Skip to main content
Search form
  • 00:05

    [Researching Public Attitudes to Personal DataUsage Using Twitter Surveys]

  • 00:10

    ANSGAR KOENE: My name is Ansgar Koene.I'm a Senior Research Fellow at Horizon Digital EconomyResearch, which is a research institute at the Universityof Nottingham.And the research that we do at Horizonis basically looking at online data,the way it's being generated, the way that it'sbeing used in research, which can be data from social media.

  • 00:33

    ANSGAR KOENE [continued]: It can be data from devices like mobile phones or internetroutes, or things like that.So it's part of the UK's focus on developing digital economy.Within Horizon digital economy research,my research focus over the last one and a half years

  • 00:53

    ANSGAR KOENE [continued]: has been on citizen-centric approachesto social media analysis, which meanswe're looking at the ethics of how data that is availableonline is used when social scientists or other researchersare trying to understand human behavior.

  • 01:14

    ANSGAR KOENE [continued]: [Why did you choose Twitter as a research tool?]So in the current research study,we use Twitter as a way to try and accessa broader kind of audience.Basically, we're trying to use the broad spectrumproperties of Twitter that peoplecan see what is being treated, even if they're not

  • 01:36

    ANSGAR KOENE [continued]: part of a close friends circle, and sousing hashtags, for instance, as a wayto get things into other screens so people can see them.The idea basically came, because we were running a studyto understand people's attitudes towards online research

  • 01:58

    ANSGAR KOENE [continued]: and what kind of information theyshould want to have in order to knowif they wanted to give consent for having their data beingused.We originally ran that as a study that was justannounced via an online article and some mailinglists, but didn't get enough participants.

  • 02:20

    ANSGAR KOENE [continued]: Especially, the participants we gotwere not across the spectrum.They were very much from the sortof left side of the political spectrum.So we wanted to use Twitter as a way of getting at a broaderspectrum of people.Also, within the research group, we'redoing some studies to look at the use of Twitter

  • 02:41

    ANSGAR KOENE [continued]: by the research community as a means of doingpublic engagement.We know that within the Responsible Researchand Innovation Agenda, so we are,researchers are asked to interact with the publicto understand the public's attitudes towards the researchthat's being done.A lot of research groups are using Twitter as one

  • 03:04

    ANSGAR KOENE [continued]: of the means of doing the communication,because it's relatively easy.You can just have a Twitter account for the researchproject, and frequently say somethingabout what you're doing.So we're doing a research project try and lookat who we actually reach when we useTwitter in that kind of a way.And in a sense, our other study of looking

  • 03:24

    ANSGAR KOENE [continued]: out accessing a broader public for our survey on consentwas also the way of gauging who is this public that wereach when we use Twitter?And so our experience has been basically

  • 03:45

    ANSGAR KOENE [continued]: that the audience that you reach is not quite asbroad as what one would hope.Approximately one in five people in the UKsupposedly have a Twitter accountand may occasionally use it.But the practice is that peoples tend to communicatewithin a kind of bubble.They only really talk to people who have common interests,

  • 04:09

    ANSGAR KOENE [continued]: established common interests.So when we use Twitter from our project research,our recent projects Twitter account,or we ask other researchers to help with disseminatingour tweets through their Twitter accounts, what you tend to getis just other people who are also doing research.

  • 04:31

    ANSGAR KOENE [continued]: And so reaching a broad audience is quite difficult, actually,with Twitter.So in order to try and expand that,we also used the promoted tweets option.So yeah, basically, we paid Twitter a small amount moneyto put some of our tweets out like advertising

  • 04:56

    ANSGAR KOENE [continued]: kind of tweets.You can specify who these should go to,what kind of a demographic you want to have.We just specified a very broad demographic, basically anybodyliving in the UK.So that way, we reached a large number of peopleto got out the tweets shown to them.So the impressions was about 30,000.

  • 05:20

    ANSGAR KOENE [continued]: However, the actual engagements with it stayed quite low.Because most people, if they're not alreadyinterested in this area, it's very difficultto convince them to get interestedin this with just the 140 charactersof a tweet, which is actually probably more like 100characters if you deduct the number of characters

  • 05:43

    ANSGAR KOENE [continued]: for the image and the number of characters for the linkto your survey that we're trying to run.[What challenges did you face and what would you dodifferently?]The lowest response rate was quite surprising to us.We had hoped that by using some key hashtags that are already

  • 06:04

    ANSGAR KOENE [continued]: popular-- so in our case, since we were looking at ethicsof data usage, hashtags like privacy or digital rightswe used-- that they would trigger more responses.And the fact that our tweets werebased on stories that had previously been in the media,so we would remind people of an instance that

  • 06:24

    ANSGAR KOENE [continued]: was in the media-- for instance the Samsung, the smart TV thatlistens in to people in their home effectively.It was a big story in the news.So we tweeted about this, just asking,do you think this kind of technology is OK-- yes, no,or depends?So this kind of a small, mini survey we were hoping we

  • 06:49

    ANSGAR KOENE [continued]: give a low threshold for people to participate.Unfortunately, apparently they didn't.It's not immediately clear why this is.One issue is probably that we weretrying to target-- make the tweets too generic, too broad.

