Computer-Mediated Anthropology
|
|
An Online Resource Center |
|
CMA Methodology: Online Surveys by Noah Porter, 2004 |
|
Bird wrote: "as any anthropologist knows, ethnography has also long encompassed a range of methods that supplement or even replace classic fieldwork, and that may perfectly legitimately be used to study media reception - especially if combined with a broader analysis of cultural context. 'Ethnography is not, in and of itself, a way to gather information'...Texts for student anthropologists outline many methods and approaches, including life histories, autobiographies, personal narratives, self?descriptions, diaries, interviews, and (dare I say it!) surveys and other quantitative methods" (2003: 8-9). While I heartily endorse the validity of multi-sited ethnography (Marcus 1995), it was not possible to use this method to study how anthropologists around the country (and the world) have engaged computing, so a survey seemed the appropriate method to use. How the CMA Survey and online surveys in general are (or, at least, can be) done: 1. Decide on a relevant area of study that a survey could address. My readings and experience indicated to me that anthropologists have had a troubled relationship with technology. Are my experiences representative of the "big picture"? A survey seemed appropriate to answer this question since there are a lot of anthropologists out there, and it seemed like it would be easy to determine one's attitude towards computers with only a few short questions. 2. Get a list of e-mail addresses to use. As Kaye and Johnson (1999) note, Internet surveys "pose a unique set of problems in guaranteeing a random sample of respondents. Unlike telephone and mail surveys in which samples can be produced through census lists random digit dialing, the Web has no central registry of users. In addition, e-mail addresses are so varied that they are virtually impossible to construct randomly, making Internet sampling very complex" (p. 325). Thanks to spammers, virus writers, and other people who prey on the virtual public, this is not as easy as it used to be. People are reluctant to give out their e-mail addresses on the web and newsgroups now because of spiders. In my case, however, I was able to use a central registry that was relevant to the population I was studying. I harvested a list of over 400 e-mail addresses from the AAA Guide. It was my hope that there would be enough of a sense of comradery in this discipline (not to mention enough people with valid e-mail addresses listed who were around to check their mail) for this survey to work. Fortunately, I got enough replies to make it work. 3. Decide on the content of the e-mail. This is a well-explored methodological issue, and does not need to be rehashed here in its' entirety. I would just like to make two suggestions beyond the usual ways of making surveys meaningful and relevant to participants: First, make sure that the subject line of the e-mail is actually descriptive. Spammers have become quite good at coming up with generic message titles to encourage people to open them (e.g. "Your response requested", "Important information!", etc). Make sure your e-mail title is descriptive enough that it will not sound like spam. The title I used was: "Computer-Mediated Anthropology survey". I doubt "anthropology" is a word that is well-known enough to be of use to spammers, despite the best intention of public anthropologists to publicize the discipline. Second, why not include something about yourself or the survey? I included my automatic signature with a link to my CV at the end of the survey, just in case anyone was curious who was sending. 4. Send it out. While this may seem like an obvious and easy step, there are some important decisions to be made in how this is done. First, if you have multiple e-mail accounts, which one do you use? I would recommend using a university account over a commercial one; a message from harvard.edu is more likely to be taken seriously than one from aol.com. Also, you want to make sure that your account has enough free space in it to handle all the replies you will get. In addition, you should consider whether you will send each survey individually or as part of a mass-mailing. Do you want the people you contact to feel like the message you sent is personalized, or do you want them to be reminded that they are just one of many that you are contacting? Keep in mind also that the former is more time-consuming than the latter. For the CMA survey, I compromised by sending out e-mails in groups of 10. That way, each person is made aware that I was not specifically targeting them on an individual level, but they are not confronted by an imposing list of hundreds of addresses, which could make their participation seem unnecessary. Posting a survey on the web is an alternative to e-mailing surveys, but which poses some different problems for researchers to address. For a discussion of these issues, see Kaye and Johnson (1999). 5. Collect and analyze data. If you've done surveys before or just taken a good methods course, you know how to do this part. You may also wish to consider some qualitative analysis software packages to aid you in this task. As with other surveys, keep in mind that "[v]oluntary participants might be more interested, informed, and concerned about the survey topic, and typically hold viewpoints that are stronger and more extreme than those of other individuals" (Kaye and Johnson 1999: 326). References Cited Bird, S. Elizabeth Kaye, Barbara K., and Thomas J. Johnson Marcus, George E. |
