Media and Generations in Portugal

Abstract

Many discourses link media with generations, ascribing particular appetence for the use of new media to youth, in contrast with older generations. This article aims to give an account of the empirical regularities but also differences found in a longitudinal quantitative analysis of Internet users, uses, and media preferences among different Portuguese age cohorts. Making use of questionnaire surveys with representative samples, the importance of generational belongings in structuring different types of relationships with media is demonstrated, as youngsters seem to prefer new media. However, this variable does not hold in itself complete explanatory power. The importance of formal education, besides other sociographic variables, is clear, indicating the existence of social disparities within age cohorts, with repercussions on the mediated access to resources and opportunities. The diversity and inequalities found through the statistical analysis help to combat the rhetoric of a supposedly innate, all-encompassing digital nativity. By adopting a social constructivist approach, with a more holistic scope, this article also aims to reconstruct part of the complexity and multidimensionality of the relationship between and within Portuguese generations with media, thereby deconstructing more essentialist and homogenising notions of youth and generations.

1. Introduction

1.1. Aims, Research Questions, and Relevance

We constantly see the category “generations” being used as a way of understanding and describing the world—generationalism [1]—and, in parallel, a re-emergence of its use in youth studies [2].

It has also been used to understand the relationship of people with media in a contemporary world characterized by the increasing importance of mediation [3] within a process of deep mediatization [4] that, dialectically, influences society. More than ever, media pervades everyday life [5] and are part of the “texture of experience” [6] (p. 2), amounting to a crucial force of symbolic power [7], for much of the social construction of reality is now mediated [4]. Therefore, media is a crucial field that must be studied if we wish to better understand contemporary societies.

Drawing on the sociological importance of generations, this article aims to provide an account of the empirical regularities, but also differences, found in a longitudinal quantitative analysis of Portuguese Internet uses and media preferences among different age cohorts—which can be defined as “those persons born in the same time interval and aging together (…)” [8] (p. 844), “who experienced the same event within the same time interval” [8] (p. 845).

Approaching the problem from a multidimensional perspective rooted in the Portuguese context—a country that ranks 16th out of the 28 EU member states in “The Digital Economy and Society Index” (DESI1) 2018 edition, therefore belonging to the medium-performing cluster of countries — the following main research question was formulated:

RQ: Is there, in the Portuguese context, anything particular about the relationship of youth with media?

Alongside which, a few subsidiary research questions were formulated:

SRQ 1.1: Are there different patterns of Internet use in Portugal? SRQ 1.2: If so, what are they like? Is Portuguese youth more predisposed to new media use and online activities than the older generations? SRQ 1.3: If it is, can we relate these differences to age cohorts?

SRQ 2.1: Do different age cohorts prefer different media—youth favouring new media vs. elderly generations favouring mass media? SRQ 2.2: If this is the case, can these differences be explained solely on the basis of age cohorts?

1.2. Generations

To discuss generations by overcoming a more naturalistic definition, one that sees generations as simply succeeding one another in linear order, we must turn to the seminal work of Karl Mannheim [9]. Mannheim proposed a historical, culturalist view of generations, stating that age cohorts on their own are not sufficient to shape a unified sense of generational belonging within coevals. He thus unfolded generations into a tripartite typology: a sense of collective generational identity is formed not merely out of being born at the same historical time and place (generation as location - Generationslagerung), but by synchronically experiencing crucial, shared historical events and societal processes, at the same biographical stage (generation as actuality—Generationszusammenhang). Hence, generations are not reducible to coevals. Generation as an actuality only emerges when a worldview, a collective memory, and self-consciousness are shared but, even within this age cohort, social differentiation will be found—the so-called generation units (Generationseinheiten), or social generations, as refined by Pilcher [10] and Aboim and Vasconcelos [11], who have shed light on this multidimensional approach to generations. This notion of generation unit intends to move beyond a reductionist, homogeneous view of age cohorts by fostering a sociological view instead.

