Microblogging Science: Methodology
3.1 Introduction
The purpose of this study is to examine why scientists use a micro-blogging to communicate their research. For the purposes of the study it was decided to limit the research to a curated account on the microblogging platform Twitter. This decision was made for several reasons. Twitter is one of the most active and popular microblogging platforms currently in operation. According to Twitter there are currently 271 million active users with 500 million tweets sent per day.
Due to the large volume of data, this study is limited to analysing the Twitter account @RealScientists. The account (@RealScientists) is curated. Researchers, Scientists, science writers, communicators and artists take turns managing the account for one week at a time. Since it started in February 2013 it has hosted curators across a multitude of disciplines and from across the world. Curators from Australia, New Zealand, Sri Lanka, America, Canada, The US and Sweden have taken part.
3.2 Approach
Given the complexity of analysing a platform as large as twitter, the researcher concluded that selecting one account, the @RealScientists account, for an in-depth study was appropriate. A mixed method approach was taken to analyse the @RealScientists Twitter account. The quantitative tradition follows that social science is no different to any other science and as such can be measured. This is most often achieved by conducting experiments, thereby generating your own data or analysing data already in existence. In contrast, the qualitative tradition allows for an inductive analysis of data. The researcher is encouraged to explore the subjects of their study with greater depth, and at a closer degree than in the quantitative tradition. Mixed methodology is becoming more common in Social Science research. According to Zina O’Leary it can help capitalise “…on the best of both traditions and overcome many of their shortcomings.” (O'Leary 2010, Chapter 8) This ability to draw on both research traditions can add depth and meaning to quantitative data.
The researcher conducted a quantitative analysis of the @RealScientists’ account archive from the 9th February 2013 to the 2nd March 2014. This was comprised of 55 separate curators and circe 25,000 tweets. Twenty curators were selected for a qualitative analysis of their tweets. Interviews were conducted with individual curators & one of the account administrators. Finally a survey of the account’s followers was conducted.
3.3 The Archive
Twitter provides access to a comprehensive archive of users’ tweets upon request. The archive is delivered in both .csv and .js form and details the time, location and content of each tweet including retweets for the history of an account. The administrators of the @RealScientists account provided the archive for research in May 2014. It contained all the data from the account since it began in February 2013, including all tweets, retweets, links and photographs sent throughout this period. As the @RealScientists account has been in operation for more than a year, it was decided to quantitatively analyse the data from inception until the 2nd March 2014 (13 months). This comprised of 55 separate curators. From within that dataset 20 curators were selected for a more detailed qualitative analysis. These were chosen at random using a lottery system.
3.4 Quantitative analysis
People on the micro-blogging platform, Twitter, have developed particular stylistic norms in conversational tweeting. For example, hashtags are employed to join in a larger discussion around an event or news story and are indicated by using the # symbol, i.e. #ScienceCommunication.
At the beginning of the analysis the following criteria were determined:
- @username at the beginning of a tweet indicates a direct, but public, response to someone on Twitter.
- RT indicates a retweet.
- @username, MT or RT in the middle of a tweet indicates engagement and/or comment on the original tweet.
Java et al (2007) determined that there were four distinct categories in Twitter communication: Conversations, Information sharing, news reporting, and daily chatter. For the purposes of the quantitative analysis, the following categorisation was applied:
@username
Conversation with other Twitter user – Intimate information sharing
.@username
Broadcast conversation – non-intimate information sharing
RT
Retweet – information broadcasting
____ RT/MT
Modified retweet – information broadcast with commentary.
Tweet
Information sharing/requesting
The database from 12th February 2013 to 2nd March 2014 was reviewed to determine the number of original stand-alone tweets, @username engagements and RTs on the account as a whole. In order to ensure that only tweets by the curators were captured, all tweets beginning with [ADMIN] were removed prior to analysis. Each curator was then separated and a similar analysis was done on an individual basis.
The curators were then sorted into groups according to their field of research or employment. A comparison was done between fields to determine:
- Which field was more active on the account
- Which field tweeted more
- What types of links were tweeted for which field of research
- Who had conversations versus who tweeted directly
As the Twitter archive includes a breakdown of links sent by an account the research looked at the following:
- How many links were sent
- How many of the links were photographs (TwitPic, Twitter, Instagram)
- How many of the links were videos (YouTube, Vimeo, TED)
- In the links tweeted, was there a dominant source
3.5 Qualitative analysis
In order to get a richer understanding of the @RealScientists account a qualitative analysis was conducted on twenty curators selected from within the dataset. These were:
- Dr Rachel Dunlop (@DrRachie)
- Dr Cameron Webb (@Mozziebites)
- Dr Paul Willis (@FossilCrox)
- Dr Helen Maynard-Casely (@Dr_HelenMC)
- Dr Kristin Alford (@Kristinalford)
- Phil Torres (@phil_torres)
- Mia Cobb & Julie Hecht (@DoUBelieveInDog)
- Eva Amsen (@easternblot)
- Marisa Wikramanayake (@mwikramanayake)
- Dr David Hawkes (@mrhawkes)
- James Hutson (@jameshutson)
- Dr James Smith (@theotherdrsmith)
- Ethan Perlstein (@eperlste)
- Luis Quevedo (@Luis_quevedo)
- Dr Katherine Mack (@Astrokatie)
- Dr Helena Ledmyr (@Helena_LB)
- Dr Darren Saunders (@whereisdaz)
- David Winter (@TheAtavism)
- Dr Will Grant (@willozap)
- Michelle Bannister (@astrokiwi)
The archive for each curator was separated out and initially analysed for content, style, pictures, links, and discussion topics. The researcher allowed themes to emerge organically as the analysis progressed. Additional categories were added to the as they occurred in the text.
3.6 Interviews
The twenty curators selected were approached for interviews. Four responded positively, one of whom was the account administrator. Due to time zones and locations, these interviews were conducted by Skype and email. The Skype interviews lasted for approximately 20 minutes each. These were recorded and then transcribed. While a number of the questions remained the same for each interview, several were based off the interviewee’s week in control of the Twitter account. The week of tweets was assessed and questions tailored to the individual interviewee were added to the general questions. The interviews were semi-open. There was room for the interviewees to develop on their answers and to offer insights on their time in charge of the account.
The data from the interviews was used to allow for a more in depth understanding of how these curators approached the task of managing the account for a week, their expectations in advance of hosting and the effects or lack thereof following their week curating.
3.7 Followers
While the focus of this thesis is on the researchers and their reasons to communicate on twitter, it was considered important to assess the possible motives to follow a curated science account. To investigate this aspect, the researcher developed a survey on Google docs. The link to the survey was tweeted by the administrators of the @Realscientists account. As of 30th June 2014 date 60 followers had responded. The researcher asked the following questions:
- Age
- Gender
- Location
- How long have they been on Twitter
- Are they a scientist or involved with research
- How did they find out about the account
- Have they interacted with the curators
- Have they considered applying to be a curator
- Have they learned anything from the curators
- Do they follow any other science based twitter accounts
The results from this were analysed quantitatively.