Psypective: Identifying Political Influential and Opinion Leaders On Social Media

According to each measure of influence, the categorization of Twitter profiles of users included in the top 20 most influential lists, made up of the qualitative portion of the study of various types of influencers, is useful for researchers to identify influence and select measures for future studies. As they relate to the idea of opinion leaders, this discussion of local indicators of power offers guidance for future studies of political discussion groups.

Defining influence, the opinion leader of Katz and Lazarsfeld, as defined in their Two-Step Flow Hypothesis, can influence his or her close personal relations by exerting social pressure and social support. There are four main facets of power that are suggested: getting a following, seen as an authority, knowledgeable/expert, and in a position to exert social control and social support/social embeddedness within their local society. In reaction to theories of direct social media impacts, Katz and Lazarsfeld claimed that a small portion of the population, opinion leaders, who then control the public, was influenced by mass media. While most support a certain mix of metrics, various ways of operationalizing influence are based on the inclusion or exclusion of different aspects of influence. In the Two-Step Flow studies, the most labour-intensive, but most popular, approach is to ask people who they are affected by and whether they think they are influential. The presence of social media and social networking sites online has ensured that there is a great deal of trace data on which networks of influence can be created. Studies using access to large-scale social data have recently assumed a concept of power that is more readily quantifiable. By the number of followers and/or how far a post travels, they prefer to quantify power. It is the following facet of impact that these studies depend on to provide an operational description. Being an expert is the key facet of impact that these studies are concerned with. To assign degrees of power to individuals, some studies use complicated approaches, such as rating language quality or monitoring URLs over time. They have several metrics to describe and operationalize power, but there is little understanding of how they play out empirically or how they relate to each other. They analysed the degree to which different metrics agree, and in turn, what different aspects of influence may exist, using a rank correlation coefficient to compare metrics. Both of this show how central a node is within a network; there is a follow-up to the facet of control discussed.

In short, they interpret these results to mean that the centrality of the index and eigenvector ranks according to a similar facet of power, one that is somewhat different from the facets calculated by our other metrics. At this point, they can conclude that users are indeed defined differently by various metrics, sensitive to different facets of control.

Cha and Gummadi pointed out that indegree is the fundamental measure of influence used across disciplines, while popularity can be calculated better than actual influence. The assumption made by those using the index to calculate influence is that a broad follow-up is the most important facet of the influence.

Those are potentially essential aspects of the mechanism of control. All considered positive indicators of influence were posting political content, mentioning the political elite, and holding political conversations and were consistently identified in this top 20 list, while it was not for the network-wide clustering coefficient list. They may be well placed to impact locally, but may be very tiny in that locale. As other Twitter effect studies have done, it may be prudent to set a minimum index standard.

As is the case for all the influence measures we have used in this analysis, clear assumptions are embedded in operationalization about which aspects of the influence mechanism are most relevant. Discussion Generally, it is believed that the most important facets of influence are based on who follows a specific user and how much they speak about that user or whether the user is regarded as an expert. While it is important to make decisions about how to categorize users and which metrics are most relevant, using out-of-context measures of influence may lead to misleading or inaccurate results.

Humans can assume that journalists and politicians are trusted experts who supervise most measures of power, particularly the standard measures of centrality. Influential is a clearer term that can be beneficial, but it also suggests that we then lack clarification about the complexity of the impact of the social process. Opinions shifted when someone paid attention to a mass message in a group and then used their role within that group to impact the other members directly.

Whereas this aspect is ignored by most impact metrics, the use of the clustering coefficient can be useful. This do not advocate the coefficient of clustering as a stand-alone indicator of impact. Measuring the power of the communicator means the ability to separate and accurately weigh the components of influence. In short, to classify the most prominent members of the #CPC and #NDP Twitter groups, our research has used several indicators of influence. These opinion leaders control others in their personal network, unlike our main journalists and lawmakers, who have network-wide patterns of influence. These findings are instructive for future studies of impact on Twitter and possibly other sites, albeit unique to our case. Dillard, J., Segrin, C., & Harden, J. Primary and secondary priorities in the formation of messages of interpersonal power. Her dissertation focuses on the ways in which people use technology to wield political control.