Internet and Politics Studies in Brazil : Mapping the Characteristics and Disparities of the Research Field *

This article focuses on the appropriation of the subject of Internet and Politics by Brazilian Social Sciences. To this end, we analysed 299 papers presented at 11 conferences in the areas of Sociology, Political Science and Social Communication from 2000 to 2011 in Brazil. The methodology was based on content analysis in order to map the main authors, research centres, political and technological objects, theoretical approaches, methods, techniques and coverage areas of the papers presented in each of these conferences. Our findings suggest significant growth in the number of studies presented over the last few years, where 40.5% were based mainly on qualitative research and 56.5% had an emphasis on the civil society strand. Finally, our longitudinal analyses indicate some changes in the field, with a growing number of studies carrying out empirical research, and a tendency to focus on more specific digital tools instead of on the Internet in general.


Introduction
D espite still being considered a 'new media', the Internet is already a lady in her forties: the Arpanet, its predecessor, was created during the Cold War period, between the 1960s and 70s, and the Internet as we know it today started to spread in the beginning of the 1990s, just after the invention of the World Wide Web in 1989.Since this happened, many authors and researchers have investigated, wondered and argued about the impact of this media on many fields, including politics and, more specifically, on the democratic system.Such reflections have tried to assess how the technological aspects of the Internet, such as its interactivity, multimedia characteristics and capacity to store information impact on political systems and civil society organizations.
Thus, several studies of political topics have addressed the changes related to the inclusion of Information and Communication Technologies (ICTs) in social dynamics.Indeed, in Western countries, the use of the Internet by the most diverse types of political actors and institutions participating in the political system has been the object of systematic and empirically oriented analysis, especially in Anglo-Saxon nations (Medaglia, 2012;Saebø, Rose and Flak, 2008;Susha and Grönlund, 2012).In fact, Chadwick and Howard (2009) have indicated that studies on Internet and Politics have overmatched research on mass media and politics in the English language. 2  In Brazil, as in many other countries, there is a broad and continuous amount of publishing in this field.There are studies on the opportunities for participation offered by the State to the civil sphere (Braga, França and Nicolás, 2009;Bragatto 2008;Nicolás 2009); on online campaigns and the impact of the Internet on electoral disputes (Aggio, 2010); debates about the effects of the Internet on party organization (Albuquerque and Martins, 2010); issues related to new political practices (Penteado et al., 2011); reflections on the potential of the Internet as a public sphere and for online deliberation (Sampaio et al., 2011); on the uses of new media by civil society organizations (Maia, 2006); and on the relation between new media and social inequality (Moraes et al., 2009), among others.
Regarding the systematization efforts of the Brazilian studies, Rothberg (2008) has revised part of the area's national publications in order to propose a research agenda in digital democracy, and, through an analysis of the research findings, has aimed to assess to what extent the Internet is allowing civil society to participate in defining the agenda and public policies.His conclusions suggest some avenues of research, among which are studies for verifying the quality of participation in the process of production and evaluation of public policies and for checking the focus and scope of such initiatives.Moreover, he emphasises the need to evaluate the forms of moderation of discussions and website design, as well as the quality of information provided to support processes of public consultation.

