Effect of Informational Divergence on the Mental Health of the Population in Crisis Situations: A Study in COVID-19

Abstract

Informational divergence emerged as a significant phenomenon during the COVID-19 health crisis. This period was characterized by information overload and changes in the communication of public health recommendations and policies by authorities and media outlets. This study examines the impact of such divergence on the population’s mental health, focusing on primary emotions expressed in comments across digital ecosystems. A media EMIC approach was used to analyze digital ecosystems during March and April 2020. Data were collected from Twitter, YouTube, Instagram, official press websites, and internet forums, yielding 3,456,387 communications. These were filtered to extract emotion-expressing content, resulting in 106,261 communications. Communications were categorized into primary emotions (anger, disgust, joy, fear, and sadness) using an exclusionary emotion assignment procedure. Analysis techniques included polarity and term frequency calculation, content analysis using Natural Language Understanding, emotion intensity measurement using IBM Watson Analytics, and data reliability assessment using the ISMA-OWA operator. The findings suggest that exposure to informational divergence from governments, health organizations, and media negatively affected mental health, evidenced by sadness, fear, disgust, and anger, which are associated with elevated levels of stress, anxiety, and information fatigue. In contrast, information perceived as reflecting coordination, support, and solidarity elicited positive emotional responses, particularly joy.

1. Introduction

Communication is a fundamental pillar in crisis management [1,2], especially in healthcare, where the transmission of clear, consistent, and reliable information not only facilitates crisis management but also protects the psychological well-being of citizens, strengthens trust in institutions and promotes positive social behavior. The evolution of digital communication platforms has amplified both the reach and complexity of crisis communication, creating unprecedented challenges for information management during public health emergencies.

The COVID-19 crisis has highlighted the fundamental role of communication in managing health emergencies. This period was marked by an information overload [3] with frequent changes and contradictions in the public health recommendations and measures disseminated by authorities and the media. This phenomenon, known as information divergence, is characterized by the issuance of inconsistent or contradictory messages by key actors, such as governments, health agencies and the media, and by the frequent changes and contradictions in the recommendations and public health measures disseminated by authorities and the press [4,5,6,7]. Unlike mere misinformation or deliberate disinformation, information divergence represents a more nuanced challenge where even authoritative sources may present conflicting narratives, creating a complex information environment that citizens must navigate.

The complexity of information divergence during COVID-19 manifested in multiple forms. Healthcare authorities worldwide often presented conflicting recommendations about basic preventive measures, while statistical reporting methods varied significantly across regions, creating confusion about the pandemic’s true scope. Economic impact assessments and recovery forecasts also showed marked disparities between different institutional sources, contributing to public uncertainty about appropriate individual and collective responses to the crisis. These discrepancies in information can make people feel trapped in a situation with no clear solution and lead to feelings of hopelessness or helplessness, triggering or exacerbating depressive symptoms, especially in crises, such as the COVID-19 pandemic [8,9,10,11,12].

Recent studies indicate that prolonged exposure to information divergence generates in the population confusion, uncertainty, distrust in institutions and the media, information fatigue, difficulty in making informed decisions, and political polarization, factors that, in turn, affect the mental health of the population through primary emotions, such as sadness, fear, disgust and anger [13,14,15,16,17,18,19]. Thus, for example, the Observatory of Suicide in Spain of the Spanish Foundation for Suicide Prevention reported an increase in suicides during the pandemic year of 7.4%. According to the report, 3941 people died by suicide in Spain, an average of almost 11 people a day: 74% of them men (2938) and 26% women (1011). These figures represent an increase of 5.7% in men and an increase of 12.3% in women; in children under 15 years of age, the figures have doubled compared to previous years, and in those over 80 years of age, there has been an increase of 20% [20]. While the World Economic Forum has established addressing information divergence as a central priority in its policies [21] and in academia, there has been a growing interest in this issue, as reflected in scientific indicators [22].

This research aims to analyze the impact of informational divergence on the population’s mental health through the primary emotions expressed by citizens in digital ecosystems during COVID-19. Based on preliminary analysis of our data and existing literature, we propose the following hypotheses:

H1: 

Exposure to contradictory information from authoritative sources increases negative emotional responses (particularly fear and anger) in digital ecosystem communications.

H2: 

Information perceived as reflecting coordination, support, and solidarity among authorities elicits positive emotional responses, particularly expressions of joy and trust.

This empirical analysis employed a media EMIC approach to examine digital ecosystems during March and April 2020. From an initial sample of 3,456,387 communications, we extracted 106,261 emotion-expressing posts for detailed analysis.

