Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Figure 1.

Number of tweets sent by Twitter users (logarithmic scale).

Shows the number of tweets that mention ‘IPCC’ sent by each author whose tweets were collected. The data is presented on a logarithmic scale showing a very skewed distribution of the tweets by tweet authors, with only a few authors sending many tweets about the IPCC and many authors sending only a few tweets about the IPCC.

More »

Figure 1 Expand

Figure 2.

Number of tweets in which a username was mentioned (logarithmic scale).

Shows how many times different usernames were mentioned in the collected tweets. The data is presented on a logaritmic scale and it clearly shows how skewed the distribution of usernames mentioned is. Few usernames were mentioned many times, while many usernames were mentioned only a few times.

More »

Figure 2 Expand

Table 1.

Example tweets from each category of Twitter user.

More »

Table 1 Expand

Table 2.

Twitter users with ten or more conversational connections, coded by attitude to IPCC.

More »

Table 2 Expand

Table 3.

Most frequently used hashtags associated with science.1

More »

Table 3 Expand

Table 4.

Most frequently used hashtags associated with Australia.

More »

Table 4 Expand

Table 5.

Most frequently used hashtags associated with political campaigns in the United States.

More »

Table 5 Expand

Table 6.

Hashtags associated with social aspects of climate change.

More »

Table 6 Expand

Figure 3.

Detecting three communities of Twitter users from conversational connections only (resolution: 1.9, modularity: 0.422, modularity with resolution: 1.104).

Each node represents a Twitter user. Size of nodes is correlated with that user's number of conversational connections. Detected communities are differentiated by color. Colors were selected randomly and should not be associated with political stance. Thickness of the edges reflects the number of conversational connections between the two usernames connected by the edge. Proximity between the nodes reflects local closeness, as nodes with more connections to each other than to the other nodes in the graph are clustered closer to each other.

More »

Figure 3 Expand

Figure 4.

Detecting communities from conversational connections with additional coding by views on climate change.

Twitter users were manually coded according to the content of their tweets and Twitter biography within the population of tweets analyzed. Each node represents a Twitter user. Size of nodes is correlated with that user's number of conversational connections. Climate change unsupportives, purple; climate change supportive, red; climate change neutral, green; did not tweet, light blue. Colors were selected randomly and should not be associated with political stance.

More »

Figure 4 Expand

Table 7.

Categorization of Twitter users by tweet content and profile information.1

More »

Table 7 Expand

Table 8.

Conversational connections between different categories of Twitter users.

More »

Table 8 Expand

Table 9.

Results from the chi-squared test after calculating the expected values (Chisquare = 203,98).

More »

Table 9 Expand