Internet and Politics
In the same way as in the advent of the telegraph, radio and television, discussions about the impact of the Internet on political activities have been accompanied by what could be termed the discourse of technological determinism -the rhetoric about radical changes caused by the social and political appropriation of new technological artefacts.Throughout history, there has been the idea that new technologies would promote, invigorate or cause ruptures in the democratic system (Wright, 2006).
Considering the impact of the Internet on democracy, Vedel (2006) identifies three main "ages".The first one emerged in the 1950s with cybernetic sciences under Norbert Wiener's 4 supervision.At that time, the idea was to use computer technology and automated systems to re-evaluate the processes of political negotiation and conflict resolution.
Thus, computers would function as new mediators, since it became possible to process large amounts of information and to arrive at more "rational" conclusions.Of course, this approach was strongly criticized for its over-simplification of politics "into a practical, scientific system that can respond to the environment in predictable manners and achieve well-defined goals, and termed (…) technocracy" (ibidem, pp.227).
The second age of electronic democracy was from the 1970s to the 1980s with the proliferation of cable TV networks and the first private computers.According to Vedel, it was highly influenced by social movements which defended that "society would be better transformed from the bottom up and the coordination of local actions rather than through the conquest of the state central apparatus" (idem, p. 228).Thus, local communities were considered the key political arenas for participation experiments.Television was used to broadcast public hearings, debates and citizen discussions, enabling interactivity through telephone callbacks.This phase, usually known as 'teledemocracy', faced several technological limitations, such as a lack of real interactivity, and interconnectivity problems with computer networks.Nevertheless, Vedel recognizes that this period of experimentation generated active interest for the democratic potentials offered by ICTs (see also Arterton, 1987).
The third stage started with the emergence of the commercial Internet in the 1990s and was accompanied by a new ideology of freedom of information and of cyberspace itself (Coleman, 2007;Vedel, 2006).Vedel (2006) defends that the first visions were based on the 'Californian ideology', which stands for social solidarity, political liberalism (i.e. less power for the State) and ecological concerns.Hence the Internet was not just a tool for democracy, "it creates a new way of being together and a novel polity, which no longer takes place within the bounded territories of nation states, but in an open, de-territorialized, non-hierarchical space" (Vedel, 2006: pp. 229).
In his turn, Coleman (2007) believes that the emergence of the Internet led to an impetus for grandiose narratives of technocratic determinism.These narratives were based on two main assumptions.Firstly, the Internet was regarded as a new frontier, a deterritorialized cyber-utopia beyond comprehension or control by the State.Secondly, the feedback possibilities would allow everyone to vote directly on every issue and it would defy 'old' forms of political representation (i.e.direct democracy).
In the subsequent years (from the late 1990s until the early 2000s), several empirical research projects were carried out to test these points of view.Not surprisingly, the majority of the studies found that the introduction of these new technologies did not result in immediate and profound changes in hard politics.Some even called this 'politics as usual', since the political system actors adapted the new technologies to their own agenda and needs (Margolis and Resnick, 2000;Wilhelm, 2000).
Nevertheless, as the analysis progressed, the initial question "what can the Internet do for democracy?"was replaced by others about how the different media, channels and digital tools that make up the Internet can be used by individuals, organizations and institutions involved or interested in the political system to enhance democratic values.
To this, we can add several variables such as socioeconomic and cultural factors, party ideology, level of Internet access, digital literacy, etc.Likewise, the willingness of political representatives to implement or accept such innovations and the motivation of citizens to create or seize such opportunities have become subjects of study (Gomes, 2011;Marques, 2010;Medaglia, 2012;Saebø et al., 2008).
In this way, we would like to highlight four important issues that have guided our research: theory, design of digital tools, strands and methodology.Firstly, regarding the theory, just as different views of democracy have always been present in discussions on political theory, the same is true about research focusing on the impact of the Internet on the experience of democracy.Liberals emphasize the need for greater visibility and accountability.Supporters of strong democracy call for a conformation of spaces in which citizens can express their will directly.Deliberative democracy theorists emphasize the importance of exchanging rational arguments in public and the public sphere issue (Dahlberg, 2011;Gomes, 2007).Regarding our study, we have analysed the different theoretical approaches used by researchers and whether or not they bear any relation to the political objects.
Regarding the design of digital tools, one cannot ignore that this is a direct consequence of political forces behind its development (Chadwick, 2006;Gomes, 2011;Salter 2004).But, more than that, several factors can hinder or foster online political participation and the adoption of online tools, and design is considered an important variable.Some research projects have already indicated that a poorly designed website can inhibit political discussion (Janssen and Kies, 2005) and participation in e-government (Bezerra and Jorge, 2010) and e-democracy projects (Sampaio, Maia and Marques, 2011).Finally, the different options and features of each digital tool should be considered in this evaluation.For instance, one could analyse the need for identification, the role of moderation, technical features of each software (e. g. how reply and feedback work, if it is voice or text-based, if it is possible to send/upload images, videos, files, emoticons and so on, if it is more "online forum-like" or if it resembles a social network website), etc (Janssen and Kies, 2005;Salter, 2004;Wright and Street, 2007).Of course, design should not be regarded as the main deciding factor for whether digital initiatives succeed or not, as this would be a sort of technical determinism (Coleman and Blumler, 2009).Nevertheless, ignoring this question could also lead to an incomplete evaluation (Marques, 2010).Hence, this study will both map the different technological tools (websites, blogs, wikis or social networks, for example) and the political objects (i.e. government or organized civil society) analysed in Brazilian studies.