The manuscript is organized as follows: the Section 2 presents the methodology used; the Section 3 presents the results obtained; and finally, the Section 4 presents a discussion and conclusions, ending with the future lines of work and the references used.

2. Materials and Methods

2.1. Methodological Framework

The study adopts a mixed-methods approach to examine the relationship between information divergence and emotional responses during the COVID-19 crisis. This framework integrates quantitative and qualitative analytical techniques, enabling systematic processing of heterogeneous data from digital ecosystems. Our selection of digital platforms as primary data sources is grounded in their demonstrated capacity to capture real-time public reactions to crisis information, providing a robust dataset for emotional response analysis [10]. The temporal focus on March–April 2020 corresponds to a period of peak information divergence in Spain’s COVID-19 response, offering an optimal window for examining the immediate emotional impact of contradictory communications.

The analytical framework combines three complementary approaches: sentiment analysis through natural language processing, emotion intensity measurement, and reliability assessment.

2.2. Data Collection and Analysis

This study collected data directly from the media and digital ecosystems (social media and mass media): Twitter, YouTube, Instagram, official press websites and internet forums. The data were extracted from the sources through their APIs or by using scraping techniques.

The study universe consisted of thirty-two million Spanish citizens, representing 68% of the Spanish population, who are considered users of social networks and the Internet according to data from the Spanish National Statistics Institute (INE) (forecast as of 1 July 2019) [23].

The criteria for inclusion of submissions were:

• Written communications from the Spanish government on the government’s actions with COVID-19.
• News communications on government actions and citizens’ reactions.
• Public communications, accessible without subscription or explicit permission from the sender of the communication.
• Declared age of the author, when available, over 18 years old at the start of the study (1 March 2020).
• Comments written in Spanish language.

While the exclusion criteria were:

• The communication did not come from governmental COVID-ACTION or any other news communication about governmental actions related to COVID-19 or was not a reaction from citizens to this type of communication.
• The communication was generated by automatic procedure methods (bots, spam, fake accounts, …).

In addition, authorship data, source, hyperlinks, location, images, and non-textual components were removed for anonymization purposes.

Data collection began on 1 March and ended on 30 April 2020. During this time, 2,567,824 communications were collected and categorized into the primary emotions: anger, disgust, joy, fear, and sadness, using an exclusionary emotion assignment procedure. Communications with more than two emotions, e.g., anger and disgust, were classified into the emotion with the highest degree of membership. The total number of communications in each feeling was as follows: 417,520 for anger, 379,657 for disgust, 550,170 for joy, 3,550,200 for fear, and 866,271 for sadness.

The analysis of the communications was conducted using the following techniques:

Polarity and frequencies of terms in communications. The Natural Language Toolkit (NLTK) [24] in Python 3 was used to calculate the polarity and frequency of the words that make up the communications sample. NLTK implements a multilayer perceptron neural network, which allows for measuring the emotional polarity of the communications: positive, neutral, or negative [25,26].

Content analysis using Natural Language Understanding allows inferences to be drawn from specific features identified in messages [27]. This type of analysis makes it possible to detect trends and reveal differences in content communication. In addition, it facilitates the evaluation of messages and the identification of intentions and appeals.

Type and intensity of emotions. IBM Watson Analytics measured the emotional intensity of each communication’s five primary emotions—anger, fear, joy, disgust, and sadness Emotion intensity is measured on a scale of 0 to 1, where 0 represents the total absence of an emotion, and 1 illustrates a high intensity of the feeling.

Reliability of the data for the calculation of emotion. The Majority Interval Aggregation of the Majority (ISMA-OWA) operator [28] was used to determine the consistency of the data [29].

Finally, the visualization of the results makes use of so-called concept clouds as a replacement for word clouds and as an improvement on word clouds, as they allow the topics of conversations to be determined more accurately by grouping the relevant concepts into a single term instead of simply counting the occurrence of words. In addition, concept clouds allow the emotion of conversations to be represented through colors. This work describes government communications, such as mass media communications, without emotion, while clouds corresponding to citizens show emotion.

3. Results

3.1. Anayisis of the Tone of Government Communications and the Emotional Responses of Citizens

The data analysis shows the main topics arising from communications emitted by the Spanish government and the response from mass media and social media. In addition, we show the concept clouds with issues expressed in the media by the citizens and the emotions associated with these topics.

The first two concept clouds present the topics obtained from official communications issued directly by the government or given as news articles, press releases and other mass media communications expressing information. Therefore, these two concept clouds do not consider emotions. The third and fourth concept clouds show the topics perceived by the general population as described in digital ecosystems, including the aggregate sentiment of the communications for each topic. Afterwards, a second analysis shows the main themes by date and the primary emotions elicited from them: anger, fear, joy, disgust, and sadness, expressed in communications emitted by the public.

