Theme 6: Qualitative and Case Study Research

Case study is the observation of one or multiple cases over a period of time to understand the dynamic present within specific instances. Cases can be a person, group, event or situation. The observation of phenomena in these cases and the relationship between them is examined in order to test theory or generate theory. When beginning a case study, the definition of a research question is important so that the research has a focus amongst the large amount of empirical data, though not necessary. This question can be redefined as empirical evidence is collected. Case study can use both quantitative and qualitative research methods. These can also be combined in order to strengthen theory. When combined, quantitative data can indicate relationships to the researcher and qualitative data can define the rationale behind these relationships. Multiple data collection methods can also be used to better ground the emerging theory. These methods can also be altered as the study goes on. The ability to make adjustments is a key part of theory building in case study. Case study is also well suited to areas of research where existing theory is not well developed or in new areas of research.

A case study I chose to analyze was titled Emerging opinion leaders in crowd unfollow crisis a case study of mobile brands in Twitter by Liyang Tang and Zhiwei Ni. The initial research question sought to explain crowd unfollow in companies’ social media accounts and whether emerging opinion leaders influence these crowd unfollow events. Perhaps these opinion leaders were the cause when an unusually large number of unfollows occur. Twitter was chosen as a typical representative of social networks as it is widely used by different groups of people to express opinions and is effective as evidence in understanding peoples’ behaviors. Then three mobile brands’ Twitter accounts were chosen to be analyzed. This was Samsung, Nokia, and BlackBerry. Multiple data collection steps were chosen. First software and algorithms were used for “burst detection” or detection of a crowd unfollow event. It was found that were three types of unfollow event: one with long-lasting duration (Nokia had one that lasted 10 days), one with short-lived duration (Samsung had one that lasted less than a week), and the no evident crowd unfollow (BlackBerry). Then different monitoring units were used to detect relationships among ex-followers. There existed some users with significant importance in only one case. Nokia appeared to have significant opinion leaders that influenced an unfollow, while Samsung unfollowers seemed to have little influence on each other. It was hypothesized that Samsung’s unfollow event might have resulted from bad rumors following a product release.

In evaluating this case study and comparing it to the "Process of Building Theory from Case Study Research" in Eisenhardt’s article, I found many weaknesses in the case study. First, while Twitter was chosen for significant reasons, there were no reasons listed for choosing the three mobile companies as cases. Eisenhardt suggests focusing on theoretical useful cases. Second, there were multiple data collection methods, but they focused on collecting the same data. I believe the case study could have benefited from qualitative data and analysis of the Twitter accounts of the opinion leaders. This could have benefited in determining why they held significant influence. Third, the hypothesis seemed to be supported by only one case. Cross-case patterns are important in developing theory. Eisenhardt ascertained that emerging relationships must fit with the evidence in each case.

I also analyzed Democracy in disguise: the use of social media in reviewing the Icelandic Constitution by Bjarki Valtysson in order to better understand qualitative research. Valtysson used qualitative content analysis to scrutinize how social media was used as a participative process in the collaborative rewriting of the Icelandic Constitution. Valtysson define qualitative content analysis as “research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns” (Valtysson, 53). Valtysson analyzed content to find themes in how the Constitutional Council received and treated the communication and discussion that were framed in social media. The content examined was the use of Facebook, YouTube, Flickr, Twitter in engaging citizens to participate.  Valtysson did not elaborate much on her methods of the content analysis. Valtysson went into more detail describing how the affordances of YouTube, Facebook, Flickr and Twitter vary. This appears and it might be a limitation of qualitative research. It is more difficult to describe one’s observations as compared to the more deliberate steps in quantitative research. She did notice a lack of responsiveness and discussion, but it seemed difficult to explain all the observation that led to this conclusion. Valtysson used figures from Statistics Iceland, which helped to support her conclusions, but I believe this made use of quantitative methods. I think that in order to improve the article, more evidence could have been described in the detailing of exact quotes and conversations that support the observations.

Sources:

Emerging opinion leaders in crowd unfollow crisis a case study of mobile brands in Twitter, Liyang Tang and Zhiwei Ni

Building Theories from Case Study Research, Kathleen M. Eisenhardt

Democracy in disguise: the use of social media in reviewing the Icelandic Constitution, Bjarki Valtysson

Theme 3: Comments

https://u10o7oqf.blogspot.se/2016/09/theme-3-part-2-reflections.html?showComment=1477682849495#c9195348587218661666

https://dm2572-16.blogspot.se/2016/09/theme-32.html?showComment=1477688417899

https://omg-dm2572.blogspot.se/2016/09/after-theme-3.html?showComment=1477690531546#c234956807848717664

Theme 4: Reflection

What I found most interesting about the lecture on Drumming in Immersive Virtual Reality was the ability to formulate the measure of expressivity using quantitative data. At first introduction of the study’s intent, I would have assumed only qualitative data would be used. Perhaps there would be an extensive interview in which the participants would analyze how they felt in each avatar, whether they were the Caucasian drum player (FL) or the dark-skinned drum player (CD). Then they could express if they felt they could play more confidently with more expressivity or if they felt the opposite. Yet, there was a substantial amount of data analysis involved in determining the movement of each participant in their baseline condition and then virtual body. This study illustrated the amount of control and structure that must be present when using qualitative data. When you are using quantitative methods, statistical analysis can only build purely on the quantitative data. I believe this is why the study used both the mapping of movement using Opitrack, the use of questionnaire to evaluate the body ownership illusion, a short interview, and yet another questionnaire to determine demographics and personality traits.

This lecture also demonstrated the importance of establishing a baseline condition. The difference in expressivity when in the illusion of the CD body, was determined by comparison to the baseline group. This could be compared to the FL body that did not differ significantly from the baseline group. If these were used without the baseline white hands, the difference could not be explained as statistically significant. This would have just confirmed the hypothesis without empirical evidence as support.

I appreciated the study of questionnaires during the seminar as this was the quantitative research method of choice in the article I chose. There were many advantages and disadvantages that I had not considered. For one, the article used a posted questionnare. A benefit of this method we discussed is that a physical copy sits around as a reminder to complete it. This is opposed to online questionnaires that are easily dismissed with the click of a mouse. A disadvantage of mailing physical copies is that the data is not immediately registered in a database for analyzation. When completed online, the results are instantaneous.

I also found it interesting that this article chose the posted method in place of the online method as it was an article examining different types of Internet users. The author did not state her reason for doing so, but perhaps the use of an online questionnaire could have benefited her effectiveness in research. It would have eliminated the irrelevant participants that were not frequent Internet users and given her a relevant study sample immediately.