Thirdly, a distinction that helps this study is the classification of the work into the social or the institutional strand (Gomes, 2007).The former focuses on (but is not limited to) the implications of the new media for civic engagement, the public sphere, political deliberation and its relation to social capital.Common to all these points is a concern for the political education and political skills of the citizenry in cyberspace.That is, how the Internet can provide suitable places for the formulation of preferences, strengthening of links between groups with the same interests, organization of social demands and ripening of political and ideological positions.
The "institutional view" addresses three things: a) the study on the digital conformation of democratic institutions in a strict sense (digital cities and governments, online parliaments) or in a broad sense (online political parties); b) the institutional initiatives in the vector that goes from State to citizens (such as the provision of online public services and e-government); and c) institutional initiatives in the citizen to State vector (opportunities for participation or offer of input by the citizenry in the form of votes, responses to surveys, budget decisions or suggestions, records and discussion of opinions in electronic forums, etc.) (Gomes, 2007: pp. 11).
The considerations here are of a more structural order, since they relate to the organization of democratic dynamics and, consequently, the link between institutions of the democratic State and the citizenry. 5These two strands will be verified in the Brazilian research papers.
Finally, one should consider the methodological approach of the different research projects themselves.According to Macintosh and Whyte (2008), incomplete or incorrect methods account for some low results from the e-participation fields.Aström and Grönlund (2012) show strong evidence for this hypothesis with their analysis on the results of several research projects in the field.They found that the better the evaluation, the better the results; in other words, badly designed research could lead to false/incomplete results. 6  Although there is some degree of determinism in this statement, since it denies the reality of the study object and establishes that low research results are due to a bad investigation method (and what is a desirable method is largely a normative definition), it also indicates the need for different methods for distinct purposes.In relation to the present study, this means that a large variety of objects should, in some way, be followed by an equivalent quantity of methods and techniques.That could be considered an indicator of an epistemologically mature field.
Given the above theoretical assumptions, which are not exhausted by the authors or the ideas mentioned, in the next section we will move on to a presentation of the methodology for analysing the papers presented at conferences in Brazil, considering these four points as starting points.Three main research questions guided the construction of our methodology and database: In Brazil, as mentioned before, there is still a gap in such comparative studies (Rothberg, 2008).There are few studies focused on mapping research related to Internet and Politics.We found only three.First, Amaral and Montardo (2011) made an analysis of all the papers presented in the Cyberculture sessions at the Intercom Meeting from 2001 to 2010 (N=443) and identified 13 major themes in the studies.Secondly, Araújo (2011) listed the Brazilian works in the area of cyberactivism from 2000 to 2010 (N=22) and classified them according to the publication period, the authors' area and institution, the research object and bibliographical sources.Finally, Bragatto and Nicolas (2011) mapped studies presented at four Brazilian conferences between 2006 and 2010 (N=71), which were categorized according to themes, objects of study, types of approach, methods and research techniques.
It is noteworthy that, in order to perform this research, there was an apparent need to define the concept of politics.We chose the broad concept of politics proposed by Maia, "not only restricted to the world of governments and the function of defining votes by citizens, but one that also encompasses the processes of civic association, negotiation and conflict among subjects in day-to-day interactions" (2006, pp.15).Initially, it was necessary to separate political activity from acts of sociability to avoid grouping very disparate researches.Therefore, we considered two key questions.The first was the object of study and the second, the objective made explicit by the authors.The papers were only included in the sample if their object and/or objective were directly related to political actors or political activities.
Regarding our sample, we considered the main Sociology, Political Science and Communication conferences held on Brazilian territory.The period covered goes from 2000 to 2011.Besides believing that a decade of analysis leads us to more general conclusions and to a better monitoring of the field's trajectory, it is also the case that the first annals of Brazilian conferences became available online in 2000.
We analysed 299 papers presented at 11 conferences.We checked all annals and proceedings available online.In order to apply the questionnaire, we used three progressive rules as follows: 1) initially, we checked only the title and the abstract, 2) if this was not sufficient, we also evaluated the introduction and conclusion and, if any doubts remained, 3) various parts of each paper were examined.
For the testing and subsequent refinement of the criteria, three pilot studies were conducted.In the first one, 30 papers were analysed and the problems and difficulties reported, which led to changes and improvements in the indicators.In the second pilot test, the application was remade with the 30 original papers and another 30 were added, but using the new categories created after the first pilot.Finally, the third test involved the application of the criteria to another 30 papers different from the previous 60.Small changes were made to the spreadsheet, which then reached its final form, as described below.