To analyze the different topics, concept clouds were used, showing in detail the main topics of conversation about managing the COVID-19 crisis by the Spanish Government. Figure 1 shows the concept cloud with the main issues extracted from communications directly emitted by the Government of Spain that generated any form of engagement on social media and mass media.

Official communications are exempt from sentiment analysis, as this kind of communication usually presents information on decisions adopted, new bills and regulations passed, purchases performed by public entities, announcements, death counts, and other information intended as statements expressing crude facts instead of opinions. Therefore, the concept cloud associated with these types of communications presents these topics but does not include sentiment values.

In Figure 2, we show the topics contradicting statements issued by the government. These topics are obtained from reputable sources, such as the Spanish Government (when contradicting itself) or news sites, that provide information. Since extracting this information is the same as in Figure 1, we also obtain topics without sentiment values for the concept cloud presented in Figure 2.

By comparing both figures, we can show the discrepancies between the information issued by the government in the first months of the pandemic. In both cases, we consider information announced by the Spanish government that was subsequently challenged by the media, rectified by the Spanish government, or admitted by the judiciary.

Lastly, the concept clouds in Figure 3 and Figure 4 show the topics related to communications directly emitted by citizens in the form of opinions about the conflicting information presented above. Figure 3 shows issues associated with negative sentiments, and Figure 4 shows topics with positive feelings. It is noteworthy that positivity has been focused mainly on health workers, business people and the army, while negativity includes more government political management actions.

3.2. Emotions by Topic

In a second study, we analyzed the emotions of the communications issued by the population about the main topics detected in the previous step, assessing the primary emotions of anger, fear, joy, disgust, and sadness. In Figure 5, we show the total communication and emotional distribution of the communications issued by the public on these topics. In this figure, we highlight that the most significant number of communications are identified with the feeling of sadness, with a total of 866,271, followed by joy with a total of 550,170 communications where we have detected that this emotion is present. These data suggest that sadness is an emotional response to the negative perception that citizens experienced in the context of adversities that arose in the period studied. The fact that the feeling of happiness occupies second place is related to those news events that allowed citizens to have a more optimistic view of the solution to the problems caused by the COVID-19 pandemic. Other emotions from the communications analyzed had somewhat lower figures than the previous ones, such as anger with 417,520 communications, disgust with 379,657 and fear with 354,206. With these figures, it can be established that citizens’ reactions during this period were also indignation, rejection, and fear, reflecting a complex emotional panorama that was analyzed during this period.

In addition, Table 1, Table 2, Table 3, Table 4 and Table 5 show the communications more closely related to each of these emotions.

3.3. Communications Associated to Primary Emotions

The analysis of emotional responses in digital communications during the COVID-19 crisis reveals distinct patterns of public sentiment over time. Figure 6 presents the daily evolution of the five primary emotions (anger, disgust, fear, joy, and sadness) from 1 March to 30 April 2020. To enable meaningful temporal comparison and control for varying communication volumes, the data are normalized as proportions of total daily communications. This normalization allows for the identification of emotional patterns independently of fluctuating communication frequencies. Each subsequent section examines a primary emotion in detail, analyzing its triggers, temporal patterns, and relationship to specific themes in government communications and media coverage. Table 1, Table 2, Table 3, Table 4 and Table 5, derived from this study’s data collection and analysis, present representative communications for each emotional response, while Figure 7, Figure 8, Figure 9, Figure 10 and Figure 11 visualize the most influential themes associated with each emotion, providing a comprehensive view of how informational divergence influenced public sentiment throughout the crisis period.

3.3.1. Disgust

Table 1 analyzes the communications during the study period from 03/01/2020 to 04/30/2020. Among the most prominent aspects are the discussions that pointed to specific groups responsible for the pandemic based on criteria of religion, gender or transmission of the virus and the controversies related to the acquisition of health material, which dominated the discourse in digital environments. Finally, Figure 7 illustrates the most influential topics associated with the emotion of disgust. This analysis highlights that this emotion is most prominent in everything related to coronavirus transmission, along with other issues that highlight the disgust among citizens, such as face masks, cases of coronavirus and fake news.

Table 1. Comments related to the emotion of disgust during the first stage of COVID-19 in digital ecosystems.