Many have followed and proposed refinements of Mannheim’s seminal contributions [10,11,12,13,14,15]. Some have combined Manheim’s legacy with the contributions of Bourdieu [13,16,17]. Based on sociological theory and drawing mainly on Bourdieu, a generation can be defined as “a cohort of persons passing through time that come to share a common habitus, hexis and culture, a function of which is to provide them with a collective memory that serves to integrate the cohort over a finite period of time” [13] (p. 93). A generation must then have a kind of self-reference and a “we-sense” [14] (p. 258).

1.3. Media and Generations

The academic discourse relating generations and media is not new [18]. One of the most strongly used terms in contemporary discourse on media and generations has been that of “digital natives”, coined by Marc Prensky [19,20]. However, other concepts attempting to refer to the relationship between generations and media have been crafted. To name a few: Tapscott [21] employed the term “net generation” for all those born between 1977 and 1997, Rheingold [22] coined the term “generation txt”, associating it with the use of text messages, Feixa wrote about the “generation@” [23] and later on the “#generation” [24], Huang [25] of the “social media generation” and Taipale of “first” and “second digital generation” [26], Sackmann and Winkler, based on the idea of technology generations, as “groups of birth cohorts whose conjunctive experience with technology is differentiated by social change” [27] (p. 493), have found empirical evidences of a technological associated cohort influence with the emergence of an “internet generation” (those born after 1980).

1.4. Media, Youth, and the Social

At an almost similar pace to this lexical fruitfulness and sometimes simplistic abuse by media discourses (often transforming complex and multidimensional concepts into shallow buzzwords) contributions have been made towards the critical deconstruction of the rhetoric present in some of the epistemological foundations of a supposedly “digital global generation” composed of “digital natives” [28,29,30,31,32,33,34]. These criticisms call for a more refined and multidimensional view that does not conceal internal variations existing within this age cohort, situating it instead within multiple contexts of existence. Special vigilance against the danger of expressions like “digital generation” is stressed [35,36], as well as against other kinds of essentialist notions, abusive, labels, and forced dichotomies between generations, based solely on the year of birth and apparently defined by technology alone.

It is important not to forget that media and youth are not monolithic categories [37]—in fact, youth and its meanings are increasingly elastic [38] (p. 6). Reference to Bourdieu’s classical text—“Youth is just a word” [39]—is already an academic social regularity in the sociological approach to youth. It is typically employed to promote an understanding of youth that is not only biological but that views youth as a socially constructed category—since the divisions between ages are arbitrary. It is a socially (re)constructed and contingent notion that depends on the historical, cultural, and social context in which it is rooted. As such, the prior delineation of youth in an age cohort is artificial and does not allow to encapsulate the juvenile condition. This condition is typically marked by a transitional phase in the life course, between adolescence and adulthood, characterised by dependence, but also by the construction of autonomy. At this articulated juncture, there is an enormous diversity of individual biographies and social contexts of existence.

Buckingham [40] supports that the repercussions of technology are not rigid, affecting all social agents, regardless of their generation. The impacts of media will depend more on the conditions of their use than on generational belonging. These individual contexts and life courses, id est “a sequence of socially defined events and roles that the individual enacts over time” [41] (p. 22) are variable and unequal, even within a specific age cohort. Buckingham also criticises the dualistic narrative of natives versus emigrants, and therefore, an exotic and essentialist take on youth, which can inflate the already distorted view of a profound and uncommunicative generational divide. Moreover, continuity, intrinsic of social change, is still often neglected and absolute power is conferred to media and technology [36,42]. In fact, a historical reading shows that many of the apparently radical innovations are, in fact, reconfigurations through new modalities [43].