Results
Altogether Regarding the evolution in the number of papers submitted each year, we can see in Figure 1 that up until 2005 the percentage of papers presented did not exceed 2% of the total sample.There is a notable curve in 2006 (6.7%), with a tendency for papers presented over the years to increase.And finally, we can observe that in 2011 there was a considerable percentage of papers (31.1% of the sample) and that out of the 11 conferences analysed, papers were collected in nine of them, indicating a significant and more homogeneous growth in publishing.We found that 225 papers of the sample contained keywords, which is a very significant amount (75.3%), especially because some conferences do not require their use.The presence of key terms is essential for searching through the websites of the conferences and through search engines on the Internet in general.Regarding the 10 most commonly used terms, we found the word "Internet" 65 times in the abstracts, making it the most frequent term.Next comes the word "cyberspace" used 18 times; "electronic government", 14 times; "blog", 12 times; "communication" and "cyberdemocracy", 11 times each; "Internet and Politics", 10 times; "politics" and "Twitter", nine times and, finally, the term "social networks", eight times.

Institutions and Authors
With regard to the authors, we analysed variables related to the institutions to which they belonged, the region and state, education and area of their last degree, as well as whether they wrote in co-authorship and if they belonged to any research groups.
Together, the seven institutions represent over half of the total sample (61.8%), 12 there being a perceptible concentration of research in the South and Southeast, with UFBA, located in the Northeast of Brazil, being the exception.We believe UFBA's leadership is probably related to the fact that it has one of the oldest postgraduate programmes in the area of Communication in Brazil and because of its pioneering research projects both in Political Communication and Cyberculture.
Table 1 also shows the analysis by institution, from which it is possible to evaluate publishing profiles by states in Brazil.Indeed, the Southeast is the predominant region, 13   with a significant concentration of publications (49.6%).This is not due to the work of only one university -four universities contributed to this.The Southern region 14 came second (16.8%), but well below the Southeast.As far as Bahia's prominent place is concerned, it is possible to establish a close relationship between its placement and publications by UFBA.It is also worth mentioning that among the seven leading institutions, six of them are public universities and five are federal institutions.
We identified 282 distinct authors in the 299 papers of our sample.Regarding the authorship of the papers, a considerable amount was written by only one author (66.2%), representing more than half of the sample.Although a large part was written in co-authorship (33.8%), a very modest amount made a partnership between institutions (8%).Of the papers with only one author, more than half were written by authors who belong to a research group (57.5%).However, several authors did not specify whether they participated or not in groups.
With regard to the title of the author, in Figure 2   If we add Sociology (9.7%) and Social Science (8%) to these areas, there is a concentration of almost 80% of the sample.These data show that a significant proportion of researchers participate in conferences pertaining to their areas of academic education and that the conferences analysed tend not to be so interdisciplinary.We found that the first 20 authors presented an average of four papers in the period analysed, representing 28% of the sample.It is a significant concentration if we consider that the total average of the rest of the researchers is one paper.