Emotion of Disgust Related to COVID-19 Comments in Digital Ecosystems
02/03/2020	The Ministry of Health accuses the Evangelical Church of Madrid of being the focus of COVID.
03/23/2020	Spanish toilets are very well protected against COVID. Getting infected with COVID depends on the gender. The 6M did not facilitate the spread of COVID; it is a gender problem. The government buys PCR tests without a guarantee. The government recommends using public health care, while ministers go to private hospitals. Ministers’ cabinets have access to COVID tests, while health workers and the population do not have access to them. Minister Iglesias does not respect the confinement he demands from citizens.
03/28/2020	The government buys defective PCR tests from Chinese companies. The government does not ask for help from businessmen with contacts in China. The government buys fake tests that private companies had rejected. The number of masks shipped is much lower than advertised. Members of the government use social media to speak well of the government. The PCR tests bought by the government do not work.
04/09/2020	Deaths in residences are very high, and they receive no help from the government. The government buys material without guarantee from companies that are not on the lists of reliable companies. The government’s death figures do not correspond to the actual deaths. The Spanish government boasts about the management of COVID, with a record number of deaths. The government of Catalonia does not want military hospitals even though it cannot attend to the citizens.
04/10/2020	The government boasts of management with a record number of deaths in Europe.
04/14/2020	The government buys equipment from ghost companies. The government forces companies to provide health material that cannot be bought, as purchases are centralized. Government vehicles are protected while there are no supplies in hospitals.
04/20/2020	Members of the government bypass confinement at events. The government changes laws to protect itself from possible complaints. The government uses the civil guard to monitor the anti-government climate over the handling of the pandemic.
04/23/2020	The government controls freedom of speech in mass media and social media. The ministry buys supplies at a higher price than private companies. They force people to return to work without means of protection. They hire opaque companies to provide sanitary equipment.
04/28/2020	The government admits it lied about the COVID tests. The official death count is false. The government uses police and confinement to control citizen complaints.

Table 1. Comments related to the emotion of disgust during the first stage of COVID-19 in digital ecosystems.

Emotion of Disgust Related to COVID-19 Comments in Digital Ecosystems
02/03/2020	The Ministry of Health accuses the Evangelical Church of Madrid of being the focus of COVID.
03/23/2020	Spanish toilets are very well protected against COVID. Getting infected with COVID depends on the gender. The 6M did not facilitate the spread of COVID; it is a gender problem. The government buys PCR tests without a guarantee. The government recommends using public health care, while ministers go to private hospitals. Ministers’ cabinets have access to COVID tests, while health workers and the population do not have access to them. Minister Iglesias does not respect the confinement he demands from citizens.
03/28/2020	The government buys defective PCR tests from Chinese companies. The government does not ask for help from businessmen with contacts in China. The government buys fake tests that private companies had rejected. The number of masks shipped is much lower than advertised. Members of the government use social media to speak well of the government. The PCR tests bought by the government do not work.
04/09/2020	Deaths in residences are very high, and they receive no help from the government. The government buys material without guarantee from companies that are not on the lists of reliable companies. The government’s death figures do not correspond to the actual deaths. The Spanish government boasts about the management of COVID, with a record number of deaths. The government of Catalonia does not want military hospitals even though it cannot attend to the citizens.
04/10/2020	The government boasts of management with a record number of deaths in Europe.
04/14/2020	The government buys equipment from ghost companies. The government forces companies to provide health material that cannot be bought, as purchases are centralized. Government vehicles are protected while there are no supplies in hospitals.
04/20/2020	Members of the government bypass confinement at events. The government changes laws to protect itself from possible complaints. The government uses the civil guard to monitor the anti-government climate over the handling of the pandemic.
04/23/2020	The government controls freedom of speech in mass media and social media. The ministry buys supplies at a higher price than private companies. They force people to return to work without means of protection. They hire opaque companies to provide sanitary equipment.
04/28/2020	The government admits it lied about the COVID tests. The official death count is false. The government uses police and confinement to control citizen complaints.

Figure 7. Themes related to COVID-19 and the emotion of disgust that have been most impactful along with their impact value.

Figure 7. Themes related to COVID-19 and the emotion of disgust that have been most impactful along with their impact value.

3.3.2. Fear

In Table 2, the communications that had the most significant impact on this temporal progression from 03/01/2020 until 04/30/2020 are shown in relation to COVID-19. Of these communications, the increase in infections, the deaths caused by the pandemic in Spain, the overflow of the health system, and the economic and political problems caused by COVID-19 are the general themes that have the most significant impact on the emotions addressed. Finally, Figure 7 shows the most relevant topics and their impact value on the emotion of fear. The issues most related to this emotion are an outbreak of coronavirus and emergency material, followed in order of least to most importance by the coronavirus crisis, state of alarm, deaths of family members, and coronavirus transmission.

