Thread by @taylor_owen

Yesterday we released our first report from the Digital Democracy Project. Over the course of the upcoming Canadian election we will be publishing weekly results of our analysis of the media ecosystem and the role of mis/disinformation in voter behaviour. Some details (thread):

Our goal is to merge large scale media ecosystem analysis with survey research. The online data analysis team is led by @derekruths and will study a wide range of online data, including Twitter, Reddit, YouTube, some public-facing Facebook, Blogs, and all Canadian news websites.

The survey team led by @PeejLoewen and will be running regular nationally representative surveys as well as a metered media monitoring survey which will collect the online consumption of 2000 Canadians during the writ period.

We are also partnering with the Digital Ecosystem Research Challenge (which @lizdubois & I are running) by providing 18 international research projects w/ the data we are collecting about the Canadian election. Details of that collaboration are here: digitalecosystem.ca

Now some detail on the first of what will be approximately ten election reports over the coming three months. This initial memo was an effort to provide a baseline analysis on which future studies can map.

On the survey side we tried to get a sense of issue preferences of voters, media consumption, knowledge levels on election issues, & trust in info sources. On the online data side, we studied four groups of Twitter users: candidates, journalists, 3rd party groups and the public.

The full report is available here: ppforum.ca/wp-content/upl… (PDF) Here are some findings we want to highlight:

1. Both news consumption (traditional print, broadcast & online media) and trust in traditional news outlets are high, across partisan lines. This is very different than the US and mirrors what @PeejLoewen, @jandrewpotter and I found in this study last yr: policyoptions.irpp.org/magazines/octo…

This reinforces one of the core rationales for this project, that Canada needs to develop policy based on analysis of its own media ecosystem, not solely drawing on what we know about mis/disinformation from other countries.

2. Exposure to both mainstream media and to a greater extent social media is associated with being more misinformed (as measured by knowledge of key facts on election issues). The effect is highest for partisans on social media.

There may be many factors at play here, including motivated reasoning; partisans bring a strong frame of reference to how they view facts. Which tells us something important about fact checking efforts. Correcting facts is not enough if the underlying narrative frame is ignored.

But there is a difference between being uniformed or misinformed and the problem of mis/disinformation. Those that were uniformed didn't know the answer. Those that are misinformed gave the wrong answer. Neither measure tells us about the presence or threat of false information.

So this initial report doesn’t tell us anything about the threat of disinformation campaigns, nor the flow of inauthentic content, (issues we will track over the election), only the degree to which voters are currently un/misinformed.

3. There is a disconnect between the issues voters say they care about (such as healthcare and the economy) and what journalists and candidates are talking about on twitter (gov ethics and foreign policy).

This is not necessarily a bad thing, journalism and politics both have important agenda setting functions. But it’s worth pointing out the disconnect and watching how it evolves over the campaign.

4. The environment emerged in our study as the dominant policy issue for Canadians, though markedly less so for Conservative party supporters. This is consistent with several recent polls, and is unique among Canadian elections.

There is lots more in the report, which is available here. Again, it is a baseline and future reports will focus on a wide range of topics and new data sets. We will be releasing reports every two weeks until the election and then every week during. ppforum.ca/wp-content/upl… (PDF)

Finally a shout out to the amazing team that is driving this project, including: @derekruths @PeejLoewen @aengus_b @EricMerkley Oleg Zhilin @smaclellan @rgorwa @jandrewpotter Meghan Keenan Beata Skazinetsky
@ppforumca @MaxBellSchool @Mozilla @luminategroup @jwmcconnell Rossy Fnd