Monday, 10 October 2016

Research Methods on Social Media

When it comes to using social media as a research tool the consensus appears to be that a broad range of methods will give the best results. As Alice E. Marwick Twitter and Society (p110:2013) remarks "Simply collecting a great deal of data may not be adequate for describing use beyond simple queries". A large amount of data does have its merits, for example it can identify the popularity of a particular subject or hash tag.

An example of this can be seen through the recent Presidential debates in America, the second of which was the most tweeted ever with 17 million tweets (I think I contributed at least 100 to this number). But what this data does not identify is anything about the intentions of those tweeting and what was motivating them to tweet what they were tweeting, such as my unapologetic need to troll Donald Trump supporters.

With this in mind more qualitative methods are essential when hoping to conduct research with more weight and findings. One example is interviews, a method which I will be looking to use during my research project, when I intend to look into LGBT youth and "coming out" on Twitter, particularly if the increased prominence of social media has made young lesbian, gay, bisexual and trans-gender youth feel more comfortable or able to come out.

I think interviewing will be essential as this topic could generate more complex responses which can't be covered in a 140 character tweet. There may also be some things which require anonymity, which again unless the individual does not use their full identity on social media, is not possible if data was only being sourced from what is posted publicly on a users Twitter profile. There are also ethical considerations to take into account, which I will discuss in a separate blog post.

I think it would also be a mistake to focus on one social network or social context as the sole method of research unless the research is to be specific to Twitter for example. As danah boyd and Kate Crawford Critical Questions for Big Data (p669 2012) note "Twitter does not represent all people, and it is an error to assume 'people' and 'Twitter users' are synonymous", again pointing out that a more varied approach to research on social media is the best way to achieve more quality when it comes to research findings.

Links:

Second Presidential Debate: the Most Tweeted

Twitter & Society (2013)

Critical Questions for Big Data (2012)

No comments:

Post a Comment