Research by the Pew Research Centre illustrates six different types of conversations on Twitter.

Divided: Polarised Crowds. They feature two large and dense groups that are not often connected and rarely interact, even though they discuss similar issues. They focus on highly divisive topics, such as politics. They ignore one another, rather than argue with one another. They rely on different sources and use different hashtags. For example, liberals link to mainstream news, while conservatives link to other websites.

Unified: Tight Crowds. Learning communities with highly interconnected people with few isolated participants discuss insular events and experiences. For example, conferences, professional topics, and hobbies. These conversations demonstrate how social media can generate mutual support.

Fragmented: Brand Clusters. Disconnected individuals discuss well-known products, services, and popular subjects, such as celebrities. This is typified by people posting about brands without talking to other individuals. These patterns show how social media can sometimes be a vehicle for reaching for institutions or popular entities, rather than exchanging ideas.

Clustered: Community Clusters. Small-to-medium sized groups that interact and exchange information, such as to discuss global events. This illustrates how some major issues can lead to diverse audiences sharing their distinct opinions and perspectives.

In-Hub and Spoke: Broadcast Network. Discussion of breaking news stories where many people repeat prominent news messages. In-hub interactions are retweets and replies from many disconnected people to a large news account. Most people are simply re-sharing information without interacting, while a minority of people will discuss the news with one another. This demonstrates the enduring power of mass media as trusted messengers.

Out-Hub and Spoke: Support Network. These are individuals interacting with official accounts to receive customer service and resolve complaints. This includes businesses and government services. Out-hub interactions describe one business account replying to many disconnected users. This pattern highlights how social media creates a measure of accountability and transparency, as individuals can benchmark and evaluate the performance of public institutions. Customer support can be measured by the efficiency and reach of these accounts.

Source: image and research by Pew Research Centre.