Introduction

In 2016 the multiplatform of news and information, Pictoline, broadcast in Spanish a GIF (Graphics Interchange Format) titled Are you inside an echo chamber? In which the following was mentioned:

On the Internet we tend to follow people and media who share our vision of the reality. Invariably, we end up sharing and consuming content that reinforce that vision. To retain your attention, social networks like Facebook are optimized to show you only what interests you and leave out the “irrelevant”. This mixture of own selection with algorithms could leave you inside an echo chamber where you believe that the world is as you imagine it… and the other visions are a minority or simply do not exist, but nothing further from reality (^{¿Estás dentro de una cámara de eco?, 2016}).

Pictoline’s campaign took place within the framework of the 2016 us electoral process, as the election was polarized between the Democratic and Republican parties. In that sense, the electoral process and mainly the victory of the Republican candidate allowed for the implications of the echo chamber² to be problematized and analyzed publicly. To cite a few examples, in May 2016 The Wall Street Journal published an interactive page called “Blue Feed, Red Feed: See Liberal Facebook and Conservative Facebook, Side by Side” (^{Keegan, 2016}), which showed the contrast between the news circulating on Facebook to indicate that American society was polarized.

Subsequently, after the election, the newspaper Independent published a column entitled “Social media echo chamber gifted Donald Trump the presidency” (^{Hooton, 2016}), in which it is mentioned that polarization between Conservatives and Republicans led to the triumph of Donald Trump. In the same vein, days later, the technology magazine Wired published the text “Your Filter Bubble is Destroying Democracy” (^{Mostafa, 2016}), which mentions that the social bubbles of Facebook and Google are designed for people to observe a specific reality about the United States.

As a whole, the above vision is anchored in the idea that people often interact with others who are related to their interests, which is known as homophily.³ Moreover, due to algorithmic mediation (^{Morozov, 2015}), the echo chamber tends to strengthen as the algorithms used by social networks such as Facebook, Twitter or Instagram start suggesting friendships, places, newspapers or products of any kind, to mention a few examples, based on the interests of each Internet user.

In this context, it seems impossible or very difficult for people to change their way of thinking or their preferences, known as frames of reference. This means that the idea of the echo chamber as a closed system is strengthened in public discourse. This hegemonic idea is also expressed in recent studies analyzing echo chambers.

In 2016, the article “Echo Chambers on Facebook” (^{Quattrociocchi, Scala & Sunstein, 2016}) was published, in which the question is whether echo chambers really exist. The study analyzed how two groups of Facebook users, in the United States and Italy, reacted to the content of webpages on conspiracy theories and scientific information. The findings indicated that, for scientific information and conspiracy theories, the more active a user is within an echo chamber the more they will interact with other people of similar beliefs. In this sense, the tendency of the users of these webpages is to promote their favorite narratives and, therefore, to form polarized groups.

The idea was reinforced in the paper “Anatomy of news consumption on Facebook” (^{Schmidt et al., 2017}). It looks at how news media publications are consumed and how patterns of user activity emerge. The interaction of 376 million users of 920 news sites that make up the European Media Monitor was analyzed during the period from 2010 to 2015, and from a cluster analysis the study pointed out that, despite the wide availability of heterogeneous content and narratives, there is great segregation and growing polarization in news consumption (^{Schmidt et al., 2017}).

However, from a critical perspective and exhaustive review of previous research that constitutes a hegemonic view on echo chamber, in “’The Echo Chamber’: Distraction: Disinformation Campaigns are the Problem, Not Audience Fragmentation”⁴ (^{Garrett, 2017}), the following is argued:

Numerous analyses of large scale observational data indicate that online news consumers do not systematically avoid exposure to content with which they would be expected to disagree…, the notion that people have constructed highly polarized online news environments, environments in which they never see the other side, is a myth (^{Garret, 2017, pp. 370-371}).

