by Sentimenti Team | Aug 21, 2024 | Politics and Social
On Platform X, The New York Times (NYT) has swept the post: „Donald Trump has boasted that, if he returns to office, he will dispatch forces to patrol the border, suppress protests and fight crime. His vision of using the military to enforce the law domestically would carry profound implications for civil liberties.”
Let’s take a look at the emotion in the comments under this post:
Sentiment in reader comments
We investigated what sentiment this post triggered in the audience – users of Platform X. For this purpose, we analyzed the comments posted under the NYT entry. In addition to the quantitative structure, we measured the intensity level of the sentiment expressed in the comments. The average intensity of negative sentiment was significantly higher than the average intensity of positive sentiment.
This simple measurement, however, does not answer the fundamental question: what did the readers-consumers of the NYT specifically feel when they read the posted entry. What emotion did the text evoke in them and does it matter in the evaluation. Only an in-depth analysis conducted in terms of measuring individual emotions provides an answer to such a question.
Dominant emotions in reader comments
The emotions we can measure include anger, fear, anticipation, surprise, trust, sadness, disgust, or joy. They are both components of positive or negative sentiment. Therefore, only the measurement of their intensity shows what we can expect from the behavior of their “owners.” After all, we react dramatically differently when we feel sadness (withdrawal) and when we feel anger (action). By merely concluding that negative sentiment prevails, we can infer very little. But to the point:
In the case studied, the dominant influence on the level of intensity of positive sentiment was the emotion of anticipation. Its intensity varied depending on the content of the comment and ranged from 10% to 58%.
In the case of negative sentiment – its level was decisively influenced by the emotion of anger expressed in the comments. Its intensity, as in the case of expectation, varied, ranging from 14% to 87%.
As can be seen from these measurements, the intensity of anger was stronger than the intensity of expectation in all cross-sections, and this built the overall negative perception of the information contained in the NYT entry.
Arousal – a measure of the involvement of comment authors
In addition to analyzing sentiment and the individual emotions that make it up, arousal was also measured, which was saturated with the content of readers’ comments. Stimulation most simply shows the level of engagement with which content authors express themselves. In this case, the average intensity of arousal remained very high reaching 59%.
In the case of arousal, too, there is a wide variation – a reflection of the commitment with which the individual author of the commentary expressed himself. The range of arousal intensity was from 42% to 83%. This, as well as the average intensity value, shows that the NYT entry triggered the need for readers to respond saturated with a very high layer of engagement.
Conclusion
Measuring sentiment alone is not enough to conclude how the author or authors will behave. Anger, present in the comments analyzed, suggests a strong commitment for the commenters, even indicating a desire to change the situation, an impulse to action triggered by the message. If, on the other hand, the dominant emotion was sadness, also seen in the category of negative sentiment. The conclusions could be quite different – indicating withdrawal of the majority of commenters, resignation and reluctant but acceptance of the state of affairs.
From an analysis of sentiment alone, all this will not be deduced and, most importantly, we will not learn about the future intentions of the recipient, reader or commenter, measuring only his mood, without going deeper, into emotions.
by Sentimenti Team | May 23, 2023 | Market research, Politics and Social
Politics, emotions and polls are closely intertwined. Politicians of parties vying for seats in the upcoming parliamentary elections are trying to reach potential voters through all available channels. They are, of course, also using social media for this purpose.
Emotions in tweets by leaders of political groups (anger and trust)
One of the channels that politicians use to convey their opinions, but also the substantive elements of their election programs, is Twitter. Individual posts on the Twitter accounts of the following politicians were analyzed for the period April 11 to May 10, 2023:
- Adrian Zandberg,
- Szymon Hołownia,
- Mateusz Morawiecki,
- Zbigniew Ziobro,
- Władysław Kosiniak – Kamysz,
- Robert Winnicki,
- Donald Tusk.
It was investigated what emotions they were and how they might be perceived by Twitter users. The analysis involved measuring the intensity of selected emotions of anger and trust. Individual analysis results were published periodically on our accounts: Twitter and Facebook.
The activity of politicians on Twitter varied, but most importantly, the intensity of anger and trust that their posts evoked varied. The chart below shows collectively the values of the average intensity of the indicated emotions over the entire period under study.
Only the entries of Mateusz Morawiecki and Donald Tusk evoked higher trust over anger. Close to balanced were the entries of Wladyslaw Kosiniak-Kamysz, where anger and trust occurred at similar intensities. In the case of the remaining politicians, the emotion of anger far outweighed trust. We emphasize it was not the task of the analysis to assess the substantive content of the entries, but to measure the intensity of the emotions they evoked.