Therefore, it is necessary to reject views of media and society imbued with technological determinism [44]. As Wajcman [45] advises, it is important to refute the grounds of this more essentialist view, realising that any technological system is never merely technical, but rather the reflection of a technical dimension, which is also political, economic, cultural, and organisational. Selwin [46] even appeals for academics to critically withdraw from this sociocentric view on youth and new media, stressing the need to “make the familiar strange”—here paraphrasing Mills’ classic [47]. To that effect, it is important not to forget the social component of digital technologies and to recapture social theory for a sociological mode of reasoning that avoids, among others, on the one hand, the danger of technological determinism and Manichaean cause-effect thinking; but also, on the other hand, the opposite intellectual trap of anti-essentialism or sociologism. In that sense, it is important to adopt anti-deterministic theories that defend the flexibility both of technology and of social interpretations. Namely, those under the terminological umbrella proposed by MacKenzie and Wajcman [48]—the “social shaping of technology” (SST) [49]—from where the line of the “social construction of technology” (SCOT) is derived [50,51], as well as theories of domestication [52,53,54,55,56]. Both proposals, epistemologically rooted in social constructivism, make it possible to understand the relationship between technology and society in dynamic terms, defending that the two are mutually shaped in a dialectical way. In other words, technologies and practices of users are co-constructed [57]. Therefore, the relationships between media and society are dynamic and dialectical, and the process of information and communications technologies (ICT) adoption and domestication is not homogeneous in all age cohorts but characterised by social disparities instead.

1.5. Digital Divide

Focusing on these differentiations and social inequalities, the problem of the digital divide emerged in academic and political discourses in the 1990s [58]. In that still, embryonic stage of Internet dissemination characterised by relatively unsophisticated platforms and uses, academic and political agents analysed the problem mainly in infrastructural terms [59] (p. 193), and particularly concerned with issues related to the inequalities existing in patterns of access to ICTs [54,60].

The passage of time—both in social reality and in technology, as well as in the ways of thinking this relation—brought the term “second-level digital divide” [61] into academic discourse. This term calls for the need of a more sophisticated and multidimensional analysis [62], one that pays due attention to the differences in activities and uses [63,64], as well as to different levels of media literacy. Essentially, with the change in nomenclature, there was a shift from the dichotomous digital divide to a more nuanced digital spectrum [65,66] and an interest in digital inequality [67], id est in the digital and networked disparities resulting from (and leading to) different social contexts and the unequal access and distribution of resources.

More recently, with increasing access to ICTs and the consolidation of research, some authors have started to refer to a “third-level digital divide” [68], linked to individuals’ capacity to effectively capitalise the potential of new media and their tangible outcomes [69,70]. Since Internet use may, or may not, be used as a tool for capital accumulation [71], the central question now is: who benefits from the use of ICTs and how? This view establishes the link between online and offline by calling for an epistemological approach that regards these two dimensions as being interdependent and superimposed [72,73,74,75]—debunking the myth of the cyberspace [73]. Likewise, it thereby rejects technological determinism, attempting to understand how new media are embedded in the social fabric, and consequently, are also socially co-shaped.

1.6. Youth and New Media Studies

An empirical observation of the disparate relationships social agents establish with media helps to deconstruct discourses on the homogenising power of media and to overthrow any deterministic, linear and rigid conception of society and generations. In fact, and focusing on young people, there is countless empirical evidence reflecting, on the one hand, their unequal access to ICTs, and, on the other hand, their varied relationships with media [31].

These studies empirically show how generational belonging has served to predict the use of new media over the years and across different national contexts, with youth being faster in the adoption of ICTs [75].

According to Mannheim’s generational reasoning, being part of the same age cohort implies the location and exposure to specific historical events, particular life chances and, at the same time, the experience of a particular media environment—defined as “the entire body of available media at any given time” [76] (p. 3). Besides this affinity on a more material and infrastructural level, media can participate at the symbolic level in this process of mediating forms of collective memory, while also promoting generational identities [18,77]. This is even more relevant, especially with “fresh contact” amid the formative stages of the life course, such as the years of late adolescence and youth [9,78]. Therefore, there is theoretical basis supporting the fact that childhood and adolescence are fundamental stages at the beginning of media socialisation and media biographies [77], which can shape future patterns of media domestication and literacy [18], as well as promote media nostalgia [79,80].

These studies tend to recognise greater appetence on the part of younger people, nevertheless, this does not happen in an all-encompassing manner that allows saying there is a unitary digital generation. It is not only technologies that promote and accelerate those differences in practices and representations in a direct, unilateral way. Therefore, it is necessary to adopt a critical perspective towards homogeneous categorisations of merely generational or age cohorts and be vigilant to contexts of use (and non-use) in order to better understand practices and their circumstances [54]. Especially because most of the research suggests that young people are interested in, and motivated for, the use of ICTs, more because of what they can access and do by using them than because of technology in itself [36] (p. 9).

