis requires a lot of time and re-evaluation of prior thoughts about content. I wasn’t sure if I was doing this activity the right way when I started, but I think I’ve found a way to help organize and orient myself to conducting a content analysis. This method of qualitative analysis will certainly prove useful to me as I continue in my program.

 

Starting into this weeks readings, I was under the impression that content analysis had some similarities with open coding. One just simply defines the research prior to consulting the text. Of course I fell into a trap of confusing simple with easy. Although Krippendorff’s writing style is fairly clear, I miss the more inviting prose style of Charmaz in the prior week’s readings. It is possible that my own state of mind was the culprit in my confusion on how to perform content analysis. On a whim I thought it would be interesting to compare text from two different mediums (hard-bound book, and blogs) written by the same author on the same topic. Both have a reflective quality while discuss an approach to performing woodworking tasks and the underlying personal philosophy of the craft. The blog tends to be more immediate and has an added element of readers responses. So now I have my text squared away, it is time to carve out some research questions. I immediately forget this interesting concept and start creating some good yet dull questions. This set me down a path to ignore the comparison of mediums that open coding would let me (re)discover.  For some reason I found the analyst’s context step confusing. If I understand this step correctly, then this step requires the researcher to understand how the text fits into the greater conversation as well as state the researcher’s understanding of the topic. Since I have read many woodworking books (here I am referring to books that are more involved than simple Ikea style construction plans), I had plenty to write about. The following steps, analytical constructs and inferences seem very “sciency” with its testable mini-theories and computer-aided dictionary of tags. This appears to be a precarious and slippery slope that one will start counting and turn the task into a quantitative project. However, it does seem to aid efficiency when compared to the “fishing expeditions” of open coding. The validation of evidence appears to be a daunting yet necessary task. This step keeps the findings in-line with the original questions thus keeping the project consistent. This provides confidence against critics who want to argue about the findings. This exercise has given me an appreciation for individuals that use this method. I feel that it is too tempting to create a structure that counts things especially if all of the text is already in a computer-readable format.

As I discovered in class Thursday, content analysis is both much more complex than it appears and much simpler.  For the activity, I looked at whenicameout.tumblr.com, a site set up so that anyone can submit their coming out story, anonymously or not, and have it posted.  I was thoroughly confused while actually doing the activity, convinced that I was not identifying any of the pieces that content analysis requires.  However, when talking about my exercise out loud, I realized I actually had all of the pieces, some more than I needed.

I think content analysis will come in quite handy when conducting my own research in the future.  I’m interested in LGBTQ literature and how it is presented in reviews, online, at libraries and schools, etc.  This method will be the perfect tool to take those apart.  In addition, my current project deals with the coming out story as information, and this method is also seemingly tailor-made for such a project.

While I realize open coding is an entirely different process, I feel that content analysis might be an excellent complement to it.  The analysis here focuses on context and can be used to spotlight influences perhaps not apparent in the coding process.

This week’s discourse and content analysis readings emphasized the “objective, systematic and quantitative description of the manifest content of communication” (Frohmann, 1994, p. 19).

For my class activity, I chose to analyze the content found in the Rants & Rave’s board of the Tallahassee Craigslist site. My research question was “What topics are most discussed on the Rants and Rave’s board of the Tallahassee Craigslist site?” There were several open letters to colleagues and bosses; manifestos pertaining to the government, religion and race relations; complaints about bad customer services or lousy restaurant patrons and so forth. I noticed quite a bit of profanity, exclamation points, ALL CAPS, ellipses, and colloquial (slang) terms. The discussion board appears to be a confessional.

My big take-away: before the analysis of documents can take place, there must be lots of pre-thinking. “Every content analysis requires a context within which the available texts are examined. The analyst must, in effect, construct a world in which the texts make sense and can answer the analysts’ research questions” (Kippendorff, 2004, p. 24). So from a disciplinary standpoint, I chose to approach the Rants & Raves board study from the Information Science domain and along the “information grounds” theoretical lines. I wasn’t sure if I did what was asked of me in terms of the class activity. Still, I think the activity was successful in driving home the point that content analysis and discourse analysis are anything but simple “soft” methodologies. On the contrary, without property scaffolding, the work of analysis will be haphazard, which jeopardizes validity and reliability. Empirical research, especially those which involve hypertextual evidence, necessitates clear controls,  strict definitions and repeatable procedures.

This week’s in-class exercise was thought-provoking, and much harder than it sounded on the surface.  I remember discussing content analysis in Research Methods almost two years ago and thinking, “Oh, that sounds like easy research.”  Ha!  Contextualizing and conceptualizing content analysis research is far from easy, even when you do so using your very specific research interests.  As I may be doing some content analysis in my own dissertation case study, this was a good opportunity for me to see what it’s really like.

As with coding, context is very important in content analysis.  I think it is especially important that the researcher have a firm grasp of the phenomenon under consideration before proceeding with analysis.  In other words, I don’t think you should just wake up one day and think, “Hey, I believe I’ll analyze some blogs about x even though I have absolutely no background in studying or understanding x.”  I believe, after completing the exercise yesterday, that a firm grounding in the history of the phenomenon is necessary to be able to frame your research.  If I had not been a school librarian, I don’t think I would have been able to do a very good job of framing my proposed content analysis yesterday, unless I had taken the time to read up on the topic thoroughly beforehand.  Of course, this is good advice no matter what methodology you choose.  Otherwise, you will just look foolish, right?

