A Study on Homogeneity between Editorials and News Sources Opinions in the Chilean Reference Press

RESUMEN Se espera del periodismo que nutra la democracia a través de la aportación de contenidos públicos relevantes. La capacidad de la ciudadanía para discutir asuntos de importancia social depende de la disponibilidad de ideas, opiniones y antecedentes factuales que contribuyan a conocer y dar sentido a la contingencia desde las más variadas perspectivas posibles. Teniendo en cuenta que se califica a la concentración de la propiedad de la prensa en Chile de emblemática, el objetivo de este trabajo es analizar sistemáticamente diarios de referencia chilenos, para establecer si las empresas de El Mercurio SAP y Copesa SA constituyen un duopolio ideológico que atenta contra el objetivo democrático del periodismo. Metodológicamente, se trata de un análisis de contenido comparativo de posturas adoptadas en editoriales y noticias en cinco periódicos para determinar si estos medios tienden a seleccionar fuentes informativas que coincidan con sus puntos de vista. Este revela que en Chile no se puede sostener de manera general o sistemática la hipótesis de una homogeneidad intra-medios. Sin embargo, los diarios analizados tendían hacia un comportamiento similar en términos de homogeneidad intra-medios, por cuanto tienden a comportarse de una manera parecida. ABSTRACT Journalism is expected to nourish democracy through the contribution of publicly relevant contents. The ability of the citizens to discuss socially significant matters depends on the availability of ideas, opinions and factual background data that contributes to knowing and making sense of the contingency from the most varied perspectives possible. Considering that the concentration of the press ownership in Chile is said to be emblematic, the objective of this paper is to systematically analyze Chilean reference newspapers to establish whether El Mercurio SAP and Copesa SA companies are a duopoly that attempts against the aim of democratic journalism. Methodologically, this is a comparative content analysis of stances taken in editorials and news stories in five newspapers to establish if these media tend to select news sources that coincide with their points of view. It reveals that the hypothesis of intra-media homogeneity cannot be sustained overall or systematically in Chile. However, the media analyzed tends towards a similar behavior in terms of intra-media homogeneity, as Chilean newspapers tend to behave in a similar way. Un estudio de la homogeneidad entre las posturas editoriales y de las fuentes informativas en la prensa de referencia chilena


A Study on Homogeneity between
Editorials and News Sources Opinions in the Chilean Reference Press RESUMEN Se espera del periodismo que nutra la democracia a través de la aportación de contenidos públicos relevantes.La capacidad de la ciudadanía para discutir asuntos de importancia social depende de la disponibilidad de ideas, opiniones y antecedentes factuales que contribuyan a conocer y dar sentido a la contingencia desde las más variadas perspectivas posibles.Teniendo en cuenta que se califica a la concentración de la propiedad de la prensa en Chile de emblemática, el objetivo de este trabajo es analizar sistemáticamente diarios de referencia chilenos, para establecer si las empresas de El Mercurio SAP y Copesa SA constituyen un duopolio ideológico que atenta contra el objetivo democrático del periodismo.Metodológicamente, se trata de un análisis de contenido comparativo de posturas adoptadas en editoriales y noticias en cinco periódicos para determinar si estos medios tienden a seleccionar fuentes informativas que coincidan con sus puntos de vista.Este revela que en Chile no se puede sostener de manera general o sistemática la hipótesis de una homogeneidad intra-medios.Sin embargo, los diarios analizados tendían hacia un comportamiento similar en términos de homogeneidad intra-medios, por cuanto tienden a comportarse de una manera parecida.

