Friday, February 27, 2009

Week 8: Eth-Know-Graphies (EKG)

Sorry for the title, I was just having some fun mispronouncing ethnographies... and a EKG reference... Enjoy!

Blog Question: What distinguishes ethnographies from case studies, how does “triangulation” impact data collection and analysis, and what must ethnographers do to ensure their work is both reliable and valid?

To answer the blog question on the basic level quickly, ethnographies differ from cases studies mostly based on the time of the study, with ethnographies observing over a greater length of time, and observing the multiple facets within an environment. In addition, a hypothesis of also generated within the ethnography. Ethnographies require multiple sources of data and the combination of the data is triangulation. This combination of data is supposed to provide a “rich account” that could not otherwise be obtained via experimentation. Ethnographers have a great number of potential threats to both their reliability and validity. Laura and Asher use Sadler’s list of ten threats ranging from the misinterpretation of cause and effect (No. 10), uneven reliability (No. 6), or pitfalls in confidence in judgment (No .2) (p. 46-7). With the basic blog questions out of the way, I will move onto the specific readings.

Doheny-Farina is through. He utilizes several theoretical assumptions as well as four sources of data; field notes, meeting tape recordings, and two types of interviews that were taken three to five days a week over the course of eight months. Not only did Doheny-Farina use four sources of information captured multiple time a week, but he also brought it back to the theoretical framework. The result of the ethnography was that Doheny-Farina showed that the process was reciprocal and that he described the process. One of the better aspects of this project is that Doheny-Farina did not overstep the limitations of the ethnography, admitted the provisional nature of his model, and noted some limitations to the study. In my personal opinion, this was a good ethnography.

Beaufort, like Doheny-Farina, establishes a theoretical framework, uses a multitude of data sources (interviews, transcripts, written documents), and participates with the research participants over a ‘long’ period, weekly over the course of a year. However, the sample size of n=4 was relatively small and precludes generalization. Beaufort straddles a dangerous line by linking independent and dependent variables and by generalizing which the study may not support.

Sheehy once again uses an eight-month long study of middle school students at SMS. What is interesting for Sheehy’s study is the role the researcher played. Sheehy noted an inconsistent and that her role as a participant/observer changed within the study. This shift may have altered the data and may be a threat to the reliability of the ethnography, but the researcher does not explain any shift in the data. Unfortunately, the effects of the change in the role of the researcher are left unknown. Other than that, Sheehy is also on the edge about generalizing within her conclusion.

To be frank, Ellis is not an ethnography, it is more of a narrative. It does not pull from a theoretical framework that I am aware of as a planner. It does not pull from multiple sources of data, although it is unique. Finally, although it comprises of several months of reflection, the data was not collected (uniformly) over a long enough time. If anything, this is, at best, a good primary source of information for historians in the future and I question the “autoethnography” academic rigor.

Speaking of autoethnographies, Anderson provides a better approach to this style of qualitative research. Overall, Anderson provides a better definition of what an autoethnography entails, mostly because of the researcher being a full participating member of the group/environment that is being studied. However, Anderson provides a historical evolution and theoretical framework for autoethnography, but if we were to look at the whether or not this work is an ethnography, then it is not (duh) . Anderson provides the approach that Ellis lacks.

1 comment:

  1. Hah! I like the "EKG" reference. It works particularly well for evocative autoethnographies.