Strands, Approaches, and Objects
Regarding information about the type of strand of the papers, we found that the majority of them were located in the social strand (56.5%).By crossing the type of strand and the conferences ( Table 2 shows the relationship between the strands and the leading academic institutions.Most notable are UFBA and UFPR, with the majority of papers belonging to the institutional strand.On the other hand, UFMG, UFRJ and USP showed research that was significantly more focused on the social strand.This is possibly explained by research areas and groups related more to one strand than the other.Table 3 provides information on the theoretical approaches and the political objects. One can see that the "participation" category (N=62) was the most recurrent and that "engagement" came second (N=36), which reveals the intention of researchers to understand to what extent the Internet can provide new ways and channels for citizens to take part in political processes.The "representation" category (N=2) was the least used, which is also symptomatic of research interests in the field. 17It is remarkable to see that the main theoretical approaches can be characterized as part of the digital democracy field (Gomes, 2007).If we add the categories 'participation', 'engagement' and 'deliberation', the result accounts for over 40% of our sample.Other approaches that are generally prominent in international studies, such as political strategy, elections, digital inclusion and political economy, have reached relatively small values in our sample.
However, we need to consider two things regarding participation overrepresentation.Firstly, there are studies that deal with political participation in democratic theory while others deal with more "abstract" concepts of participation based on Cyberculture.
Although they are different, both were classified as "participation" -something that other researchers could try to differentiate in future research.Nevertheless, we believe that the participation category would still be the most present one, not only because of the recent Brazilian history of struggle against dictatorship, but mainly because of the even more recent redemocratization process and all the related concerns with civil society entering the decision-making processes (a concern present in the 1988 Brazilian Federal Constitution), which emphasises the role of political participation in democratic processes.
Regarding the political objects, the category "non-organized civil sphere" was the one that showed the maximum value (27.1%) -however, "government" showed a small difference (25.1%).One can notice a preponderance of studies related to spontaneous manifestations by society, and of research involving issues close to the executive, legislative and judicial branches.The longitudinal study revealed no significant differences between the political objects throughout the years and was not included.Similar results were found in the work of Medaglia (2011) about research on e-participation.
However, when crossing data on approaches and research subjects, as seen in Table 3, some significant findings were made.First of all, the participation approach was mainly used to analyse the government (41.9%).Considering that e-gov was not a prominent topic in our sample, this finding suggests that Brazilian researchers are interested in seeing participation in government-led initiatives or, at least, participation with a direct impact on government.This echoes the studies from the e-participation field in the English language context (Saebø et al., 2008;Susha and Grönlund, 2012).On the other hand, the online deliberation studies in Brazil were more related to the non-organized civil sphere.
This specific result may be better assessed in future studies.We believe that these studies were more interested in verifying the political conversations among citizens; however, one could argue that the problem is that the government itself is not interested in promoting online deliberation initiatives.
Still regarding the objects of study, the actions by the civil sphere (whether organized or not) represent more than 43% of our corpus and the majority were analysed using the engagement approach.This could be connected to a more "grassrootist" notion of democracy, in which the hope for democratic improvements lies in social movements.According to this vision, the State is an agent to be beaten or even conquered.Future studies may check the reason for this focus on social institutions 18 .
We can also highlight the low value achieved by research regarding political parties (2%), which confirms the impressions of Albuquerque and Martins (2010)