INTRODUCTION
In the case of the Chilean press, the concentration of communications media ownership can be considered emblematic.With an enormous part of national readership grouped around two news consortiums, María Olivia Mönckeberg (2009) has said about this fact that "in very few countries in the world does this take on such extreme forms, where diversity in the written press is non-existent" (p.13).
This phenomenon is worrying because of its possible consequences.In fact, the increasing concentration of news media observed in many countries for decades continues to be a subject of analysis on the part of a variety of disciplines, including journalism.Understanding that journalism responds to a social contract involving a reciprocal commitment between communicators and their audiences (Sjøvaag, 2010), it is expected to nourish democracy through the contribution of a volume of verified and publicly relevant information, and that the receivers should act accordingly.In this sense, the ability of the citizens to discuss socially relevant matters depends on the availability of ideas, opinions and factual background data that contributes to know and make sense of the contingency from the most varied possible perspectives.
The most recurrent premise is that this desirable and necessary diversity of information and approaches would be better safeguarded by a plural media system.Therefore, a functional system that serves social communications would require a variety of media with different owners, reflecting different viewpoints, acknowledging diverse cultural representations and offering mutual interaction possibilities (Klimkiewicz, 2010).
This fundamental discussion is contained in this paper 1 which had as its objective to systematically analyze Chilean reference newspapers and establish whether the El Mercurio SAP and Copesa SA companies are a duopoly that tends to select news sources that coincide with the papers viewpoints and stances.In this case, the assumption was that a homogeneous press rules the markets with the corresponding lack of diverse perspectives required for deliberation and the exercise of public control.
The specific point was to see if this concentration implies a tendency to a uniformity of the positions reflected in their editorials and the positions of their sources when commenting on the same issues, and if this would generate intra-media homogeneity, as described by McQuail (1998). 2 A content analysis was applied on the basis of the theory of media diversity, comparing events that had been editorialized and given news coverage in five newspapers over a period of three years, three of which were published in the capital and two which were regional reference media.The study showed that the intra-media homogeneity hypothesis is not generally and systematically sustained in Chile.What is observed is a tendency to media homogeneity, as Chilean newspapers tend to behave in a similar way.