Likewise, the text points out that beyond being inside an echo chamber, the constant disinformation campaigns are what cause that certain ideas have greater weight in public discussions, since they respond to political objectives. Garret’s contrast makes it possible to nuance the general idea, both in public discourse and in the social sciences, about the impenetrability of the echo chambers. Furthermore, it would be an analytical error to think that the strength of these cameras operates equally for all people and they are not susceptible to sociodemographic variables such as age, education, income, access to the Internet or to different social and political contexts. Without forgetting the absence of research showing social interactions at the micro level from a qualitative perspective.

In that sense, the purpose of this research is to show for the Mexican case that echo chambers in an online environment and in an electoral political context are not hermetic. Therefore, it is possible to change people’s frames of reference and electoral preferences, since echo chambers have multiple cracks that make them flexible. The investigation will be placed in the 2015 electoral process in Mexico, based on three case studies: the candidacies of Jaime Rodríguez “El Bronco”, Enrique Alfaro and Pedro Kumamoto. The question that will frame the discussion is the following: What are and how are the mechanisms of social interaction mediated by Facebook that influence the change of electoral preferences?

In order to answer this question, a conceptual framework on the interaction mediated by social networks will be developed, which will be called sociodigital interaction. Subsequently, through a series of stories from a corpus of interviews, it will be argued that Primary Group and Weak Ties are the mechanisms of sociodigital interaction that influence electoral preferences. Finally, a diagram will be presented to know how these mechanisms are articulated and the final reflections will be presented.

Methodology

This research arises from a doctoral thesis and is framed in the qualitative tradition. The selection of the three case studies was not random but oriented towards the information they provide to the research phenomenon, since “atypical or extreme cases tend to reveal more information because they activate more actors and more basic mechanisms in the situation being studied” (^{Fyvberg, 2004, p. 45}). In this sense, research on the digital phenomenon and mainly journalistic sources have indicated that in the cases of Jaime Rodríguez, Enrique Alfaro and Pedro Kumamoto, the presence of social networks was central to their victory (^{Atilano, 2016}; ^{Petersen, 2015}). In addition, the selection of these cases will make it possible to distinguish the cracks of the echo chambers from different political-administrative divisions and subdivisions. That is to say: Federative Entity, Municipality and Local District.

People over the age of 18 who voted for any of the previous candidates are the unit of analysis, and from a corpus of 33 stories some will be taken to answer the research question. These testimonies come from the application of semi-structured interviews in situ, during the period July 2016 to March 2017 in the Mexican states of Jalisco and Nuevo León. The selection of interviewees was random and they were contacted using different Facebook groups.

Sociodigital interaction

Sociodigital interaction is dynamic and has become more complex with the development of the World Wide Web, the exponential growth of Internet users, broadband access and the consumption of new generations of mobile phones. From this scenario, the deployment of interaction in the digital space makes it possible to create contents and share those created by other users in a process called “mass selfcommunication” (^{Castells, 2012}). To define the concept of sociodigital interaction I take up again Goffman’s definition of interaction, but I modify its physical dimension.⁵ In this way, the sociodigital interaction is: one-way or reciprocal influence of one Internet user on the actions of another, or also the total interaction of Internet users, using technological and digital devices.

It is important to point out that although social interaction is in an online-offline continuum, empirically it is relevant to distinguish it from sociodigital interaction, since the latter has its own elements that only occur in a digital environment. For example, when the sociodigital interaction is exercised with the use of writing, such as in a post on Facebook, it is possible to edit any post or comment after writing it, which can change or nuance a certain message and, therefore, modify the meaning of the action. Moreover, in this type of interaction the message can be nourished from the material and symbolic resources available to the Internet user. That is to say, a Facebook user can strengthen a publication with information coming from some page and add a link, or it can contrast a piece that has published a newspaper with some video of YouTube or another type of evidence; among so many other possibilities. In that sense, ^{Scolari (2008)} points out that our ways of relating them are under a scheme of hypermediation, understood as “processes of exchange, production and symbolic consumption that develop in an environment characterized by a great quantity of subjects, media and languages technologically interconnected in a reticular way among themselves” (pp. 113-114).