  • 07:13

    ANSGAR KOENE [continued]: And that way, nobody felt directly addressed.So if we were to do this kind of a study again,we would probably want to, instead ofrun one broad kind of campaign, run multiple specific campaignsthat would target different segments of the demographics,so do one targeted at our student population, one

  • 07:36

    ANSGAR KOENE [continued]: targeted at Tory voters, et cetera,and then have each of these have their own min surveys.Obviously, wouldn't be able to compare resultsfrom these different mini surveys,because it's a different population.And you have to make a different kind of questionin order to address this different population.

  • 07:60

    ANSGAR KOENE [continued]: But if each of these links to sort of a broader survey,then you would still be able to use the data on that broadersurvey that all of them responded to.So that is one way that we're thinkingof potentially going in.The other thing is that having experimented with the using

  • 08:20

    ANSGAR KOENE [continued]: promoted tweets, I'm now thinkingthat the monetary advantage that you might get out of usTwitter, instead of using, say, expert surveyingcompanies, probably not that much of a difference.You might actually save some time and effortby going with a established kind of surveying group.

  • 08:41

    ANSGAR KOENE [continued]: [Were there any ethical issues you had to consider when usingTwitter?]So for our study, I think the ethicswas reasonably straightforward.Because we were effectively not doing a Twitter analysiskind of study.But we were doing, effectively, a standard survey kindof study.

  • 09:01

    ANSGAR KOENE [continued]: So participants were clear that theywere answering a survey question that was clearlygoing to be analyzed.There is an inherent problem with this kindof a use of Twitter when it comesto getting good communication for consentwith participants, which is that if somebody just-- especially

  • 09:26

    ANSGAR KOENE [continued]: if you use the built-in type of surveying toolthat Twitter has, where you can put some questions.And they can just press yes or no.You don't really get any feedback as to who it was.So asking them back again, do youagree with having something done with thiswill be very challenging.

  • 09:47

    ANSGAR KOENE [continued]: On the other side, since you don't reallyknow anything about the person, it isn't really personal data.So we don't really have that kind of a problem.[When should you use Twitter for research?]In a sense, Twitter actually does help researchers,especially-- I mean, the use of this feature of embedding

  • 10:11

    ANSGAR KOENE [continued]: is small survey into it is potentiallysomething that's easy to use if you have small question.Probably, it's best for doing some piloting,for just getting some quick rough responses about an issueand just to see whether it's a direction wherethere is interest.Generally, the number of questions that you can ask

  • 10:34

    ANSGAR KOENE [continued]: is quite limited.The text that you put into these survey questions in the bitgets subtract it from your 140 charactersso that you get even less space.And you can't combine the survey bitwith an image, which is a big drawback.Because generally, you need an image in your tweet

  • 10:55

    ANSGAR KOENE [continued]: in order to attract people to have a look at it.But then you wouldn't be able have the survey bit in itanymore.[Conclusion]Thank you for watching.I hope this has been informative.We tried to cover the various aspects thatrelated to the use of Twitter, the types of challenges

  • 11:19

    ANSGAR KOENE [continued]: that we faced doing our study, and someof the advantages of using Twitter,some of the nice features that it has.

Video Info

Publisher: SAGE Publications Ltd

Publication Year: 2017

Video Type:Video Case

Methods: Social media research, Survey research

Keywords: access to information; attitudes and beliefs; audiences; challenges, issues, and controversies; communication breakdown; media coverage; practices, strategies, and tools; public engagement; targeting; Twitter ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



Dr. Ansgar Koene explains how he used Twitter to research public attitudes toward data use. He embedded mini-surveys into Twitter posts, asking the people how they felt about data use instances that were widely covered in the media. He discusses the benefits and drawbacks of this project and of researching with Twitter in general.

Looks like you do not have access to this content.

Researching Public Attitudes to Personal Data Usage Using Twitter Surveys

Dr. Ansgar Koene explains how he used Twitter to research public attitudes toward data use. He embedded mini-surveys into Twitter posts, asking the people how they felt about data use instances that were widely covered in the media. He discusses the benefits and drawbacks of this project and of researching with Twitter in general.

Copy and paste the following HTML into your website