THEORETICAL FRAMEWORK
There is a consensus of the legitimacy, and even the duty, of the press to be coherent with its editorial line and express in its editorials the positions of the newspaper consistent with that line (Horwitz, 2005;León & Blanco, 2009). 3According to González (2005), "the editorial is the press discourse par excellence, as it openly contains the interest of the news company to participate in the organization of the public space and its ideological positioning vis-a-vis political-social contingency" (p.525).
But there is also a general agreement that the positions of the medium should be complemented with different ideas and opinions that expand and enrich public debate (Schudson, 1995;Carpentier, 2007;Harp, Loke & Bachmann, 2010).The news pages are preferentially reserved for this as it is here that a newspaper contributes to the consolidation of democracy with a plurality of opinions on issues in which there is no consensus or that are controversial, so as to reach the best possible decisions or informed opinions (Sjøvaag, 2010;Puente & Mujica, 2003-2004).Apreza (2005, p. 66) points out that news pluralism does not necessarily require more media but rather "the greater number of opposed sources of information accessible to the public or audiences." The selection of sources and their statements is one of the main factors that determine the way in which people will appreciate reality and in which the media will influence the attitudes and opinions of the audience (Shoemaker & Vos, 2009;Chia & Cenite, 2012;Mayoral, 2005).This is why Casero and López (2012) conclude that the use of sources is an indicator of relevant journalistic quality.According to Carlson (2009), sources are an epistemological representation of an essential form of evidence and Aller (2012) says that a newspaper that lacks sources is nothing but propaganda.
Even with the growing concern about the homogeneity of the social actors represented in the media and attributed to the phenomenon of deregularization of the media and concentration of media ownership (George, 2005;Llorens, 2003;Horwitz, 2005) there are also studies that consider that others are responsible for this lack of diversity.After an in-depth study of the literature regarding research into political bias in the news, Hopman, Van Aelst andLegnante (2012) concluded, as Sánchez-Tabernero (2006), that the exact establishment of the presence and origin of these imbalances requires a systematic and extended analysis to test the different explanations and factors that would affect the phenomenon, such as political systems and journalistic routines.
The subject of the voices represented in the media has also been studied in Chile, in a context in which Sunkel and Geoffroy (2001) affirm that, "the instauration of a society whose access to information is a guarantee of pluralism and diversity, has been rather scarce in the past decade" (p.5).Gibbs and Parrini (2009) investigated the presence and treatment the press gave to issues on social development, economics, politics and culture in Chile, in which the readers ended up suggesting that newspapers should be more open to the diverse positions of the citizenry, including minorities, and that they should incorporate new voices that renew sources and multiply the viewpoints published.Furthermore, Valenzuela and Arriagada (2009) detected a high level of thematic uniformity between news agendas by comparing media agendas in Chile.However, our study detected in its first phase a significant thematic and news topic diversity in the editorials of five Chilean reference papers analyzed (Gronemeyer & Porath, 2013a;Gronemeyer & Porath, 2013b).Díaz (2008) focused on a review of two historical sequences -diversity and pluralism in the Chilean press and the treatment that national newspapers have given to "popular" issues over the past 50 years, concluding that the popular is trapped in the El Mercurio SAP and Copesa SA duopoly, omitting issues of inequality and exclusions.Along the same lines, Loreto Rebolledo (2000) used focus groups to study the perceptions that popular classes have of the media and observed that "there is a lack of pluralism in the media, which only contain the opinions of those whose points of view are similar to those of their owners" (p.12) And Porath (2000) approached the ideological bias of the oligopoly at a macro level and that of specific cases.These showed that newspapers in general tended to present headlines that opposed Ricardo Lagos, the Concertation's candidate of the 1999 presidential election.
Several authors say that with such a strongly concentrated and functionally centralist media system like the one that exists in Chile, many of the issues that are important to society -especially in regions or environments that are removed from the spheres of influence -depend on the editorial interest of the owner or owners of the communications companies and on the standardization of their contents (Mayorga, Del Valle & Nitrihual, 2010).Thus, greater concentration equals less participation of individuals and social groups, which ends up by the media providing scant support to social development processes (Del Valle, 2012;Jiménez & Muñoz, 2008).In a recent paper Navia, Osorio and Valenzuela (2013) looked into an eventual ideological bias in El Mercurio and La Tercera, analyzing the coverage given to the first one hundred days of government of a newly elected president.They established that both media are different and that La Tercera tends to be more critical of the center-left coalition than El Mercurio, although public perception is that the latter is the more conservative.
But a study like the one presented here had never been carried out in Chile before.Based on McQuail's (1998) theory on media diversity it consisted of a systematic comparison between the editorial position taken by the medium regarding a socially relevant news issue and stances taken by the sources of information in their news spaces when both were discussing the same events.The comparative content analysis of the opinions voiced by both actors led to an intra-media investigation (McQuail, 1998) to test the premise of homogeneity in the newspapers of the duopoly which, according to Sunkel and Geoffrey (2001) exists in the Chilean press.
The authors affirm that this would have originated in the ideological uniformity of Chilean entrepreneurs, who support the neo-liberal economic model and conservative values.In this context, for example, the national reference newspaper analyzed here, El Mercurio, has repeatedly been considered the outlet of the political economic philosophy of the Chilean right wing (Soto, 2003;Sunkel 1986).
Studies by Matthes (2012) and Scheufele (2006) contributed to this work in the understanding that the media usually use the selection and placement of certain topics or attributes in an organized context to provide readers with a meaning or interpretation.
Also, the selection and classification of editorial material that was compared here with news stories was supported by research which focused on using attribution of responsibility variables to establish the position of the media.The Semetko and Valkenburg (2000) study based on Dutch newspapers and news reports discovered that this attribution tends to be the most common focus in journalistic reports.Similarly, Neiger, Zandberg and Meyer (2010) contributed to the construction of variables pertinent to the objectives of this paper, as their research centered on the analysis of criticisms contained in press reports.
On the basis of the theoretical discussion above, with this study we tried to test the hypothesis and answer the research queries detailed below: H. Intra-media homogeneity.The news spaces of the five media analyzed (El Mercurio, La Segunda, La Tercera, Diario El Sur de Concepción and La Discusión de Chillán) show a significantly greater presence of sources with positions aligned with the editorial stances of events published, in detriment of the inclusion and display of sources contrary to these lines.
Question 1.Is there any significant change in this trend in the regional paper Diario El Sur de Concepción, after it changed ownership in 2006 and was purchased by El Mercurio SAP? Question 2. Is there any significant change in this trend after the change of government coalitions in March 2010?
Question 3. Are there some newspapers that have higher degrees of intra-media homogeneity than others?