Another peculiarity of the sociodigital interaction, which is not less important as it is gaining more and more strength in the debates around the social networks, arises from the intrinsic logics of the own platforms and their algorithms, which favors an algorithmic mediation (^{Morozov, 2015}). In this way, Facebook is not only one of the spaces where sociodigital interaction unfolds; its algorithms acquire an active role that influence the sociability process. For example, in the words of van Dijck: “Contacting a friend you have not seen since high school may be a thoroughly human act, but if performed online, a People You May Know algorithm typically prompts this deed” (2016, p. 163). Along these lines, ^{Bucher (2017)} points out that “Algorithms are not just abstract computational processes; they also have the power to enact material realities by shaping social life to various degrees” (p. 40).

Continuing with the sociodigital interaction, it enables social change based on what ^{Collins (1996)} has called “Interaction rituals chain”, which, as ^{Dettmer (2001)} points out, “produces feelings of solidarity and stratification among and within social coalitions, while recreating people’s cognitive beliefs about social structure” (p. 85).

In short, sociodigital interaction, from its definition, has the possibility of influencing the actions of another person individually or in a group without the need to be in physical presence. In addition, this interaction is in a framework of hypermediation and algorithmic mediation that makes the processes of sociability more complex, with which it is possible to reproduce and modify the social order.

That said, it will be shown below how the primary group and weak ties are mechanisms of sociodigital interaction that weaken the echo chambers and influence electoral preferences.

Cracks in the echo chamber through the primary group and weak ties

Facebook platform is a set of services to be in contact with other people and also to create communities, since it is a space where social networks converge: childhood friendships, school and work groups and close or distant relatives, to mention a few. The contacts that each person may have can be distinguished by the strength or weakness of the ties that unites them. In addition, the same political or ideological affinity may or may not exist among the networks of contacts. In this way, what each contact publishes does not necessarily reinforce the belief system of their contacts network.

On the other hand, in order for there to be changes in electoral preferences and turn them observable, it is not only necessary for the echo chambers to be flexible, but also for there to be frames of reference that activate and give meaning to the action. In other words, “The frame organizes more than meaning; it also organizes involvement. During any spate of activity, participants will ordinarily not only obtain a sense of what is going on but will also (in some degree) become spontaneously engrossed, caught up, enthralled” (^{Goffman, 2006, p. 345}). In this sense, sociodigital interaction has the capacity to influence electoral preferences since echo chambers are flexible. But how does that influence happen?

Two mechanisms will then be considered: Primary Group and Weak Ties, to show how this interaction has the capacity to influence in the light of the three case studies.

Primary groups as information shortcuts (remote and timeless)

The primary group is one of the most influential mechanisms in electoral behavior. This thesis comes from the sociological perspective of voting and the classical studies The People’s Choice (^{Lazarsfeld, Berelson & Gaude, 1944}) and Personal Influence (Lazarsfeld & Katz, 1955). However, with the development of the Web, the proposal on the primary group⁶ has been modified. Specifically because sociodigital interaction has disrupted the assumption of face-to-face interaction, in addition the creation of primary groups can be carried out at remote or timeless.

Likewise, the sociodigital interaction and the access that people have to multiple resources via the Internet has energized and broadened the information that enters the primary group. This is relevant to understand the influence that this group may have on the preferences and electoral behavior of each member, since “the response of an individual to a campaign cannot be accounted for without reference to his social environment and to the character of his interpersonal relations” (^{Lazarsfeld & Katz, 1979}). Moreover, because such groups, “actively influence and support most of an individual’s opinions, attitudes and actions” (^{Lazarsfeld & Katz, 1979, p. 53}). For example, the opinions and preferences that enter the primary group may arise from a Facebook discussion, form a news item, a meme, a video, o any discursive artifact.