Additionally, the use of new media does not, on its own, miraculously instil in its users neither digital erudition nor a conscious, critical and sophisticated media use. Hargittai [81] portrays different levels of literacy and autonomy in the use of new media through the so-called Internet generation—from digital natives to “digital naives”. In other words, uses among young people are not always as sophisticated and rich as is commonly believed, nor are these agents as techno savvy [82] as some less attentive, homogenising interpretations portrait them. In this fashion, even among young people with access to ICTs, gradations of digital inclusion/exclusion were found; not all were capable, to the same extent, of a selective, filtered appropriation, guided by the achievement of individual and collective goals. This means they reached different positions on the ladder of opportunities [83]. What qualifies a basic, autonomous user in the use of media is to be found in a complex network of variables, which includes, for example, the experience and scope of the use, but also attributes like age, gender, and education [30] (p. 506), as well as the person’s stage in the life course. Recently a new body of knowledge tries precisely to promote a generational perspective to social change [84] and address the importance of media in the formation of generational belonging [85] alongside the role of individual life course regarding media adoption and use [86,87,88].

2. Materials and Methods

The data analysed here stem from the collection of data lead by the Portuguese Observatory of Communication (OberCom) and Centre for Research and Studies in Sociology—CIES-IUL (Centro de Investigação e Estudos de Sociologia), through the wave of surveys “Sociedade em Rede” (Network Society)2, to which the 2015 questionnaire “ERC—Públicos e Consumos de Média” (ERC—Media Audiences and Consumption)3 from the Portuguese Media Regulatory Authority (ERC—Entidade Reguladora para a Comunicação Social) was added.

They are face-to-face questionnaire surveys representative of Portuguese subjects living in mainland Portugal aged 15 or more. The analysis combined the use of univariate and bivariate descriptive statistics, as well as multivariate methods of analysis carried out with the SPSS program.

Using the theoretical insights presented in the introduction, data was analysed in order to find the importance of age of birth in shaping different domestications and preferences regarding the media, understanding that the concept of generation is problematic, as it simplifies multidimensional features. Following Mannheim, one should be vigilant in order to not take the risk to conflate generation as location as generation as actuality or unity, knowing that the family of three time based, historicising concepts of generation, cohort and life course are many times, erroneously, conflated and used interchangeably [89]. Nevertheless, in order to operationalize empirically the research question age cohorts were drawn, knowing that in a generational analysis the difficult question is inevitably were to draw sharp cuts between age cohorts [84,85,86,87].

3. Results

3.1. Growth in the Number of Internet Users and Digital Divide

First of all, considering the rate of Portuguese Internet users, considerable growth was witnessed in Portugal (Figure 1). Whereas in 2003 Internet users accounted for less than a third (29%) of the population above 15 years, in 2015 that number had risen to nearly two thirds (60.5%)—in the intervening years the rate was, respectively: 33.8% in 2006; 40.7% in 2008; 44.5% in 2010; 49.2% in 2011; and above the half threshold (55.2%) in 2013.

Focusing on the 39.5% who claimed in 2015 to be non-users, 7.2% stated they had used before. That is, in overall terms, 2.8% abandoned their use and can be considered dropouts. Among the reasons for non-use and digital disengagement [90] nearly half (48.3%) of those who stopped using indicate price, but motivational elements are also important: 27.6% point to a lack of interest and 24.1% to lack of time.

3.2. Internet Use by Key Sociographic Variables

A bivariate analysis helps to understand that being, or not being, an Internet user is not randomly distributed across contemporary Portuguese society. It is in fact structured by age cohorts, but not only.

