How to use and interpret the UK Elections 2015 charts
The charts show how the level and nature of discussion on Twitter of the 6 main UK political parties (Conservatives, Labour, Liberal Democrats, UKIP, Greens, and SNP) has changed over time during the 2015 general election period. Most of the charts are based on the automated identification, collection, and analysis of around 160.000 tweets per day (excluding retweets) on the topic of the election. As there is likely to be a close relationship between news stories and discussion of the election on Twitter, the charts are overlaid with key election news stories, so that the level of their impact can immediately be viewed.
You can use the drop-down menu at the top centre to view different charts. The following charts are currently available:
- Number of Tweets: Percentage of total tweets (that we collected) that refer to each party.
- Number of Users: Percentage of total users (who sent tweets that we collected) that refer to each party.
- Positive Tweets: Percentage of tweets that expressed positive sentiment towards each party.
- Neutral Tweets: Percentage of tweets that expressed neutral sentiment towards each party.
- Negative Tweets: Percentage of tweets that expressed negative sentiment towards each party.
- Sentiment Tweets: The percentage of positive tweets decreased by the percentage of negative tweets towards each party.
- Users Positive Tweets: Percentage of users that expressed positive sentiment towards each party.
- Users Neutral Tweets: Percentage of users that expressed neutral sentiment towards each party.
- Users Negative Tweets: Percentage of users that expressed negative sentiment towards each party.
- Sentiment Users: The percentage of users that have expressed positive sentiment, decreased by the percentage of users that have expressed negative sentiment towards each party.
- Prediction: Our prediction of the result if the General Election was held today.
Because the negative aspects of politics are discussed frequently on Twitter, simply interpreting the total number of tweets about a party as an indicator of their popularity with voters is likely to be flawed. As a result, some of the charts incorporate automated sentiment analysis in order to determine whether the sentiment being expressed in tweets mentioning a party is positive, negative, or neutral.
Given that most of the data published by the major polling organizations – including YouGov,Lord Ashcroft, and Ipsos-Mori – suggests that neither Labour nor the Conservatives are likely to win an overall majority, the next government is likely to be based on a formal or informal coalition between parties. As a result, all of the above charts can also display data for the most likely of these coalitions. You can use the button in the top left corner to switch between data relating to individual parties and potential coalition partners.
Throughout the election period, opinion polls have shown that public often have differing views on the parties and their leader – and that these may change independently of one another. As a result, you can use the button in the top left corner to switch between data relating specifically to the party leaders and data about the parties more generally. Again, all of the above charts are available for both of these options.
The circles on each line indicate the point at which a major election news story relating to a particular party was published. Clicking on a circle will display a pop-up window showing the title and a brief description. You can click the ‘Read More…’ link to view the story in full as it appears on a particular news website. In most cases, this link points to the BBC News website. Some news stories feature more than one party, and these are marked the relevant line for each party.
The charts are normally updated on a daily basis. You can mouse-over any part of each line to view the exact figure for a particular day. The figures typically refer to the percentage share of the total for each party (e.g. for the ‘Number of Tweets’ chart the figure indicates what proportion of the total number of tweets collected refer to a particular party).
You can also click on the data labels in the top right corner to show or hide the data relating to each party.