Table 2. Comments related to the emotion of fear during the first stage of COVID-19 in digital ecosystems.

Emotion of Fear Related to COVID-19 Comments in Digital Ecosystems
03/06/2020	Unemployment increases.
03/18/2020	The government closes the borders. Many older people die in residences because of COVID.
03/25/2020	Hospitals are overwhelmed by the virus. Spain has more deaths than China from the virus. The economy is collapsing.
03/27/2020	COVID reaches pandemic level. Government management is overwhelmed by the virus.
04/26/2020	The government limits citizens’ freedom through censorship. Defective sanitary equipment.

Table 2. Comments related to the emotion of fear during the first stage of COVID-19 in digital ecosystems.

Emotion of Fear Related to COVID-19 Comments in Digital Ecosystems
03/06/2020	Unemployment increases.
03/18/2020	The government closes the borders. Many older people die in residences because of COVID.
03/25/2020	Hospitals are overwhelmed by the virus. Spain has more deaths than China from the virus. The economy is collapsing.
03/27/2020	COVID reaches pandemic level. Government management is overwhelmed by the virus.
04/26/2020	The government limits citizens’ freedom through censorship. Defective sanitary equipment.

Figure 8. Themes related to COVID-19 and the emotion of fear that have been most impactful along with their impact value.

Figure 8. Themes related to COVID-19 and the emotion of fear that have been most impactful along with their impact value.

3.3.3. Anger

Table 3 analyzes the communications during the study period from 03/01/2020 to 04/30/2020. From this analysis, the predominant themes concerning anger are loss of employment, political discrepancies and problems in acquiring health supplies. Aspects such as the collapse of the healthcare system, the differences in the data on the number of infected and deceased, and the control of the media suggested by members of the Spanish government are the predominant conversations in the digital ecosystems.

Finally, Figure 11 shows the most relevant topics and their impact on the emotion of anger. It shows that the topic that most influences this emotion is the interest in coronavirus deaths. The most recurrent topics, from lowest to highest, are coronavirus cases, the increase of coronaviruses and resources for their possible cure.

Table 3. Comments related to the emotion of anger during the first stage of COVID-19 in digital ecosystems.

Emotion of Anger related to COVID-19 Comments in Digital Ecosystems
03/04/2020	COVID-19 test phones do not work. The evangelical community is angry at being accused of spreading the virus. In February, the first COVID-19 deaths occur in Spain, and the government hides it.
03/15/2020	The government declares a state of emergency. The number of people infected by the 8 March (8 M) demonstrations soars to 28,000, and there are more than 800 deaths. Disputes between the prime minister and the vice-president hamper the second anti-pandemic plan. Spain exported medical supplies and PCR tests, knowing that there were excesses.
03/17/2020	The government delayed the health alert for 8 M, although there were more than 10,000 cases of COVID-19. Unemployment rises by more than 100,000 people.
03/20/2020	Hospitals prioritize the sick for lack of supplies. Politicians do not respect confinement and attend private meetings. The public health service is collapsing.
03/29/2020	The government takes no action on the pandemic. Government compares faulty equipment and bans purchases from companies and autonomous communities.
04/19/2020	China provides erroneous information on the pandemic. The government’s death toll is not credible. Deaths of elderly people in residential homes are very high.
04/21/2020	Lack of food in senior citizens’ homes. The government restricts freedom of speech using security forces.
04/23/2020	The socialist party says that if parents want to walk their children, a dog is understood. The president uses the state of alarm to cut benefits to officials. The government forces a return to work without providing protective sanitary supplies. The faulty supplies purchased by the government for COVID-19 were produced by a Chinese company at double the market price. The coronavirus was present in Spain in the month of February.

Table 3. Comments related to the emotion of anger during the first stage of COVID-19 in digital ecosystems.

Emotion of Anger related to COVID-19 Comments in Digital Ecosystems
03/04/2020	COVID-19 test phones do not work. The evangelical community is angry at being accused of spreading the virus. In February, the first COVID-19 deaths occur in Spain, and the government hides it.
03/15/2020	The government declares a state of emergency. The number of people infected by the 8 March (8 M) demonstrations soars to 28,000, and there are more than 800 deaths. Disputes between the prime minister and the vice-president hamper the second anti-pandemic plan. Spain exported medical supplies and PCR tests, knowing that there were excesses.
03/17/2020	The government delayed the health alert for 8 M, although there were more than 10,000 cases of COVID-19. Unemployment rises by more than 100,000 people.
03/20/2020	Hospitals prioritize the sick for lack of supplies. Politicians do not respect confinement and attend private meetings. The public health service is collapsing.
03/29/2020	The government takes no action on the pandemic. Government compares faulty equipment and bans purchases from companies and autonomous communities.
04/19/2020	China provides erroneous information on the pandemic. The government’s death toll is not credible. Deaths of elderly people in residential homes are very high.
04/21/2020	Lack of food in senior citizens’ homes. The government restricts freedom of speech using security forces.
04/23/2020	The socialist party says that if parents want to walk their children, a dog is understood. The president uses the state of alarm to cut benefits to officials. The government forces a return to work without providing protective sanitary supplies. The faulty supplies purchased by the government for COVID-19 were produced by a Chinese company at double the market price. The coronavirus was present in Spain in the month of February.