3.2.1. Internet Use by Age Cohorts

Making an analytical cut across the different age cohorts, Figure 2 clearly shows the effect of this variable. Starting with the most striking value, we realise that 96.3% of the respondents born between 1991 and 2000 (15–24 years old) described themselves as Internet users, whereas this percentage decreases at every jump in the generational ladder. While in the age cohort of 1981–1990 (25–34) the value is not so different (93.1%), the drop is more accentuated at every rise in age—between 1971 and 1980 (35–44) with 81.4%; 1961–1970 (45–54) 63.4%; 1960–1951 (55–64) 36.1%; and, in a more contrasting way with the youngest generations, among the respondents born before 1951, we only find 11.5% of users. Hence, this evidence about a (pre)disposition in Portuguese youth for new media use converges with results unveiled by previous studies.

In fact, if we observe the measures of central tendency, the average age of the total surveyed sample4 was 47.4 years, with a standard deviation of 18.5. Now, using the two box plots (Figure 3) and focusing on the average age of the Internet users surveyed, we observe that it was 38 (standard deviation of 14), and, among non-users, 62 (also with a standard deviation of 14). Moreover, the comparative analysis of the quartiles supports the fact that non-users tend to be older in Portugal.

However, focusing once again within the youngest, we see that 3.7% consider themselves non-users. This evidence helps to refute the minimally shared commonsensical view that all young people are equal in what concerns access to ICTs and shows that generational belonging does not suffice to explain the relationship with media since we witness a digital divide even within the younger generation in analysis. This digital divide, even amid Portuguese youth, goes along with findings from former research.

Measuring the strength of the relationship, we have a strong relationship with a Cramer’s V of 0.650 and an Eta of 0.650. Treating age as a continuous variable (not fragmented into different age levels)—which allows us to measure the relationship in a more refined and robust approach—we have an equally strong Pearson’s r5 of 0.640. Therefore, corroborating previous findings, there is a statistical relationship between the two variables, where the probability of finding an Internet user increases when we look at a younger generation.

3.2.2. The Importance of Formal Education

Examining formal education (Figure 4), in relation to the highest level of education completed, we notice there is not a single Internet user among the illiterate in Portugal. Moving on to those who never attended school, but know how to read, we find 3.3% of users; in the so-called first cycle of education in Portugal (4th year of school), we have 12.7% of users; and in the second cycle (6th year of school) that percentage increases substantially to 38.4%. However, it is in the third cycle (9th year of school) that we notice a double increase in relation to the previous one (82.9%). Already at secondary education level, the rate reaches 93% and, at the level of baccalaureate/intermediary/technological specialisation courses (CET), we have 91.7%. Moving into higher education and, unlike what could be expected, from those who completed a bachelor’s degree, 97.6% claimed to be users—which refers us to 2.4% of non-users. Finally, at second-level (master’s) and third-level (Ph.D.) degrees, all respondents presented themselves as users.

Using a Cramer’s V to measure the strength of the relationship between the two variables, compared with the remaining variables tested, we get to the strongest value of 0.760. In sum, in

Table 1. Coefficients of association/correlation with Internet use.

Independent Variable	Dependent Variable: User or Non-User
	Association/Correlation Coefficient
Level of education completed	0.759
Age group	0.650
Age	0.631
Work situation	0.595
Professional group	0.482
Marital status	0.417
Net monthly income of the household	0.401 *
Net monthly income	0.392 **
Size of the household, including the respondent	0.284
Type of employment contract	0.214
Gender	0.056
Region of residence (NUTS II)	0.002

Source: ERC2015—The new dynamics of audio-visual consumption in Portugal; n = 1018; * DK/NA = 435 (43%); ** DK/NA = 413 (41%).

, ordered in descending order, and by adding other sociologically relevant variables, we understand the explanatory power of education—a variable, without a doubt, closely linked to generation. This evidence is in harmony with previous studies and reinforces the importance of cultural capital in the institutionalized state [93] as a predictor of new media use.

Making use of the several previous rounds of surveys on the network society and measuring the relationship of Internet use with the highest level of education completed6 and age in continuous numerical terms7, we notice, in a powerful and recurring way, the importance of education in explaining Internet use (

Table 2. Coefficients of association/correlation with Internet use in previous years.