METHODOLOGY
Methodologically, this study made a comparative content analysis of editorial and news items taken from five Chilean newspapers in the month of July in three specific years.A systematic analysis of the contents spread by the media has proved to be a valid mechanism for describing and evaluating their products, because it gives access to quantitative data on public access material by means of replicable methods -an established methodology in research into communications for quite some decades (Neuendorf, 2002;Equipo VAP-UC, 2003-2004).

THE SAMPLE
The sample (see Table 1) included El Mercurio and La Tercera, the most important national newspapers belonging to Chile's two largest newspaper companies: El Mercurio SAP and Copesa SA.Together, and due to the influence of their editorial and information spaces, they are said to constitute a press duopoly.La Segunda, an evening paper, was also studied since it is another reference medium of the El Mercurio SAP chain, especially in the political arena.Its inclusion allowed for comparisons between newspapers of the same enterprise.The Copesa SA group does not own an equivalent newspaper.Diario El Sur de Concepción (hereinafter El Sur), a regional newspaper of record in southern Chile, was added.After a century in the hands of the Radical Party and then the Lamas family, in 2006 El Sur was purchased by El Mercurio SAP.This acquisition has increased the company's alleged media monopoly to the rest of the country.The inclusion of this paper in the study was considered relevant to compare its informative treatment before and after it changed ownership.The fifth paper was La Discusión de Chillán (hereinafter La Discusión), chosen because it is independent from the two news companies and because of its similarity with El Sur (both being regional reference media in their geographic area), which allowed the comparison of regional newspapers with those printed in the capital.
The years chosen for the study were 2005, 2009 and 2011 in order to cover a sufficiently extended period that enable to detect trends, but which is also a period contemporary to the study.The July printed version of the newspapers were studied, as this is considered in Chile to be a "normal" month, because it contains no foreseeable events that set a routine tone every year.
Ten socially relevant case studies were selected for the analysis.To qualify, the issue had to be editorialized and covered as news story in at least two of the five media within a same month (when there was more than one editorial, the choice involved those that were more coincidental in terms of date).Initially the intention was to include a larger number of case studies.However, due to the significant diversity of topics discussed in the editorials that were detected in the first phase of this study (Gronemeyer & Porath, 2013b), in this second stage there were few news events that met the requirements to qualify for the comparison of editorials and news stories.
In 2005 five events that fulfilled these requirements were selected: the approval of a series of substantial reforms to the Chilean Constitution, drawn up during the government of General Augusto Pinochet; the crossed accusations between government officers and the opposition owing to irregularities in fiscal contracts and lack of transparency in public institutions; the sanctuary that Switzerland offered Patricio Ortiz, a former member of a terrorist group accused in Chile of the murder of a senator; the controversy regarding the setting up of casinos in the south of Chile and the establishment of a compulsory assessment for teachers, which generated strong resistance on the part of secondary school teachers.
In 2009 the following three cases qualified: the government-opposition agreement for funding Transantiago, the public transport system for Santiago that had shown enormous deficits over the years; the prolongation of the Gendarmery strike, demanding a restructuring of the institution after negotiations with the Ministry of Justice were broken; and discussions regarding the applicability of a mass presidential pardon proposed by the Chilean Church on the occasion of Chile's Bicentennial, with the eventual inclusion of members of the military condemned for human rights crimes.
In 2011 two relevant events fulfilled the established criteria: the first cabinet reshuffle in the government of Sebastián Piñera, which involved a shift from technocratic cabinet members to ministers with a political background, searching to establish a party balance; and the continuation of student strikes.
The collection of news material included the search for pertinent texts up to five editions prior to and after the publication of the corresponding editorial.The sample consisted of 221 recorded units -chronicles, reports, and interviews, with their corresponding boxes -published in the national or political news sections of the five newspapers under study.The opinions of the five media contained in 33 editorials and the opinions of the events editorialized given by 812 sources of information.