My parents are pAn⁷ supporters and they don’t vote for someone who’s not a pAn member. Then with closest people [she refers to her family and friends] I did make sure that to vote for Kumamoto. In the social networks everything that Kumamoto shared, what seemed to me more interesting, it was what I shared on my wall and what I asked other people to do (Daniela, 26 years old, student).

I saw that my grandson was bringing Pedro’s [Kumamoto] propaganda to the house, caught my eye. Then I was with my grandson, we talked a lot, so I was convinced that Pedro was a decent person. Then Pedro came to the house and I saw that he was decent (Roberto, 94 years old, retiree).

These stories indicate that the primary group continues to have weight for the construction of electoral preferences and that young people provide political information to their group, mainly from social networks. For example, Roberto’s grandson found out about Pedro Kumamoto’s proposal on Facebook, then he had his grandfather meet the candidate and convince himself. Daniela’s case was similar; she found out about the independent candidate’s proposal via Facebook and later brought that information to her primary group.

Another point that stands out previous stories, which shows the strength of the primary group in the change in electoral preferences, is that both Daniela’s parents and Roberto had historically voted for the pAn (National Action Party). However, for the 2015 election and specifically for District 10 they decided to do it in favor of Pedro Kumamoto, that is to say, they voted for different political projects.

On the other hand, primary groups are not reduced to the family circle, they are also groups of friends, colleagues at work, school, or where opinions and attitudes are interpersonally shared, as ^{Lazarsfeld & Katz (1979)} point out. Thus, access to information held by these groups can be represented by discursive artifacts. For example, interpersonal opinions circulating in a group of friends may come in the form of a meme.

During elections the language of the meme is something that is always brought to the meets. Modalities from talking about the meme, describing it, even taking the cell phone out and showing it. Memes of all kinds. Through a very simple image, with a few words you understand the message and remember it enough to take it to a conversation (Carlos, 30 years old, editorialist).

Moreover, in an electoral context the information circulating in social networks is vast and constantly dynamic. Within the primary groups the circuit of information and interpretations can also become dynamic, since different narratives and interpretations circulate about candidates, jokes, anecdotes and the process itself in general.

People may overlap in different primary groups and these may have different sizes, and also various levels of cohesion. In other words, “primary groups are usually characterized by their small size, relative durability, informality, face-to-face contact and manifold, or more or less unspecialized purpose” (^{Lazarsfeld & Katz, 1979, p. 48}). Below are two cases of primary groups made up of friends.

I don’t follow “El Bronco” on Facebook, I found out because my friends shared publications about the campaign and they came to my timeline. I don’t even follow any news pages, so if I get any information it is because of my friends. In my house I only live with my mother and my younger siblings, and there is almost never any talk about politics (Daniel, 23 years old, student).

I heard about “El Bronco” on Facebook when he was mayor of Garcia. I voted for him and my mom voted for the PAN for everything. I don’t follow him, but I was attentive when something appeared in Facebook. Actually my friend convinced me, she posted all the time a lot of things from “El Bronco”. I even thought at some point of nullifying my vote (María Fernanda, 21 years old, student).

In the previous cases the information and opinions of the group of friends circulated mainly via Facebook. It is observed that the influence of the group of friends towards the people interviewed was superior to that exercised by the group that makes up the family. The influence generated by friendships may be due to a “need” to identify with the group. On this point, Lazarsfeld and Katz call it the “benefits” of conformity and mention the following:

If an individual desires to attain, or maintain, an intimate relationship with others, or if they want to ‘get somewhere’ either within a group or via a group, they must identify themselves with the opinions and values of these others (^{Lazarsfeld & Katz, 1979, pp. 57-58}).

When we focus on Daniel’s story, we see that his family did not talk about politics and that he did not follow any news pages or any candidate. However, the political opinions he received came from his friends, as did his electoral preference. This may respond to informational shortcuts that people use. On this point, ^{Unt, Solvak and Vassil (2017)} indicate:

Even if vote relevant information is not really discussed in the group and only the vote choice is shared, it should still have an impact given that the decision of a group member is a cognitive informational shortcut for other members (p. 3).