Methodological Variables
As far as the methodological variables are concerned, a majority of empirical studies can be observed (65.9%).In Figure 5, if we compare the types of studies throughout the years, one can note that empirical studies have become more frequent since 2007 and that theoretical papers are decreasing.Specifically, there is a noticeable difference in 2011 between the studies of an empirical (78.5%) and a theoretical (21.5%) nature.Regarding the type of method used, a considerable number of papers were of a qualitative nature (40.5%), followed by bibliographical papers (33.1%) and, finally, quantitative papers (26.4%).A small percentage applied some kind of statistic (27.8%), in contrast with the 216 papers that did not, representing more than 70% of the sample.Of the 83 papers that used statistics, we highlight the preponderance of simple frequency (69.9%), followed by the use of descriptive statistics (22.9%) and only a small use of inferential statistics (7.2%).
However, according to the longitudinal analyses in Figure 6, the use of statistics has been increasing in the last few years.In particular, one can verify a remarkable increase in the use of descriptive statistics from 2007 onwards, and of inferential statics from 2010.Moreover, according to Figure 7, a significant number of papers have employed "content analyses" (72%  Finally, concerning research question 3, we could identify the main elements of the field and some changes over the years.Regarding the strands (Gomes, 2007), there was a slight preponderance of the social strand (56.5%) over the institutional one (43.5%),which is probably connected to the fact that the area of Communication has favoured this branch of study (since there was a greater balance between Political Science and Sociology).In turn, the technological object was the one that underwent the greatest change over time.The Brazilian researchers initially focused on major questions and conducted their research on the Internet in general.Over time, more specific objects began to be addressed and, more recently, blogs and social networks have gained prominence compared to traditional objects, such as websites.
On the methodological variables, we can highlight the presence of empirical studies (65.9%), but there was great variation from 2000 to 2011.Initially, there was a preponderance of theoretical papers and currently the majority focuses on empirical studies.On the other hand, it is still noticeable that few apply statistical techniques (27.8%) and, among these, the majority (69.9%) applies only simple frequency.Only 7.2% used techniques of inferential statistics.
Throughout this text, we have indicated some possibilities for future research.For example, the concentration of research in the Southeast and South of the country can be related to either a greater number of researchers in these regions and/or to the presence of some prominent researchers in institutions located in these areas.Moreover, our mapping indicates that the Brazilian Internet and Politics field has paid more attention to issues related to digital democracy and to the civil sphere.Moreover, these results could initially be compared to other Latin American countries and, later, to Europe and the United States, searching for differences and similarities among these groups.
Finally, one can see that certain methodologies are still underused by Brazilian researchers and a survey might better understand the reasons for this.There is still the possibility of future studies analysing the articles published in Brazilian journals of Social Communication, Political Science and Sociology, which would allow us to both verify our findings and compare the differences between initial papers (conferences) and final articles (journals).
Overall, research on Internet and Politics in Brazil is growing significantly and our study indicates a greater consolidation of an autonomous and interdisciplinary field.A significant tendency towards empirical works with more specified objects can be verified, which demonstrates that the area is moving from a more general reflection to an analysis and examination of specific objects, uses and promoters of digital initiatives, as well as of the possible effects and consequences.
This research has some limitations, however.Firstly, we opted for classifying the articles using their "emphasis", which can lead to distortions.Some articles may have more than one theoretical approach or even use two empirical techniques.Nevertheless, we believe that our methodological choice not only allows statistical analysis but also effectively highlights the emphasis of the studies.
( We only took into account the main author's institution.
The South-eastern region includes the states of São Paulo, Rio Janeiro, Minas Gerais and Espírito Santo.
The Southern region includes the states of Rio Grande do Sul, Santa Catarina and Paraná.
The longitudinal graph showed that there was no significant variation in education and therefore was not included.
The dominance of Communication researchers is not a Brazilian characteristic though.According to Tomasello et al. (2010: 538), "media-oriented journals publish nearly half of all internet and related digital new media articles" and "a core set of 14 communication, or related, journals currently publishes new media research" in English.
Again, the longitudinal graph indicated similar growth among the areas and was not included.
Longitudinal analyses have not revealed significant differences and were not included.