THE MEASURING AND ANALYSIS INSTRUMENT
The study implied the creation of an instrument to measure the degree of homogeneity/diversity of editorial and sources opinions when dealing with a same event in the same newspaper.This instrument -a data sheet and its corresponding code book -contains context-related variables (like date, medium, title of the journalistic text, editorial analyzed, identification and categorization of sources) in addition to variables that classified the opinions of each of the sources by comparing their judgments with the different opinions in the editorials.For this reason, they were different for each event and medium.
Each position of the medium regarding specific aspects of the editorialized event originated a variable of opinion or judgment that was faced with the opinions of the source analyzed.In all cases, the opinion variable categories were the same: 0 = source does not comment on the opinion mentioned in the editorial; 1 = opinions were similar; 2 = the source tends to disagree with the editorial's position on this issue.
These categories were used to record each of the sources quoted by the newspaper which took up positions regarding the event discussed in the editorial.The positions of the sources were taken from the opinions they expressed in direct or indirect quotations (paraphrased by the journalist), without taking into account on this occasion the length of the quotations.
In order to ensure the reliability of the instrument, the recording processes were carried out by a team of two coders who were totally independent from the principal researcher and co-researcher, as suggested Source: Own elaboration.
by Krippendorf (1990).Fourteen additional coding assistants took part in two stages of the project and recorded the material in an activity that ran parallel to that of the official coders.In order to minimize the possibility that errors and biases might affect the final quality of the data base, coders divided their work in such a way that each recorded and coded a given portion of material, preventing a single coder from concentrating on material from a same medium.The newspapers were rotated so that each text was recorded independently by at least three coders.
Being aware of the difficulties involved in the categorization of opinions, the entire recording process included weekly meetings to ensure that the group had a coincidental comprehension of the code books, to discuss different criteria and solve those cases in which differences had arisen.
The database created from the information on use of sources on the media's part in relation to their explicit editorial position permitted the measurement of intramedia homogeneity, or the proportion in which the medium reproduces the opinions of sources that tend to disagree or agree with the explicit editorial position of the medium.

FINDINGS OPINION LEVEL OF SOURCES
The first thing that stands out is the high percentage of sources quoted by the media who did not give an opinion on at least one of the judgments that the same medium issued or would issue in its editorial regarding the event under discussion.In effect, 382 (47%) of the 812 sources involved did not give their opinion on the issues to which the medium did mention in its editorials.
This might already be taken as an example of diversity, because these sources gave opinions and made judgments far beyond the position established by the medium.This is to say that the newspapers opened their pages to other "voices", different from those of their editorial pages.But our instrument was not designed to determine if these opinions were plural or similar among themselves, as the comparison was only made in terms of the positions specified in the medium's editorial.
The rate of sources voicing opinions on editorial stances is similar in the five media analyzed without statistically significant differences in their distribution.But there exists a difference according to the year analyzed.The year in which the sources used had less to say about the judgment editorialized with a total of 55.6% was 2011.In 2005 and 2009 there was a clear majority (almost 60%) of sources that were quoted as giving opinions on the same aspects editorialized or to be editorialized in the cases studied here (See Table 2).Source: Own elaboration.
We should bear in mind that it does not mean that the sources did not have their own position regarding both 2011 conflicts, but that there are few opinions quoted by journalists on the same specific aspects discussed by the medium in an editorial on the first cabinet change in the Piñera government and the persistent student strikes.

THE HYPOTHESIS OF HOMOGENEITY
But the core question in our study is whether Chilean reference newspapers tend to use sources that support their positions or if the opposite is the case and they reflect in their informative spaces a plurality able to contradict editorial stances.
To answer this, we constructed an index that shows the degree of agreement/disagreement of the sources with the opinions established in the editorials, called the Homogeneity Index.This was done by subtracting the number of times that a source tended to disagree and the ties in which it tended to agree with the opinions contained in the editorial, and that result was divided by the number of opinions issued by the medium in its editorials.In this way, the Homogeneity Index can vary between -1 and 1, where the positive values reflect the percentage of coincidence of a source with the opinions given in the editorial, while the negative values indicate that the source tended to disagree proportionally with those opinions.Note that 0 represents both those appreciated: in 2005 the opinions that coincided with the editorial positions came first; in 2009 opinions that tended to disagree with the editorial were prevalent and in 2011 there is a return to a slight tendency towards the predominance of coincidental opinions (See Table 3).