On the other hand, the meaning of the vote is not necessarily shared homogeneously with all the people who make up the group, since resistance may exist. Specifically when of the group does not agree with an idea that someone is trying to present.

My friends, who are also teachers, almost don’t like to read or meddle in thing of politics. They inform themselves very lightly, read a piece of news and do not go into it in depth. Instead, sometimes I read a news item, I see the information and I delve into the news to see and to compare those. Then I get my interpretation, but they don’t. I tell them: “look, look what this other story said” and they don’t listen, they just laugh, clinging to what they already read. Then I see that they are satisfied with what they read, and I say “okay, no problem” (Martha, 59 years old, retired teacher).

In the previous story it is shown that what is not coherent with the frame of reference of the group will be rejected and the pre-established ideas on some subject will be reinforced, since one of the functions of the group is to provide meanings to understand social phenomena. However, it should be noted that even if external ideas are rejected in a group, as in the previous story, some of its members may think differently. This again shows the porosity of the echo chamber.

In short, the primary group is one of the mechanisms that influence the construction of electoral preferences because identifying with the members of the group creates benefits, gives meaning to its members and operates as a cognitive shortcut to understand what is commonly known as politics. In addition, the primary group is not an atomic figure, it can articulate with other groups through the interaction of its members. This leads to a dynamic circuit of information and that each person is not inserted in a single primary group. In this sense, the fact that people are in different groups and that there is a constant input of information and opinions, many of them coming from sociodigital interaction, favors that the echo chamber is undermined and makes possible the change in people’s electoral preferences.

To give way to the next mechanism it is important to point out that society is not only grouped in primary groups, there are also many interactions between people who are not in a scheme of interpersonal trust. For example, people interact through weak ties. These ties acquire great relevance with the massive use of social networks because they are the ones who have the potential to link a person with other social groups, due to the fact that they are in latency. This will be discussed in more detail below.

The strength of weak ties

The strength of weak ties is a mechanism of sociodigital interaction to influence electoral preferences. This idea is inspired by the classical proposal put forward by ^{Granovetter (1973)} in “The Strength of Weak Ties”, since it is pointed out that the strength of weak ties is given because they are “indispensable to individuals’ opportunities and to their integration into communities” (p. 1378). In addition, the analysis of social ties and specifically weak ties allows us to observe the social structure and relationships of groups or individuals that are not easily defined in terms of primary groups. To determine the density of a tie, ^{Granovetter (1973)} mentions that:

The strength of a tie is a (probably linear) combination of the amount of time, the emotional intensity, the intimacy (mutual confiding), and the reciprocal services which characterize the tie. Each of these is somewhat independent of the other, though the set is obviously highly intracorrelated (p. 1361).⁸

With the advent of social networks and the development of sociodigital interaction, Granovetter’s proposal has been reinforced. In “Social capital: the benefit of Facebook ‘friends’” (^{Johnston, Tanner, Lalla & Kawalski, 2011}), the interaction of a sample of students from seven South African universities is analyzed and it is concluded that the intensity of interacting on Facebook reinforces three types of social capital: Bridging Social Capital, Bonding Social Capital and Maintained Social Capital.⁹

Also, the paper “Cultivating Social Resources on Social Network Sites: Facebook Relationship Maintenance Behaviors and Their Role in Social Capital Processes” (^{Ellison, Vitak, Gray & Lampe, 2014}), shows the strength of weak ties. On the one hand, the technical possibilities of Facebook provide people with the expansion of their social network, and on the other hand, the access to friends of friends provides informative resources. In addition, social capital is generated through small efforts by users to maintain their relationships. For example: answer questions, congratulate, and be empathetic with the life and relevant events of others (pp. 867-868).