Figure 9. Themes related to COVID-19 and the emotion of anger that have been most impactful along with their impact value.

Figure 9. Themes related to COVID-19 and the emotion of anger that have been most impactful along with their impact value.

3.3.4. Sadness

Table 4 analyses the communications during the study period from 03/01/2020 to 04/30/2020. The topics derived from the communications studied that have the most significant impact on the emotion of sadness are those that have to do with political interests, bans on the purchase of health supplies in particular, the use of security forces to control citizens, or the increase in deaths in the elderly; these are some of the conversations that predominated in the digital ecosystems.

Figure 9 shows the most relevant topics and their impact value on the emotion of sadness. The first ones that stand out the most are those related to coronavirus victims, followed by patients. Other relevant topics are elderly, infected workers, death and COVID-19 infection.

Table 4. Comments related to the emotion of sadness during the first stage of COVID-19 in digital ecosystems.

Emotion of Sadness Related to COVID-19 Comments in Digital Ecosystems
03/06/2020	The Vice President of the Government of Spain threatens journalists with prison for reporting on the COVID-19 pandemic. The government hides COVID-19 death figures.
03/15/2020	The president of the government paralyzes aid to companies and workers due to COVID-19. The government that denied the pandemic declares a state of emergency. Sanitary supplies are exported when there is a shortage in Spain.
03/26/2020	The central government cannot buy sanitary supplies for COVID-19 and prohibits the autonomous communities from doing so. The purchased sanitary supplies made of porcelain are defective. Deaths in nursing homes grow. Layoffs in companies increase due to COVID-19. The government had information about the infections at the end of February and supported the large March 8 demonstration for political reasons. The number of infections in Spain exceeds 56,188 and the missing 4000.
04/01/2020	The government puts political interests before the interests of citizens. The government of Catalonia asks health workers whether it is better for people to die at home than in the hospital. The number of deaths from COVID-19 continues to rise.
04/11/2020	The government withdraws the experts on the management of the pandemic and replaces them with politicians. Deaths are increasing due to lack of means of protection. The government hides the figures of the deceased.
04/13/2020	The government does not have the capacity to provide protective equipment to companies and at the same time prohibits them from searching for them on their own.
04/22/2020	The government provides faulty protective equipment. The government uses the police to censor and monitor citizens.
04/29/2020	Mass job loss. The political interests of the government prevail over health care during the pandemic. The death of elderly people in residences increases massively. The government buys defective sanitary equipment.

Table 4. Comments related to the emotion of sadness during the first stage of COVID-19 in digital ecosystems.

Emotion of Sadness Related to COVID-19 Comments in Digital Ecosystems
03/06/2020	The Vice President of the Government of Spain threatens journalists with prison for reporting on the COVID-19 pandemic. The government hides COVID-19 death figures.
03/15/2020	The president of the government paralyzes aid to companies and workers due to COVID-19. The government that denied the pandemic declares a state of emergency. Sanitary supplies are exported when there is a shortage in Spain.
03/26/2020	The central government cannot buy sanitary supplies for COVID-19 and prohibits the autonomous communities from doing so. The purchased sanitary supplies made of porcelain are defective. Deaths in nursing homes grow. Layoffs in companies increase due to COVID-19. The government had information about the infections at the end of February and supported the large March 8 demonstration for political reasons. The number of infections in Spain exceeds 56,188 and the missing 4000.
04/01/2020	The government puts political interests before the interests of citizens. The government of Catalonia asks health workers whether it is better for people to die at home than in the hospital. The number of deaths from COVID-19 continues to rise.
04/11/2020	The government withdraws the experts on the management of the pandemic and replaces them with politicians. Deaths are increasing due to lack of means of protection. The government hides the figures of the deceased.
04/13/2020	The government does not have the capacity to provide protective equipment to companies and at the same time prohibits them from searching for them on their own.
04/22/2020	The government provides faulty protective equipment. The government uses the police to censor and monitor citizens.
04/29/2020	Mass job loss. The political interests of the government prevail over health care during the pandemic. The death of elderly people in residences increases massively. The government buys defective sanitary equipment.