INTRA-MEDIA HOMOGENEITY
In order to determine if the values of the index are nearer to homogeneity (values nearer to 1), its values were compared separately to the 0 value in terms of medium and year because 0 indicates that the medium tends to give an equal chance for coincidental or non-coincidental opinions in its news pages.In this case 0 is also taken to represent the sources that did not give opinions regarding the specific judgments of the newspaper in its editorial, but did give opinions on the issue analyzed, discarding those sources that limited their participation to provide information without defining their position.
Specifically the value of the Homogeneity Index was subjected to the Student´s t test regarding the 0 value for each medium and year.This test helps to establish if the difference between 0 and the observed Index value is statistically significant when taking into account the size of the sample and the variance of its values 4 (See Table 4).
In 2005 all the newspapers analyzed, with the exception of La Discusión (whose 20 cases are too few for a significant analysis) tend to significantly surpass the zero value.In other words, they all tended to publish opinions that coincided with the editorial positions expressed or to be expressed by the medium.
In 2009, despite the tendency to include more opinions that disagreed with the editorial position (expressed by the negative sign of the differences), and with the exception of El Sur (a newspaper with only 14 cases, which do not allow for a significant analysis), none of the differences regarding the 0 reference value are statistically significant.A partial exception could be La Tercera, with a statistical significance of less than 6% (0.058), which does not reach the standard value of 5%.In other words, La Tercera could be the only medium which that year tended to publish opinions that contradicted its editorial positions.But in the case of El Mercurio, La Segunda and La Discusión there is a proven tendency towards a certain balance between coincidental and non-coincidental opinions of its sources and its editorial positions.
Finally, in 2011, only two media show a significant tendency to separate themselves from the interme- who had nothing to comment on any opinion of the editorial and those who expressed the same number of coincidental and non-coincidental opinions.
The results show that with the exception of El Sur, in which during the three years of the study there was a predominance of opinions of their sources that coincided with their editorial positions, in the other four newspapers analyzed the same tendency was diate value or from "pure" neutrality (0) and show clear intra-media homogeneity: El Mercurio, and notoriously El Sur.These results allow to affirm that in the case of the Chilean press it is impossible to refer to the hypothesis of intra-media homogeneity without establishing differences between the newspapers and also apparently between the events editorialized.A more generalized confirmation would only be given in 2005.Considering the behavior of each medium in particular, the only case that approaches a systematic performance towards homogeneity was El Sur, before and after its change of ownership.Another medium with an eventual tendency towards intra-media homogeneity might be El Mercurio, if we consider that in the case of 2005 events the sources used by all the media tended to contradict editorial positions.But this statement could only be generalized with an increase of cases analyzed.

COMPARED INTRA-MEDIA HOMOGENEITY
In order to compare the behavior of the media according to the Homogeneity Index value, an ANOVA analysis was carried out, which determines if the annual differences between the five media are statistically significant among each other.The results show that in 2005 alone, El Mercurio showed a higher homogeneity index than La Tercera and El Sur.In other words that year, when four of the five media tended towards intra-media homogeneity, El Mercurio showed a higher tendency to publish opinions that coincided with its editorial line than La Tercera and El Sur (see Table 5).
Considering this index, all the other cases and years of the media analyzed tended towards a similar behavior in terms of intra-media homogeneity.Meaning that, in 2009, the heterogeneity of the media analyzed (with negative Index values) shows a similar tendency in statistical terms, at least for four of the five newspapers (that year El Sur presented too few cases to be analyzed to make statistically significant statements).performance with other media in that same year, their data do not permit referring to a statistically significant difference.So this reaffirms the conclusion of the previous paragraph, in the sense that a broader study is necessary to declare general tendencies towards intra-media homogeneity.

