What is Trump tweeting about and how his tweets resonate among the Twitter publics?
Curious and triggered by the noticeable influence Trump’s tweets have on national and international issues, when they often weigh more than official press releases by the White House, together with Mouloud Kessir and Sangam Slipakar, we decided to observe how his tweets resonate with the wide Twitter publics.
We decided to observe: 1. which issues that Trump tweets about are gaining the most attention among the Twitter publics, 2. are these issues national or international by nature and 3. how their resonance changes when comparing pre-election and post-election time period.
In order to gain deeper insights, we track his tweets from the moment he became the official and only Republican candidate – May 3rd 2016. And because we’re aware that pre-election and post-election issues will differ, having in mind his new role as a President, we are dividing the analysis in two phases. Also, having in mind that most of the issues will be national and international in nature, we are curious to additionally observe how the interest among the twitter publics differs when it comes to these different types of issues, given the fact that he has increasingly international followers.
The research question we focused on is:
Some of the findings include:
Resonance of issues and of categories of issues overtime
From the data available we are able to investigate the categories of issues that resonated the most in the two time periods (pre- and post-election). It can be observed that most of the issues in the pre-election are from the category “national issue” while few issues belong to the category “international issues”. In the post-election period, although the number of “international issues” increases, the “national issues” still outnumber them.
This visualization also gives us insight into the issues that resonated the most overall, and in each time period separately. We can observe that the most resonating issue overall are the issues of North Korea (international issue) and the issue of media (national issue). They both belong to the post-election phase. The most resonating issue among the Twitter publics in the pre-election time period is the issue of his presidential opponent – Hillary (Clinton). By observing the data set, we can see the proportion of issues in both data sets that were labeled “national” and “international” and to compare resonance of the categories of issues (by looking at the total number of favorites they gather per category).
How the resonance of issues changes pre- and post-election
This diagram allows us to observe how the same issues resonated in two periods, during the presidential campaign, and after the election. It gives us insight into the evolution and flow of the issues. From this diagram, we can observe that in post-election, the issues that resonated the most among the Twitter publics are issues concerning the USA, North Korea, radical islamic terrorism, Trump himself, “take a knee” and media.
Also, it allows us, on a surface level, to see how the interest and resonance changed and if some issues are exclusive to only one period. For example, if we take a look at the issue “take a knee”, concerning the protests of the Afro-American athletes from the National Football League (NFL) against rising racism and white nationalism, we
observe that this is a post-election issue only. The same is the case with the North Korea and Muslim ban issues. As some of these might be “events-related” (being dependent on external factors and global events), others, such as the Muslim ban are related to policies developed by POTUS himself.
Some issues are present in both time periods, such as his presidential opponent Hillary Clinton, USA, radical Islamic terrorism and Campaign being some of the issues.
Frequency of issues vs resonance of issues
The frequency of the issues is another aspect we can observe. But, the frequency of an issue in a dataset is not an indicator of resonance of the issue. If by frequency we understand how many times it appears in the data set, by resonance we understand the attention it draws, by looking at the number of favorites certain issue gathered.
What we can observe from the analysis is that not always bigger frequency results with bigger resonance. If we take, for example, the “USA” issue, we can see that pre-election its frequency is 2 but it resonated better in comparison with the post-election period when the frequency is bigger (13), but proportionally the resonance is smaller.
If we observe “Hillary” issue, we can observe that the frequency is four (4) pre_election vs five (5) post_election, but the resonance pre_election is way bigger than the resonance post_election.
And, for everyone curious, here is the most resonating tweet:
What we investigated was limited by the nature of our task – we were very careful to not get into a network analysis or content analysis, but to simply “observe”. However, further steps are needed if we would like to investigate more thoroughly the content Trump is creating while tweeting and the reactions to that content. Some of the additional insights we could gather are: investigate the frequency of issues more thoroughly; look into the media content (by looking at the URLs in his tweets) and see the most resonating ones; look for topical conversation by looking at retweets and quoting; conduct a content analysis of the topical conversation, identify clusters of actors around certain issues and observe what these clusters are about; look at the mentions to see Trump’s proximity with certain actors to identify the actors network and additionally identify the issues network before finally do a content analysis to get insights into the issues they are discussing.
The full report, where you can find more info on the data collection, sampling, categorization of issues and the analysis itself, can be accessed here. It will also allow you to inspect the visualizations in better resolution 🙂