With the previous studies we observe the relevance and validity of the perspective of social ties in sociodigital interaction. In such a scenario, strong ties may become even stronger while weak ties create social capital. Likewise, weak ties are a window to new ideas, information or worlds of life, which also weakens an echo chamber. In other words, “the fewer indirect contacts one has, the more encapsulated one will be in terms of knowledge of the world beyond one’s own friendship circle; thus, bridging weak ties (and the consequent indirect contacts) are important in both ways” (^{Granovetter, 1973, p. 1371}).

By focusing on the Mexican context, it is observed that Facebook allows each person to build their own networks of contacts, update it from time to time, apply filters and determine the density of their social ties.

At work my partners told me that there was Facebook and that there could be friends. Now in Facebook I have contact with my family, friends that I know since I used Messenger and that they’ve already moved to Facebook. I’ve also met friends from elementary school (Nohemí, 56 years old, employee).

If it hadn’t been for Facebook, I wouldn’t have had any contact with childhood friends. However, many years have passed, each one has taken his own path and now we have different opinions (Marcela, 54 years old, housewife).

One of the main contributions provided by Facebook is the possibility of contacting people just by searching for their names. In particular, this contribution is an advantage for those people who were born before the development of the Web and who lost contact with their old friends. That’s one of the reasons Nohemí opened her account. In addition, it is observed how her contact network is being expanded and how it is nourished with people that it has known in her trajectory as an Internet user. Regarding Marcela, the contribution that represents Facebook to contact her friends is also present. Likewise, it should be noted that Marcela’s echo chamber is not airtight when you point out that the opinions of her contacts have changed over time.

Each person’s management of their own contact networks responds to different reasons, one of them is their own experience and the use that everyone makes of the account.

At first, when I opened Facebook, I accepted almost every friend request that came to me, but I started getting messages from men harassing me. Then, I decided that I wasn’t going to accept people I didn’t know, even though we had friends in common (Angelica, 22 years old, student).

I use my Facebook account to be in communication with my colleagues, with people from work or dance classes, friends and family. I have deleted many contacts. For example, now my contacts are like 400, but they are people that I can find in the university, near my house or around my neighborhood (Mario, 21 years old, student).

In the previous cases we can see how the contact network changes over time. They are looking for a network that generates trust and comfort. For example, in Angelica’s case, an experience of harassment motivated her not to accept people she had not met. In Mario’s case, he eliminated the contacts he didn’t know personally. He only left his primary groups and whom he might see at some point. In both cases, the weak ties observed can be represented by known people, although the frequency of interaction is low. In that sense, a person’s contact network can take many forms, but it is possible to distinguish between strong and weak ties. Usually people themselves make that distinction, differentiating between friends and acquaintances.

At first when I started using Facebook I accepted anyone, female or male, because I was seeing what was going on with Facebook. After a while I started to clean my account and eliminate people; I saw who I knew and who I didn’t know. Then my network was among the smallest circle of friends and acquaintances. For example, if I met someone at a party and it felt right, then we added ourselves to Facebook, or if you want to know something about someone you send a friend request (Jimena, 26 years old, internationalist).

In high school I went to a convention in the United States stayed there for two weeks and still I retain these people because they bring back memories. Sometimes we talk (Valentín, 20 years old, student).

As can be seen in Jimena’s case, trust is a central element in adding or accepting people with whom she interacted face-to-face. While in Valentin’s case, it is the memories that influenced him to accept people he had not seen in a long time. Both stories show how emotions work to expand the network. Likewise, the fact that a person has among his contacts people who are not a part of their primary groups, and that the ties are maintained through sociodigital interaction, promotes the construction of social capital. However, the type of capital that is built is that promoted from weak ties. ^{Putnam (2000)} mentions that “to build bridging social capital requires that we transcend our social and political and professional identities to connect with people unlike ourselves” (p. 411). In other word, the strength of weak ties is observed.

Returning to the Mexican electoral context, it is possible that changes in people’s electoral preferences may occur when they are exposed to proposals, news, stimuli and interaction with other people.