•
Research Question 1: Is the Internet and Politics field in Brazil showing signs of growth, stabilization or reduction?• Research Question 2: Was such a tendency followed by an equal distribution of research among the different regions of Brazil?Or could one identify a concentration of leading institutions and objects of research?• Research Question 3: Finally, what are the main elements of the Internet and Politics field regarding its theoretical approach, research object and methodology?Methodology Our methodology was structured so as to construct variables allowing a comparative analysis of papers about Internet and Politics mapped in the selected conferences.Similar research has already been conducted, especially abroad.We highlight the work of Kraemmergaard and Schlichter (2011), which lists 450 articles on e-government from 2000 to 2009, of Saebø et al. (2008), who found 131 articles on e-participation published until 2006, and of Medaglia (2012), who, in continuity with the research by Saebø and team analysed 123 articles on e-participation in the period from 2006 to 2011.In turn, Susha and Grönlund (2012) worked only with articles presenting e-participation frameworks (N=50) and Aström and Grönlund (2012) dealt only with research covering specific e-participation projects (N=58).Nevertheless, there are still few research projects covering countries other than Europe and the United States, especially comparative ones(Chadwick and Howard, 2009).
a) Conferences: we considered the seminars, conferences, meetings and symposia held in Brazil.We adopted certain criteria in this selection: the existence of a work group, area or session that included studies on the relationship between the Internet and Communication, Sociology or Political Science; regular periodicity; nationwide coverage; possibility of participation by graduate students.Each paper was classified as to (1) the name of the conference, (2) the institution hosting the conference, (3) the year of the conference, (4) the name of the session.b) General characteristics of the paper: we collected information related to the name of the paper, the keywords and, lastly, the level of detail required for analysing the paper: (1) abstract, (2) abstract, introduction and conclusion (3) various parts of the paper.
f) Technological object: that is, the media or digital tool analysed by the paper being evaluated.We considered the following categories: (1) Discussion forums and chats, (2) Blogs, (3) Websites, (4) Social networks (Orkut, Facebook, Twitter etc.),(5) Wikis (Wikileaks, Wikipedia etc.), (6) Open source software, (7) Content share sites (YouTube, Flickr), (8) Internet, (9) Media, (10) Other.The categories "Internet" and "Media" were created after the first pilots of the analysis because some papers (especially the theoretical ones) did not address specific technological objects.g)Political object: we chose categories that would systematize the main points of attention of the studies on Internet and Politics: (1) Communication policies, (2) Electoral campaigns, (3) Political parties, (4) Government (Executive, Legislative and Judiciary), (5) Social movements and civic organizations, (6) Non-organized civil sphere ("spontaneous" political movements 9 ), (7) Other.h) Methodological variables: in this item we included the variables related to the methodology issues of the papers.It consists of the following: 1. type of approach: (a) Theoretical, (b) Empirical; 2. type of method: (a) Qualitative, (b) Quantitative, (c) Bibliographical -the classification results from the type of technique used in the research; 3. type of empirical technique: we sought to identify the main analytical techniques used by Brazilian researchers, namely: (1) Survey, (2) Ethnography, Interview, (4) Discourse Analysis, (5) Content Analysis, (6) Mapping of networks, (7) Did not apply any of the aforementioned techniques.4.use of statistical technique: firstly, we applied a dummy variable on the use of statistics (or lack thereof), and then verified the type of statistics used.They were: (1) simple frequency, (2) descriptive statistics, (3) inferential statistics.

Figure 1 .
Figure 1.Papers per year one can see a concentration of authors with a high level of education: Ph.Ds (35.5%), followed by Master's students (27.8%) and doctoral students (20.1%).15

Figure 3 .
Figure 3. Area of the Researcher

Figure 4 .
Figure 4.The longitudinal analysis of technological objects

Figure 5 .
Figure 5. Type of study per year

Figure 6 .
Figure 6.Use of statistics per year

Figure 7 .
Figure 7. Type of empirical technique

bpsr Maria Alejandra Nicolás, Rachel Callai Bragatto and Rafael Cardoso Sampaio
c) Authors and institutions: we included variables such as (1) name, which allowed us to check whether or not there was a concentration of authors in the sample; (2) the author's institution, seeking to note if there are major research centres in the area; (3) the Brazilian state the institution is located in; (4) the title of the author (2013) 7 (2) 114 -140 Politics, so we chose to create broader categories 8 that would summarize the main theoretical approaches in the area: (1) Transparency, (2) Information, (3) Participation, (4) Deliberation, (5) Engagement, (6) Digital inclusion, (7) Social capital and political culture, (8) Identity and citizenship, (9) Political and electoral strategy, (10) Accountability, (11) Representation, (12) Political economy, (13) Other.