Figure 10. Themes related to COVID-19 and the emotion of sadness that have been most impactful along with their impact value.

Figure 10. Themes related to COVID-19 and the emotion of sadness that have been most impactful along with their impact value.

3.3.5. Joy

Table 5 analyzes the communications during the study period from 03/01/2020 to 04/30/2020. Of the communications analyzed, the main topics related to the emotion of joy are those related to the resolutions of the Government of Spain to deal with the pandemic, the opening of new hospitals, economic aid, businessmen and the participation of the army to help the population are some of the conversations that predominated in the digital ecosystems. Figure 10 indicates which related topics had a greater impact on the emotion of joy. In our analysis, we can observe that this emotion has a higher index in topics related to workers’ health. It is followed, in order from least to most important, by European Union aid, employer aid, and the role played by the Spanish army during this pandemic.

Table 5. Comments related to the emotion of joy during the first stage of COVID-19 in digital ecosystems.

Emotion of Joy Related to COVID-19 Comments in Digital Ecosystems
03/05/2020	The government reassures us that COVID-19 will not have an impact on unemployment. Italy closes schools and universities because of the virus.
03/08/2020	The government encourages people to attend the 8 M street march. 8 March street march in Madrid: “I’d rather be killed by the coronavirus than by sexism”.
03/11/2020	Europe pledges 25 billion against the COVID-19 storm. The EU activates an immediate 7.5 billion shock plan. Deaths and infections in China continue to fall. “The supply of food to the population is guaranteed”, Juan Roig, CEO of Mercadona. Preparations begin for field hospitals.
03/16/2020	Around 350 troops of the Military Emergency Unit (UME) are deployed in some Spanish cities. The parachute brigade is mobilized to help. Central banks announce coordinated action to ensure liquidity. A recruitment drive is announced. Massive recruitment of resident doctors to deal with the coronavirus crisis.
03/21/2020	Madrid provides seven additional hotels for coronavirus patients. Communities agree to receive assistance from the army (Basque Community). Minimal services for the accommodation of transporters, etc. Health supplies donated by companies. Inditex stands out. Applause for health workers. Construction of IFEMA hospital.
03/28/2020	Congratulations to Amancio Ortega, CEO of Inditex for the aid. Mercadona assures there will be no shortages. The army helps with the dead. Management pay cut. Sacking forbidden.
04/04/2020	European aid fund for Spain and Italy. Manufacture of respirators by SEAT. The executive is denounced by civil servants for reckless homicide. The EMU disinfects residences. Medicalized hotels. The King appears with the army. Companies help with COVID-19. Inditex stands out.
04/17/2020	On this day, the self-employed who asked for cessation of activity due to the virus begin to receive payment. Zara’s sanitary gowns arrive in hospitals. More masks arrive. Companies manufacture respirators (Nissan), gowns, etc. Inditex does not apply ERTE. National companies manufacture masks.
04/19/2020	Coronavirus deaths fall to 410, the lowest number since 22 March. IFEMA Trade Fair and Conference Centre, Madrid announced as no longer necessary. Antivirals work with COVID-19. Children’s outings.
04/25/2020	The number of people cured of the disease exceeds the number infected. The de-escalation begins, going outside for sports.

Table 5. Comments related to the emotion of joy during the first stage of COVID-19 in digital ecosystems.

Emotion of Joy Related to COVID-19 Comments in Digital Ecosystems
03/05/2020	The government reassures us that COVID-19 will not have an impact on unemployment. Italy closes schools and universities because of the virus.
03/08/2020	The government encourages people to attend the 8 M street march. 8 March street march in Madrid: “I’d rather be killed by the coronavirus than by sexism”.
03/11/2020	Europe pledges 25 billion against the COVID-19 storm. The EU activates an immediate 7.5 billion shock plan. Deaths and infections in China continue to fall. “The supply of food to the population is guaranteed”, Juan Roig, CEO of Mercadona. Preparations begin for field hospitals.
03/16/2020	Around 350 troops of the Military Emergency Unit (UME) are deployed in some Spanish cities. The parachute brigade is mobilized to help. Central banks announce coordinated action to ensure liquidity. A recruitment drive is announced. Massive recruitment of resident doctors to deal with the coronavirus crisis.
03/21/2020	Madrid provides seven additional hotels for coronavirus patients. Communities agree to receive assistance from the army (Basque Community). Minimal services for the accommodation of transporters, etc. Health supplies donated by companies. Inditex stands out. Applause for health workers. Construction of IFEMA hospital.
03/28/2020	Congratulations to Amancio Ortega, CEO of Inditex for the aid. Mercadona assures there will be no shortages. The army helps with the dead. Management pay cut. Sacking forbidden.
04/04/2020	European aid fund for Spain and Italy. Manufacture of respirators by SEAT. The executive is denounced by civil servants for reckless homicide. The EMU disinfects residences. Medicalized hotels. The King appears with the army. Companies help with COVID-19. Inditex stands out.
04/17/2020	On this day, the self-employed who asked for cessation of activity due to the virus begin to receive payment. Zara’s sanitary gowns arrive in hospitals. More masks arrive. Companies manufacture respirators (Nissan), gowns, etc. Inditex does not apply ERTE. National companies manufacture masks.
04/19/2020	Coronavirus deaths fall to 410, the lowest number since 22 March. IFEMA Trade Fair and Conference Centre, Madrid announced as no longer necessary. Antivirals work with COVID-19. Children’s outings.
04/25/2020	The number of people cured of the disease exceeds the number infected. The de-escalation begins, going outside for sports.

