[Air-L] Research on Twitter - a couple of questions

sava saheli singh savasaheli at gmail.com
Mon Jan 16 21:42:09 PST 2012


I don't claim to be an expert, but I am doing research on twitter (very
different aspects and data collection) and I will say this:

   1. yes. many people do not geolocate their tweets or identify where they
   are, some have fictitious or humorous locations as well. while this in
   itself is not a problem, if it is being used as the only legitimate
   geographical identifier for data, I'd say that was problematic.
   2. same story with images - you are right. people use graphics, older
   pics, abstract pics... and even if they use their own pics, "guessing"
   their age or gender based on the avatars and reporting that as legitimate
   data is definitely not responsible data gathering.

I hope this helps,
sava



On Tue, Jan 17, 2012 at 12:29 AM, Monica Barratt <tronica at gmail.com> wrote:

> Dear AoIR members
>
> Recently there have been more scholarly articles published using
> Twitter data in the health field. I've found myself as a peer-reviewer
> of one such article. In case the authors happen to be subscribed, I'll
> keep the exact topic quiet... but basically what the authors are doing
> is using Twitter's advanced search function to compile a database of
> Tweets which they then subject to a simple content analysis.
>
> I have done my own research and discussed the issue with local expert
> Axel Bruns, however I'm still a bit unclear on a couple of points, and
> wondered if anyone reading could assist me:
>
> 1. The authors draw samples of Tweets over a specific time frame from
> specified US cities. They appear to assume that by using the advanced
> search option 'Near this place' and entering the city name will bring
> up all tweets from that city or surrounds (see
> https://twitter.com/#!/search-advanced ). But wouldn't that only bring
> up the tweets from people who have nominated their home city or
> geolocated their tweets? Don't many users do neither of these things
> and therefore the corpus of data would be incomplete?
>
> Can anyone verify this ... and also does anyone know if there is any
> research on the proportion of Twitter users whose location would be
> known and therefore would be included in such a dataset?
>
> 2. The authors use profile images to ascertain the approximate age and
> gender of account holders. In my experience, many people use profile
> images that do not represent themselves - eg. celebrities or past
> images of themselves or images of themselves with others.
>
> Is this an accurate or useful way of dealing with Twitter profile data
> or is it too flawed as a technique to be useful?
>
> While I'm not an expert in Twitter research, it seems to me that
> Twitter research is on more solid ground when using a hashtag to
> identify a corpus of Tweets, as this is the method Twitter users
> employ too.
>
> Thanks for your thoughts/help
>
> Monica
>
> Research Fellow @ National Drug Research Institute, Curtin University
> _______________________________________________
> The Air-L at listserv.aoir.org mailing list
> is provided by the Association of Internet Researchers http://aoir.org
> Subscribe, change options or unsubscribe at:
> http://listserv.aoir.org/listinfo.cgi/air-l-aoir.org
>
> Join the Association of Internet Researchers:
> http://www.aoir.org/
>



More information about the Air-L mailing list