I feel like content analysis and open coding overlap in many ways, and it will be interesting to see which approach I choose to take in my dissertation research.  I feel like I need to do some more comparing and contrasting before I make a final decision, and perhaps I will end up using a hybrid of the two.  Right now, I lean toward open coding because of its flexibility, and because I know that I have the information worlds codebook to guide me; but, as always, I remain open to the possibilities.

Grounded theory

I was really struck by the copious attention to semantics and linguistics as well as the patience which entail the grounding theory process. Dr. Rodriguez-Mori showed that numerous iterations of coding, axial coding and then even more coding are necessary for a solid theory to bloom. This methodology 1. calls for serious information organization 2. requires thorough evaluation rubrics 3. is such that the data collection is yoked with the analysis procedure (Corbin & Strauss, 1990). The readings, the guest lecturer as well as Dr. K emphasized that

• one must repeat both the data collection as well as coding processes until the information reaches a point of saturation
• the need to consult theory for the sampling method (for instance, Dr. Mori used the information grounds theory to recruit Puerto Rican participants at churches, stores and other gathering places)
• process is a part of the theory in that “meaning has to be broken down into stages” (Corbin & Strauss, 1990, p. 241)
• and, finally, that patterns and variations must be accounted for. I very much appreciated the buckets analogy!

Coding

My literature background plus tendency to over-analyze affected my ability to code. I was distracted by what I interpreted as awkward interpersonal skills and interesting banter. I did, however, manage to find instances where the authors (especially P07) capitalized things which shouldn’t have been, such as Specialist Degree and I believe Cap & Gown and Graduation. This capitalization demonstrated that these were very important constructs. Furthermore, the use of emoticons and slang (“kewl”) indicated that the forum is an informal, if not, safe place. Finally, the students seemed to use very encouraging and positive language throughout the forum, which highlights the fact that it was used for peer supported.

I have been practicing coding throughout my course project but the class activity was another animal. The data was  jumbled and, at times, unrelated. Still, it was a great learning experience. Coding requires that researchers “empty themselves” and approach words (which are loaded with meaning) in a literal and mechanical way. Try as I might, I couldn’t be as objective as I would’ve hoped. I was sharpened by my classmate’s perspectives.

The grounded theory and coding practice in this Class is fun. After reading through all the grounded theory guidelines, I felt that coding is a systematic procedure which requires a lot of time and close reflections. And different person has different ideas when they tried to code the same materials in that 10 minutes.

For me, when I started reading through the material. I first thought about what research questions I held to analyze the data. To understand how information was posted for online courses? Second, I went through and found only one user were present, which means a lot of contextual information might be missing. Then I was thinking how I could analyze the data. From topics covered? Emotional change? Time of posts? How topics are related to the course? However, I didn’t notice about the funny or weird pieces as commented by my classmates. And it makes me to think about the reliability and validity issues in grounded theory coding. We all have our barriers so the coding results cannot get rid of its subjectivity. Another concern for me is whether personal judgment or comment are appropriate as the codes in analyzing the data. Shouldn’t us keep the objective perspective towards our data and keep ourselves as neutral?

Like many of my classmates, before this activity I had always heard of coding qualitative data, but I never really had an idea of what was involved or how to actually do it. The only coding experience I’ve had was simply reading statements and assigning the statement to a pre-determined category.  I knew coding from scratch is a much more involved process, but I was unsure of where to even start.

At the start of this activity I quickly became a fan of the memo writing process. Writing down my initial thoughts, no matter how mundane or unimportant they seemed at the time, helped me to begin to form a larger picture from the tiny amount of data we examined.  Organization skills are extremely important in the coding process, which I feel will allow me to find enjoyment in the process as I code in the future.  I enjoy putting the pieces of a puzzle together, especially when those pieces lead to a larger narrative.

I know we have to stick to the time constraints imposed on the class, but I wish we could have had more practice with coding. I feel like I was only starting to get the hang of it right as the activity ended, and I would have appreciated more time spent to help get in more of a routine or groove with the data.

Like most other things related to the eventual dissertation I’ll have to write, “coding” was a vague, fairly upsetting concept I kind of didn’t want to talk about.  After the coding exercise in class, I realized I’ve already done it in some form before this.  As an undergraduate, I completed an honors thesis, and I actually did a fair amount of coding for that project.  None of it was as structured as the coding discussed in class; it was mainly through color-coded notations in margins of books and print outs; and I certainly didn’t call it coding.  However, it served the same basic function:  condensing and organizing ideas and concepts into larger groups that can be better analyzed.  Having realized this, I am less worried about the coding in my dissertating future.

By nature, I’m a hyper-organized person who makes massive amounts of notes on almost everything–including leisure reading–so I think coding will end up being an enjoyable process for me.  Of course, it is always possible it will end up driving me out of my skull, but at least I will begin the process expecting to enjoy it.  This type of coding appeals to me in that it bases conclusions on the data and actual situation rather than allowing for preconceived notions to take over the analysis.

I could certainly see the need for another coder to check one’s notes though.  My partner in the coding exercise caught a lot of the same things I did but also managed to note multiple items I either didn’t note or didn’t notice.