I was doing pre-campaign, supporting Movimiento Ciudadano and one day a friend “liked” Pedro Kumamoto’s page; this happened before the collection of signatures. On that page they began to say: “if you want to know more and you want to support, attend the meetings”. Then I went to a meeting and they told us what the project was about. I started by collecting some signatures and then supporting Kumamoto’s campaign (Damian, 20 years old, student).

It highlights how partisan identification can be undetermined from sociodigital interaction. That is to say, although a person feed solely on the news media related to his interests, the legitimacy that a contact within his contact networks can have can be stronger to modify his electoral preferences than, at times, the sympathy towards a party.

I remember my history teacher started sharing some videos of Kumamoto and the first thing that caught my attention was the visuals. He teaches at ITESO but he was my high school teacher, so I have him on my contacts. That’s where I started to get to know Kuma and then I saw all the stickers on the cars (Valeria, 18 years old, student).

In the previous story we can see the capacity of influence of weak ties. However, such influence is mediated by variables such as legitimacy, trust, charisma and even leadership. These elements can remain in time an even be strengthened from what, in this case the teacher, shows on his Facebook wall. In other words, although the last meeting between Valeria and her teacher took place when she was studying in high school, the teacher’s opinion is still important for her.

On the other hand, when weak ties are found in latency, as a result of mediation generated by Facebook, the possibility of interacting at any moment is intensified. Such latency promotes strategies on the part of users to influence electoral preferences.

What I did was gradually invite people. I was trying to send them the information to people who knew they had an interest in politics and a greater knowledge. For example, if I saw that this kind of people posted on Facebook some note of indignation or something like that, I thought they were prospects for signers and voters (Damian, 20 years old, student).

Damian’s strategy not only involved interest in promoting Pedro Kumamoto’s campaign, it also led him to know his own contact networks and to know when and to whom he could send such information. In that sense, there is a constant reflexivity and strategic vision when a person’s objective is to influence electoral preferences. This is the modus operandi of the adherents or sympathizers of a political proposal. On the other hand, the influence from a weak tie will be mediated by digital objects. This can modify the message and its density. In other words:

On Facebook, users can be exposed to posts from a variety of different people, including acquaintances, colleagues, best friends, and family members. Thus, the emotional outcomes of reading a post might not only depend on the content of the post, but may also be influenced by the relationship between the poster and reader (^{Lin & Utz, 2015, p. 30}).

Finally, the strength of weak ties is reinforced in the framework of sociodigital interaction. In addition, it becomes a mechanism that influences electoral opinion and preferences by allowing access to other types of information. These ties are in latency, so they may become strong ties or they may be diluted. As a summary and taking into account the evidence obtained in this research, Figure 1 shows how the articulation of the primary group and the weak ties generate the cracks of the echo chamber from sociodigital interaction.

Source: the author.

Figure 1 Cracks in the echo chamber 

Conclusions

Although the study on echo chambers is not recent, the evolution of the Web and the widespread use of social networks have led to a new wave of research in this regard. Above all, because it is common to point out that there is a social polarization and consequently echo chambers are generated, as closed systems, during contemporary electoral political events. Faced with this, this study points out that sociodigital interaction causes cracks in the echo chambers because people are exposed to different points of view, interpretations and stories about the political universe, due to their participation in different primary groups and their relationship with weak social ties in an online-offline continuum.

In parallel, the findings of this research help to understand the transformations of the sociological perspective of voting in the light of sociodigital interaction and the use of social networks. In this sense, the classical perspective has been revitalized with the observation of contemporary electoral processes

Although the three case studies in this research respond to an urban context and the people interviewed are above average levels in terms of schooling and Internet access, it is still relevant and necessary to analyze the social implications of the Internet in public life, emphasizing the Mexican context, within the framework of the 2015 electoral processes and from a comparative perspective.

Finally, within agenda of this research phenomenon it is important to contrast sociodigital interaction and its mechanisms with other more recent electoral processes, as well as analyzing the porosity or strength of echo chambers in non-electoral contexts.