	Coefficient of the Measure of Association/Correlation
Year of Data Collection	2003	2006	2008	2010	2011	2013	2015
Highest level of education completed	0.597	0.591	0.658	0.662	0.685	0.707	0.759
Age	0.469	0.537	0.516	0.585	0.520	0.613	0.631

Source: SR2003, SR2006, SR2008, SR 2010, SR2011, SR2013, ERC2015; SR2003, n = 2450; SR2006, n = 2000; SR2008, n = 1039; SR2010, n = 1255; SR2011, n = 1250; SR2013, n = 1542; ERC2015, n = 1018.

). In a longitudinal view of all the rounds, education was always the variable with greater explanatory force—not the year of birth, although its importance is also great.

3.3. Internet Uses: Online Activities

Transposing this to Internet uses, we can indicate many different uses of ICTs—from more communicational ones (and, as such, relational and social) to others that are more informational, as well as activities driven by entertainment, consumerism, or online expressive and productive practices.

Focusing on the total surveyed Internet users (Figure 5), in decreasing order of the listed activities, those carried out the most are of the communicational/relational type: 87% send and receive emails, 79.9% use websites of online social networks, and 68% use instant messaging programs8. Next in the list, 62.8% of all Portuguese Internet users usually use the Internet to search for information with recreational purposes; over half (58.9%) to watch videos by amateurs in sharing websites9 and 57.5% to watch videos by professionals10. More informational and news-oriented uses follow, with a little over half (53.6%) using the Internet to read contents from newspapers or magazines and 52.9% to search information for professional purposes. Half also use the Internet to make/receive phone calls. With values below the average of Internet users, less than a third (30.4%) use the Internet to read blogs, 28.4% to play online games, and 25% for professional social networks. Moving on to media consumption through the Internet, 22.7% use it to listen to radio programs and 21.4% to watch television programs. Finally, only 11.5% say they take part in online courses and, giving reason to criticisms of an excessive optimism in relation to an apparent growth of a participatory culture in new media [94,95,96,97], we witness instead also in Portugal to a participatory divide [98] or digital production gap [99]: only 6.3% of Portuguese users produce contents for their own blog.

Compared to the younger generation of Internet users (Figure 5), the ordering is practically maintained but nearly always with higher values, except in the search for information with professional purposes and in pure news contents, namely the reading of online newspapers and magazines. At the top, we still have a communicational dimension of usage, where 89.2% of young people use the Internet to send and receive emails, 87.7% for online social networks, and 86.9% for instant messaging programs. An audio-visual dimension follows, with 72.3% using the Internet to watch videos by professionals and 71.5% to watch videos by amateurs. Moving on to voice communication, but through the Internet, we see 64.6% of young users making use of online phone calls. We still observe that 62.3% use the Internet to search for information with recreational purposes. It should be noted that a little over half (51.5%) claim to play games online.

Already with values below the average, 46.9% use the Internet to search for information with professional purposes, 45.4% to read newspapers or magazines, and 35.4% to read blogs. The young users who use the Internet to access other traditional media amount to nearly a third, 32.3% to watch television programs, and 32.3% to listen to radio programs. Finally, 27.7% of young Internet users use the Internet for professional social networks, 15.4% to take part in online courses, and only 6.2% to work on their own blog.

Creating an index11 with these 16 activities12 and then measuring it with age13, we obtained a correlation coefficient of −0.617. Therefore, we have a negative correlation of strong magnitude: the more age goes up, the more the cumulative number of activities carried out online decreases.

In the 2013 version, a quantitative index including a battery of 71 online activities (spanning the dimensions of communication, entertainment, information, user-generated contents, goods and services, and social networks) was combined. Relating this index with the variable age in scale, an average negative correlation (−0.589) was also found. Likewise, in 2010, creating indexes for activities of communication14, information15, entertainment16, user-generated contents17, goods and service18, education19, and measuring the association with age groups, significant values were found in the measures of correlation20—except for the correlation with the indexes goods and services and age. The correlation was negative, in the sense that the levels of use were higher in the younger ages and would decrease in more advanced ages. In other words, the more varied uses would decrease as age would go up. Hence, in a recurring way in all the survey rounds analysed, the more one advances in the age cohort of the sample, the smaller was the number of accumulated online activities and, therefore, the number of users not involved with this type of resources was superior.