bpsr Maria Alejandra Nicolás, Rachel Callai Bragatto and Rafael Cardoso Sampaio
, we analysed 299 papers presented at 11 conferences on Brazilian territory.We listed 36 papers from the Meeting of the National Association of Post-Graduate Programmes in Communication (Encontro da Associação Nacional dos Programas de Pós-Graduação em Comunicação -Compós) from 2000 to 2011 (representing 12.0% of the sample), excluding 2003 and 2004 because there were no papers on the subject.Regarding the Meeting of the National Association of Post-Graduation in Social Sciences (Encontro da Associação Nacional de Pós-Graduação em Ciências Sociais -Anpocs), 27 papers (9.0%) were evaluated from 2010 and 2011.From the Meeting of the Brazilian Association of Researchers in Communication and Politics (Encontro da Associação Brasileira de Pesquisadores em Comunicação e Política -Compolítica), we found 52 papers (17.4%) from 2006, 2007, 2009 and 2011 (therefore covering all of the association's meetings).From the Brazilian Congress of Sociology (Congresso Brasileiro de Sociologia -CBS), there were 10 papers (3.0%), from 2005, 2009 and 2011.From the Meeting of the Brazilian Association of Political Science (Encontro da Associação Brasileira de Ciência Política -ABCP), we analysed 13 papers (4.3%) from 2008 and 2010.From the National Seminar of Sociology & Politics (Seminário Nacional de Sociologia & Política -UFPR), until 2011.And, finally, from the Latin American Congress on Public Opinion -Wapor (Congresso Latino-americano de Opinião Pública -Wapor), there were 19 (6.4%) from

Table 1 .
States in Brazil and Education

Table 2 )
, one can see that those conferences related to the areas of ca, CBS) had more papers on the social strand.However, in the conferences related to the area of Communication (ABCiber, Compós and Intercom), papers on the social strand prevailed.

Table 2 .
Conferences, academic institutions and types of strand

Table 3 .
Theoretical approaches and political objects Source: Authors.

bpsr Internet and Politics studies in Brazil: Mapping the Characteristics and Disparities of the Research Field sample
about the low interest of Brazilian researchers in the relation between political parties and ICTs..In turn, the technological objects of "web 2.0" (blogs, wikis and social networks) still represent only 24.5% of the corpus -if combined together.However, after the longitudinal analysis, one can see that the "Internet" category was the most studied until 2003.Thereafter, other platforms become more prominent, such as studies on websites.In 2006, studies on blogs and social network websites started to appear, with a significant increase until 2011.Therefore, we can affirm that Brazilian research was able to keep up with technological evolution and that there is an indication that new technological platforms (e.g.social media) are going to be significant in Brazilian academic research.

bpsr Maria Alejandra Nicolás, Rachel Callai Bragatto and Rafael Cardoso Sampaio
Brazil, the Occupy Wall Street movement and the Arab Spring, for example.Some online annals were not available.The Intercom makes all the annals from 2001 to 2011 available, but the 2004 annals were organized in a way that made it impossible to locate the papers.Therefore, this year is not in our sample.The ABCP has only the last two conferences (2008 and 2010) available online.Regarding the Brazilian Forum of Political Science (Fórum Brasileiro de Ciência Política -Ufscar), we could only find online the annals of 2011 onwards.Finally, we only found the 2011 online annals of WAPOR, which were included because it was held in Brazil.In the case of CBS, the annals were available from 2000 onwards, but we only found papers on the theme in the annals from2005, 2009and 2011.Federal University of Bahia (Universidade Federal da Bahia -UFBA); Pontifical Catholic University of São Paulo (Pontifícia Universidade Católica -PUC-SP); Federal University of Minas Gerais (Universidade Federal de Minas Gerais -UFMG); Federal University of Paraná (Universidade Federal do Paraná -UFPR); University of São Paulo (Universidade de São Paulo -USP); Federal University of Rio de Janeiro (Universidade Federal do Rio de Janeiro -UFRJ); Federal University of Rio Grande do Sul (Universidade Federal do Rio Grande do Sul -UFRGS).