Share of Voice

Hillary Clinton vs Donald Trump on Twitter

Election Share of Voice Project

Social media data are front and center in the U.S. national debates, often presented alongside a livestream of the debate itself by major news networks as a share of voice (SOV) pie chart or line graph, but few understand how social media share of voice data are collected and displayed. These charts are being generated by the social media companies via keywords, which act as filters to measure their data, but if the keywords are not inputted correctly, or are biased, the share of voice metric will not be accurate.

The Election Share of Voice Project is meant to provide transparency into how these charts are generated, demonstrate a best practice for displaying keywords used to create SOV, and serve as a resource for journalists who are looking to use Twitter SOV in their election reporting.

Methodology

The project uses Sysomos MAP, which provides thirteen months of real-time and daily access to the Twitter firehose.

We selected and counted posts for top campaign hasthags: #MAGA (for Trump) and #ImWithHer (Clinton) which were highlighted using Sysomos MAP’s analysis tool from the period of June 26-September 25, 2016 (Boolean: trump OR donald; clinton OR hillary OR hilary). Hashtag analysis will be added on a rolling basis. We also include mentions of @realDonaldTrump and @HillaryClinton, which are the official campaign handles.

Contact

Jennifer Grygiel is a team member of the Illuminating 2016 project. She is the lead on the Share of Voice project.

Contact: jgrygiel@syr.edu, @jmgrygiel