3.4. Media Preferences

Broader media consumption, beyond new media alone, seems to be increasingly wider, diversified, fragmented, and structured through several media, whether this phenomenon is called “cross-media” [100,101], “polymedia” [102], “transmedia” [103,104] or “media manifold” [105]. Now, this progressive mediation and cross use of several media further complicates a holistic view of media choices that is vigilant towards the consumption of each individual across different media and devices, whether theoretical one calls it media diets21 [106] or media repertoires [107,108,109,110].

Measuring the growing plurality of uses in a combined way and across the different media seems to be increasingly difficult [111]22. In view of space limitations, we chose to use a variable with, we believe, a high heuristic potential. It implies ranking the importance of the different media and activities (television, radio, print newspapers/magazines, Internet, video games) and selecting the favoured medium in everyday life.

Television and its consumption are clearly central: when confronted with this choice, from the total population, more than half (65.5%) chose it as their favourite medium from which they could not be deprived on daily life. Second, we have the Internet among 26.7% of the respondents, then the radio (chosen by 4.4%), and 2.6% prefer to read newspapers and/or magazines. Finally, with minor relevance, we have video games for 0.8% of the respondents.

With these national figures set out, is there anything specific about the relationship of the youngest age cohorts with media? A bivariate analysis shows that the orderings vary according to the age groups, almost promoting a ladder effect (Figure 6). Contrastingly, whereas the Internet is the most important activity for the young generation, receiving 66.7% of the votes (against 28.1% for television), in the sample of respondents aged above 64 the importance of television is overwhelming—90.3% (against 2.8% for the Internet), with radio in the lead (4.6%) in this age group.

Moving this to a multivariate analysis of the same question and bringing in some of the sociographic variables23 we have previously seen on this article to embrace greater sociological relevance and explanatory power vis-à-vis the different relationships with media, an interesting social phenomenon is observed.

To carry this exercise out, a multiple correspondence analysis was conducted with the aim of graphically representing the data on social context together with this key variable (media preferences), since this technique enables an integrated view of the structural interdependencies among the multiple variables in question, managing the multidimensionality of their relationships. Graphically projecting these variables as quantifications facilitates a topological reading and its corresponding comprehension, allows us to interpret and clarify the relational structure underlying the relationship dynamics between the participating variables. As a methodological reminder, it should be mentioned that these are favoured associations, which does not mean that the whole subsample displays all of the distinct and distinctive features of each profile.

Starting our analysis in the first quadrant (+X; +Y) on Figure 7, by quantifying the categories, we see a favoured association between those married, with average to low educational qualifications (second and first cycle of education, respectively), aged between 45 and 64, and those who prefer more traditional media like television or print newspapers and magazines.

Right opposite to these profiles, in the second quadrant (–X; +Y) we have—and by comparison with the first dimension, structured according to age and educational capital—respondents with an average or very high educational capital (third cycle of education and PhD, respectively), between the ages of 35 and 44, divorced or separated, closer to the radio. The third quadrant (−X; −Y) points to a favoured trend associating the categories of young people (15–24) and young adults, who are single or in non-marital partnership and have completed secondary education, a bachelor or master’s degree, being closer to the Internet and videogames.

Finally, focusing on the fourth quadrant (+X; −Y), we have an older population profile (65 or more years of age), with little or no education, widowed, and more connected to the television. Within this type, which points to the most common and overrepresented categories in each subgroup—where space is essentially structured by age and education (dimension 1) and marital status (dimension 2), these social profiles are the most distanced from a digital and networked way of life. In other words, they are in the field of digital exclusion and, therefore and concurrently, social exclusion.

4. Discussion

This article attempted to contribute to a reflection on the relationships of Portuguese generations with media by providing some empirical clues based on a quantitative methodology and an age cohort analysis focusing on youth.

The centrality of mediation was demonstrated, whether this mediation is digital and networked, or done through the more traditional media. Television seems to still act as a strong vector structuring national media repertoires, particularly among the older generations who elect it as their favourite medium. However, with greater incidence in the younger generations, the emergence of digital and networked media repertoires and cumulative consumption patterns is noticed and here the importance of television seems to dilute itself, bearing in mind the atomisation, but also the convergence, of mediation platforms.