THE TEMPORALITY FACTOR
Another element taken into account is that this base also contains articles published before the medium gave an editorial opinion (the sources expressed opinions that were quoted without having obtained the explicit position of the medium).Therefore, for the objective of this study it is important to analyze the temporality factor: when did the sources issue the opinion referred to by the medium, and whether this occurred before, during or after the day in which the editorial analyzed was published.
This information is relevant for two reasons.On the one hand, the accumulation of opinions collected before setting an editorial position might become a guide that enables the medium to know beforehand if its position is accepted, and among whom it is accepted.And secondly this is important because one could suspect that once the medium's editorial position is established, journalists would tend to collect more opinions that coincided with the official position of the medium.
Considering the analysis of a total of 812 sources who expressed opinions on issues editorialized by the media regardless of the fact whether or not these opinions were related to the opinions specified by the newspaper in its editorial (both types of opinions would be important when it came to establishing an editorial position in a news medium), the general tendency is that at least half of the sources had already given their opinion on the issue before the medium published its  editorial on the subject.This tendency is slightly higher in La Segunda and El Sur.On the other hand, the two media that quoted fewer sources after establishing their editorial position are El Mercurio and especially La Segunda (See Table 6).
Here again there is a difference according to the year studied.The year 2011 shows the use of a lesser proportion of sources who had already voiced their opinions when the medium finally defined its editorial position (37.8% compared to 65.4% in 2005 and 61.2% in 2009).However, this did not involve a corresponding increase in sources giving opinions after the date of the determination of the editorial position of the medium (which varies between 20% in 2009 and 25.5% in 2011).What evinced a notable increase that year of the amount of sources quoted by the medium in terms of the events editorialized on the same day in which the editorial was published could be related to the urgency and political importance of the two events selected in 2011: the change in the cabinet and the student protests, which made the media publish editorials on the development of these current events.
In order to analyze the relationship between the temporality factor or the occasion in which a source issued its opinion, and whether this opinion tended to agree or disagree with the position editorialized by the medium, an index based on the moment the source gave the opinion was constructed.This index consists of subtracting from the date of publication of an opinion of the source the date of the publication of the editorial in such a way that the positive values reveal how much time after the editorial the paper gave the opinion of the interviewee and the negative values show how many days before the editorial the opinion of the source was published.
The correlation between the Temporality Index and the Homogeneity Index reports on time variations in terms of the number of coincidental or non-coincidental opinions that exist regarding the editorial position published in a news medium.
Table 7 shows that El Mercurio is the only case with a statistically significant correlation.This means that in all the other media the tendency revealed by their sources, regardless of whether they agree or disagree with the editorial point of view, is constant before and after the publication of the editorial.
The situation of El Mercurio is not only unique in this sense, but also in the type of relationship shown by the correlation.The negative sign shows an inversely proportional relation.In other words, as time passes in terms of the publication of the editorial, the opinions published by the medium tend to dissent from the opinion editorialized in it.So, before the publication of an editorial it was more feasible to find opinions that coincided with the editorial position of El Mercurio, and that after the publication of the editorial there would probably be opinions from sources that disagreed with what was published.To a certain extent, we could say that El Mercurio tends to establish an editorial position that is concordant with the opinions it has published, after which its reporters tend to give more space to those opinions that disagree with it.

