Have you heard that the most interesting news scares all of us?

Introduction
This study aims to determine whether there is any correlation between the popularity of different news headlines by looking at what types of headlines are eye-catching to the readers. Since most newspapers are available online or on platforms such as Twitter, it’s fairly easy for readers to share information by retweeting.

In this study, we took arbitrary Twitter accounts of two newspapers for analysis - The Guardian (@guardian) and The Huffington post (@HuffingtonPost). Unique headlines for the period from 1 September 2015 to 28 February 2017 were collected and analyzed using the IntenCheck text analysis engine.

The headlines were divided into two groups based on the number of retweets: greater or equal to 100 - were deemed Popular (≧100) and those with less than 100 were deemed Unpopular. A specific template was created from the headlines for each of the news sources (publishers) from the Unpopular group of retweets (<100).

Then, from this template, a language norm (a baseline) was calculated using IntenCheck tone and sentiment analysis categories.

Popular headlines (retweets ≧100) were evaluated with respect to the specific baselines for each of the newspapers.

Emotions
The results of the Emotions category clearly showed us tendencies that are actively used in the media. News with strongly pronounced negative emotions such as Anger, Fear, and Sadness attract most of the attention. Positive and Joyful news or headlines attract little attention and are not shared as much as the negative ones. Once again, this demonstrates that the media not only informs but also forms the mood of the public. This creates a very strong unconscious inner tension in people who read this type of news.

Attitude (Semantic Differential)
Similar trends are shown in the results of the analysis of the Attitude or semantic differential. The Attitudes which are expressed most are Negative, Strong and Weak. Again, it can be observed that Negative emotions are more popular. The simultaneous expression of Strong and Weak characteristics is an indicator of internal conflict. It can be interpreted as some uncertainty, which creates the desire to share and discuss information with other people.

Communication style
Quite an unexpected result was shown in the analysis of the Communication channels. Most of all retweeted headlines use predominantly Auditory words (Who said what?). This is probably due to the fact that the readers of such headlines feel as though they are participants of a conversation which begins on the pages of newspapers. On the other hand, perhaps these are people with a pronounced Auditory representative system. This may be related to a new form of communication, as writing blogs or tweets is a continuation of the oral dialogue. It is an opportunity to meet the needs of any conversation with the help of short comments such as retweet blogs.

Timeline
The result of the analysis clearly shows that most of the headlines that attract attention and want to be shared reflect the events of the Present period of time and are focused on the Present tense.

Insincerity
The analysis showed that the most attention-grabbing headlines consist of some uncertainties which intrigue the readers. It is pronounced in the Insincerity category which indicates vagueness of the text. Thus, we can conclude that when writing headlines it is very important to create some sort of intrigue and uncertainty, which attracts most of the attention.

Perceptual positions
Interestingly, the results for the Perceptual Position-IV (we; ours) which is a “uniting” category were the strongest. On the one hand, this can be interpreted as an indication that events concerning a large group of people attract more attention than individual experiences Position-I (I, me). Considering the results of the Emotions and the Semantic Differential categories, the dominance of the 4th Perceptual Position support the tendencies of uncertainty and tension in readers which leads to their desire to share information and to find support in other like minded people.

Conclusion
In this case study, a collection of 18 months of headlines from two well-read newspapers were analyzed using the unique IntenCheck text analysis engine. The results showed that most shared headlines contain Negative emotions such as Anger, Fear, and Sadness, and carry some uncertainty (Insincerity). Present time headlines with the dominant unifying IV Perceptual Position (we; our) are most popular.

Choosing the best headline text for your article can be hard. This is why instead of just guessing, the IntenCheck system can provide scientific insight into effective patterns that may help produce the most attractive headlines for your readers.