Concerning the Internet, the younger age cohorts are those who ascribe greater importance to it, preferring entertainment uses to traditional news information (when compared to other generations). However, intra- and inter-generational diversity reigns in Portugal. Inequalities in access and uses were demonstrated by a cross-generational analysis, which helps to deconstruct certain types of discursive rhetoric of the so-called digital generation based solely on age or of digital natives as a homogeneous global whole.

In fact, through bivariate and multivariate readings, the co-importance of formal education in shaping media uses was recaptured. The interpretation of the graphs allowed to perceive favoured associations between some attributes of social contextualisation and their relationships with media. The younger generations, with more institutionalized cultural capitals and not in a formal partnership, tend to ascribe more importance to new media. It is certainly important not to forget that the different relationships with media are not only shaped according to the generation, but also by individual life courses, contexts and resources, as well as by what people do with technology and why—what is their involvement and engagement, what are their particular motivations.

5. Conclusions

Returning to the main research question (RQ) of this paper—is there, in the Portuguese context, anything particular about the relationship of youth with media?—some conclusions can be drawn. The statistical analysis showed that there are specificities of the relationship of Portuguese youth with media that set them apart from the older age cohorts.

So, are there different patterns of Internet use in Portugal (SRQ 1.1)? Is Portuguese youth more predisposed to new media use and online activities than the older generations (SRQ 1.2)? Indeed, different patterns of Internet use in Portugal where discovered in this longitudinal analysis. In all of the scrutinised years, Portuguese youth tended to dedicate more time to online activities than the older generations. Nonetheless, can we relate these differences to age cohorts (SRQ 1.3)? Age is important, but these intergenerational differences cannot be explained solely on the basis of the age criterion, since variations in inter-age cohorts were found in the samples.

Regarding the subsidiary research questions relating to media preferences in Portugal, do different age cohorts prefer different media—youth favouring new media vs. elderly generations favouring mass media (SRQ 2.1)? Younger age cohorts said they preferred new media in daily life, in contrast with the elderly generations, which preferred television. Hence, not all the Portuguese are the same with regard to (new) media uses and representations, and age seems to be an important variable to consider. However, can these differences be explained solely on the basis of age cohorts (SRQ 2.2)? No, since homogenous age blocks were not found.

Generational commonalities, such as a shared history, social, and cultural frames, but also a particular and situated media environment, seem to be important but are not sufficient to produce sharp different modes of media use and appraisal. These differences cannot be explained solely by generational belonging or age cohorts. This kind of reasoning can shade internal differences and crystallise a false sense of unity among individuals born in the same space-time macro context, since not all coevals will experience the same event in the same way—there are always differences in social context at a more micro level, as well as biographical specificities.

This paper attempted to link social demography with patterns of preference and use of (new) media, while also overcoming the reductionism of cohort analysis by introducing other social variables into the discussion. This exercise shed light on inter- and intra-age cohort specificities, but also on heterogeneity, which can be explained by education and other important variables.

As a limitation of the study, focusing merely on quantitative data collection can obscure finer relationships and nuances that escape the large sieve of extensive research methods. More qualitative and mixed methods are needed in order to better and holistically understand relations with media. Additionally, as already stated [86,88], future research agendas should explore the life course as a critical dimension in shaping people’s relationships with media. Future research should move beyond a demographic approach to media uses and summon not only social and cultural forces and processes, but also biographical traits that, together and dialectically, shape different forms of social appropriation of media. With that research programme in mind, in order to analyse the individual and societal level, agency and structure, contextualised in space and time, life story interviews could be useful to, in a sociological sense, be able to “hear, behind the solo of a human voice, the music of society and culture in the background” [113] (p. 168).

All in all, we should not deny a kind of generational order, but it is important to acknowledge that it must be intersected by other variables. The generation concept is essential to understand the different types of relationships the Portuguese establish with media “insofar as it keeps its complexity” [34] (p. 187), because it does not explain everything—instead, it conceals a large variety of media uses, domestications, preferences, and repertoires.