Figure 11. Themes related to COVID-19 and the emotion of joy that have been most impactful along with their impact values.

Figure 11. Themes related to COVID-19 and the emotion of joy that have been most impactful along with their impact values.

4. Discussion and Conclusions

Most previous studies on the impact of information on the population’s mental health [30,31,32,33] have focused on disinformation, understood as the deliberate dissemination of false or misleading information intended to manipulate, confuse or influence public perception. In contrast, this study focused on informational divergence, defined as the coexistence of different—and even contradictory—narratives or versions of the same event or issue generated by reliable sources, such as the media and government institutions. Although information divergence is not necessarily intentional, it shares with disinformation the capacity to negatively influence the perception of reality and decision-making processes at the societal level.

The analysis conducted during the COVID-19 health crisis allowed us to explore how information overload and constant changes in official recommendations affected people’s emotions. The results show that the primary emotions experienced significant variations over time:

Sadness: remained relatively constant, linked to news about pandemic victims, especially older people.

Disgust: associated with news stories about lack of transparency, such as unclear data on contagion and the procurement of faulty supplies.

Fear: related to mismanagement, insufficient emergency equipment and restrictions on individual freedoms, such as declaring states of alarm.

Anger: triggered by news of the effects of mismanagement, such as increased deaths, shortages of health resources and unfavorable comparisons with other countries.

Joy: linked to identifying positive role models in civil society, such as health workers, the armed forces and well-known entrepreneurs.

Analysis of the digital ecosystem communications during the COVID-19 crisis validates our research hypotheses regarding emotional responses to information divergence. Regarding H1, our findings demonstrate that contradictory messages from authorities consistently triggered negative emotional responses. When health organizations and government bodies issued conflicting guidelines about preventive measures or presented contradictory statistics, public communications showed predominant patterns of fear and anger in digital ecosystems.

For H2, results confirm that coordinated communication efforts positively influenced public emotional responses. When authorities presented unified messages about health measures or economic support, communications expressing joy and trust notably increased. This effect was particularly evident in responses to coordinated announcements from healthcare institutions and in public reactions to unified crisis management strategies.

The research study also highlights that, in crisis contexts, when the population perceives divergence in the information provided by the authorities, they tend to seek reassurance from alternative social actors with high reputations. This fact reinforces the importance of ensuring clear, consistent and aligned public communication between different institutions.

In practical terms, these findings underline the need to implement unified and reliable communication strategies during crises. Furthermore, they support EU recommendations on developing media literacy policies [34,35] to strengthen citizens’ capacity to manage and evaluate information in critical contexts. Such actions could mitigate the adverse effects of information divergence on mental health and improve social resilience in future crises.

The scope of the research encompasses specific temporal and geographical boundaries. Focusing on Spain during March–April 2020 provided concentrated data but limits generalizability. Digital ecosystem analysis excludes non-digital communications and may underrepresent certain demographic groups, particularly older populations and those with limited internet access. Language constraints (Spanish-only) may miss nuanced emotional expressions in regional languages. Data collection faced technical challenges including API rate limits and potential sampling biases in social media platforms.

Future investigations should explore cross-cultural variations in emotional responses to information divergence. Longitudinal studies could track long-term mental health impacts of sustained exposure to contradictory information. Integration of alternative data sources, including traditional media and official health records, could provide more comprehensive understanding. Development of predictive models for emotional responses to information patterns could enhance crisis communication strategies. Research into effective coordination mechanisms between government agencies could improve information consistency during future crises.