Emotions in leaders’ tweets vs. poll results
The results of polls are a product of both the substantive content of programs, speeches, publications, but also undoubtedly the emotions they evoke in the audience, i.e. potential voters. We asked ourselves whether it was possible to find a correlation between the emotions evoked by the tweets of political party leaders and the results of published polls. We based our research on surveys by the Institute for Market and Social Research IBRiS:
In the case of the United Right, the May 17, 2023 poll recorded a 1.5% drop in support compared to the April 11, 2023 results. During this period (which the poll covered), an increase in the intensity of anger with a concomitant decrease in trust was noted in the entries of the group’s leaders. The reason for this – as the results of measuring the intensity of anger and trust showed – may have been the tweets published on Zbigniew Ziobra’s account.
At the same time, the Civic Coalition recorded a nearly 2% increase in support. Whether and what impact Donald Tusk’s Twitter posts had on this situation is difficult to judge. However, it can be noted that the increase in poll support occurred in parallel with an increase in the intensity of emotions of trust in the statements published by the politician.
Looking at the emotions evoked by the tweets of Szymon Holownia and Wladyslaw Kosiniak-Kamysz, one can ask the question – is it thanks to the latter’s balanced statements that the PL2050 + PSL bloc recorded a slight increase in support? The question is legitimate, since in Szymon Holownia’s tweets one can notice a significant disproportion in the saturation of emotions of anger and confidence.
In the case of the Confederation and Left support polls, it can be noted that one of the factors in the lack of growth (Left) and decline in support (Confederation) in the May 17, 2023 poll may also have been a decline in audience trust in the content published on Twitter by the leaders of these formations.
The emotions that parties and politicians arouse in voters are important
Observing the Polish political scene, it is hard to resist the impression that the “actors” on it use the simplest tools to influence the audience. So what if they can have the best political and social programs, etc. if they don’t know (after all, hardly anyone measures emotions):
- How to present your program, your vision in a way that inspires confidence,
- That one should control negative emotions (in this case, anger). Too strongly and too insistently politicians try to “play” on negative emotions.
Looking at the results of emotion analysis and survey results, we should conclude that it is worthwhile to study emotions, and just as worthwhile to bet on positive emotions. It is worth monitoring emotions in spontaneous statements, and not just conducting surveys, because in the latter the respondent thinks about his emotions, rather than expressing them.
Part of the political scene trusts Sentimenti. By commissioning it to conduct regular surveys, it seems to be doing well at it.
by Sentimenti Team | May 8, 2020 | Politics and Social
The presidential debate took place on the 6th of May in the TVP headquarters. First of all, it was the first debate before the presidential election with the full line-up, i.e. with all 10 participants – candidates for the head of state’s seat. On the basis of the publication placed on the portal polskatimes.pl – “2020 Presidential Debate: Who won? [SONDA] Ten candidates debated in TVP before the election [RELATION]” the Sentimenti team of analysts prepared an analysis of emotions that were evoked by the statements of the individual candidates. At the time of the analysis polskatimes.pl was the only such extensive source presenting the debate.
2020 Presidential Debate – Confidence and Fear
The analysis was done for the two most important opposing emotions, trust and fear. The intensities of these emotions were examined for, respectively:
- the entire debate;
- for individual 5 rounds of questions;
- for all 10 presidential candidates.
Results of the analysis
The presidential debate as a whole – strong emotions of trust and fear in candidates’ statements
5 rounds with questions by candidate
Free speech by the candidates at the end of the debate
Politics is a game of emotions, and these were lacking
This time we will not present the conclusions, so we leave the presented results of the analysis to your own interpretation. However, it is worth mentioning that the format of the debate left much to be desired. From the point of view of emotions it was, above all, a spectacle for the eyes (we can not complain about the visual side). It’s hard to say anything good about the substantive layer, that is the one that shouldn’t and shouldn’t cause any emotional stimulation.
Politics is above all a game with emotions. One can risk a thesis that we did not deal with a fight for the future of Poland and its citizens but, at the most, with a more boring version of ‘One of Ten’. There was no real discussion, which would open any space for candidates to present themselves and their idea for presidency.
There was also no opportunity to show the strengths and weaknesses of the candidates. Each presented himself according to his abilities measured by few minutes of speaking time. To sum up – it is hard to say anything good about the show, during which the most stimulated person was the sign language interpreter in the corner of the screen, who probably had the opportunity to catch our attention more often.