CONCLUSIONS AND DISCUSSION
According to the results given, we can first state that in the cases analyzed for this paper, almost half of the 812 sources quoted by the five media did not comment on any of the editorial judgments on a given event issued or to be issued by a same medium.Furthermore, no statistically significant differences were detected in this area among the newspapers involved in the study.This might indicate that the Chilean press has made an editorial decision to publish opinions that discuss aspects that are different or complementary to the medium, which might be an indication of diversity.Therefore, if this factor is taken into account, the theory of a general and systematic homogeneity in the stances of media and news sources in the so-called press duopoly is untenable.Source: Own elaboration.
Concerning the core issue of this study, which explores the possibility that reference Chilean newspapers tend to use sources that support their positions or, on the other hand, reflect in their news spaces a plurality that is capable of contradicting their editorial stances, this study reveals that it is not possible to support the general thesis of intra-media homogeneity.
What is seen is a tendency towards intra-media homogeneity, in the sense that Chilean newspapers tend to have similar behaviors, bearing in mind that, with the exception of El Sur, in which there is a permanent predominance of the opinions that are aligned with their editorial judgments, there was a similar trend in the other four newspapers.In 2005 there was a propensity to intra-media homogeneity; in 2009 there was a big heterogeneity among the media in which opinions tended to dissent with the editorial and during 2011, in spite of the differences in El Mercurio and El Sur, which were once again more prone to intra-media homogeneity, there are no differences in the trend towards greater homogeneity in these media.
Hence, one could then affirm that this similarity in media behavior might obey newsroom routines rather than the decision of the media to systematically try to see their opinions on current events strengthened by the opinions of their sources in their news items.
As an average, there is a higher tendency towards intra-media homogeneity.And because it would seem that all the newspapers behave in such a similar way, perhaps both audience and critics tend to perceive intra-media homogeneity even stronger.More judgment criteria to explain this perception could be obtained from considering the space that the newspapers give to the source, and the opinions that agree or disagree with the editorial judgments.This was not part of this study, but it does appear as a necessary next step.
Furthermore, the relevant factor of the temporality of the opinions given should be remembered, in view of the fact that part of the articles analyzed were published before the medium published its editorial stance.The issue of the time of publication revealed that in the case of the Santiago reference press, most of the sources analyzed in this study gave their opinion before the medium determined its editorial position.In the case of the regional media this tendency evinced its greater percentage in El Sur, and was lower by half in the case of La Discusión.This could once again neutralize the homogeneity in the duopoly thesis because the newspapers would not be orienting or biasing the work of journalists through the airing of their official stances on the news items treated.
Another way of testing the duopoly theory was to ask if there was a significant change in El Sur when it was purchased by El Mercurio SAP in 2006 in terms that if the regional medium would show a greater tendency to similarity to the papers of that company, which was not detected by this piece of research.
Finally, in terms of the objective of this study, it was pertinent to assess if there was a significant change in the eventual tendency towards homogeneity after the change of government coalition in Chile in March 2010.In 2011 there were substantial changes regarding 2005 and 2009, although they do not necessarily refer to an increase or decrease of a tendency towards homogeneity.
One of the differences detected is that in the 2011 cases analyzed, the total newspapers showed a lower percentage of information sources that pronounced themselves on the editorialized judgments as opposed to 2005 and 2009, when an ample majority of sources was quoted as giving opinions on the same subjects editorialized or to be editorialized in the issues studied herein.Again, that this does not mean that the sources did not have a position regarding Sebastián Piñera's first cabinet shift, or about the student strike in 2011, but that there are no quoted opinions on the specific aspects on which the medium gave an editorial pronouncement.A future and more qualitative analysis might contribute to determine the issues on which the sources made their comments, and which were not given editorial coverage by the medium in cases of high political connotation, bearing in mind that these 2011 events were editorialized and given due coverage in the five papers analyzed.
In addition, in 2011 a relevant difference regarding the temporality factor is seen.That year had a lower proportion of sources that had already commented when the medium finally discussed the event in its editorial, although the consequence was not an increase in the proportion of sources giving opinions after the day the editorial was published, but an important increase in the amount of sources quoted on the day in which the editorial was published.This might be related to the same urgency and relevance of the cabinet shift and student demonstrations.
On the basis of this study, it would be important to go on with this investigation, giving an in-depth treatment to the analysis of source opinions and of the spaces attributed to them, in view of the fact that there are so many persons quoted that do not necessarily refer to the ideas given by the medium in its editorials.It is also important to increase the number of case studies in order to be able to carry out a representative analysis of the regional media in all cases.

Table 2 :
Rate of sources giving opinions on editorial positions

Table 3 :
Average value of the Homogeneity Index of sources regarding the editorial position of the media in terms of newspaper and year Source: Own elaboration.

Table 4 :
t Test for Homogeneity Index (test value: 0) according to newspaper and year analyzed Source: Own elaboration.

Table 5 :
Statistical significance of the differences between averages of the Homogeneity Index of the different newspaper for every year analyzed

Table 6 :
When the sources express their opinions in relation to the date of the editorial

Table 7 :
Correlation between Temporality and Homogeneity Indexes