Mediatization of (Im)mobility Experiences: New Media Use of (Im)mobile Groups

. theoretical literature and ethnographic studies suggest that individual spatial (im)mobility and new media use are becoming increasingly interlaced through mediatization. in this article, we tested this assumption quantitatively, by examining possible associations between media use and emigration wishes which we interpreted as an indicator of spatial (im)mobility potential. analysis of data collected in autumn 2014 in estonia via a representative survey (n = 1,503), enabled the following research questions to be answered: 1) How have the emigration wishes of estonia’s population changed over time, in regard to mediatization? 2) What kind of media use patterns describe individuals with different emigration wishes? the results showed that, compared to earlier surveys, the (im)mobility potential of the general population had not considerably changed. the mobility potential was indeed interlaced with more active and versatile use of new media, and immobility potential with less active and diverse use of new media. Yet, these associations could be explained by generational differences, as younger respondents prefer to emigrate and use new media, while older respondents wish to stay in the country of origin and follow traditional media.


Introduction
Individual spatial mobility experiences are increasingly interlaced with new media.Both aspiring and current migrants use online social networks and news sites to plan their routes and navigate while being on the move, as well as to create and maintain social and cultural ties with people and places all around the world (see: Al-Rawi 2017;Marino 2015;Oh 2016;Collin 2014).
This paper strives to provide empirical evidence on the patterns of new media use across different types of (im)mobility potential.As the intertwinement of media innovations and social transformation intensifies over time through mediatization (Couldry, Hepp 2017;Jansson 2018), we are interested in whether and how this deepening integration has changed emigration aspirations.
Recent theoretical works (Couldry, Hepp 2017;Jansson 2018) suggest that the development of information and communication technology (ICT), especially the rise of social media, has facilitated individual spatial mobility.Moreover, access to, and use of media are seen as the cornerstones of an individual's ability to control their own mobility throughout the various phases of migration (Jansson 2018).
Migrants use ICT to prepare for moving from one country to another, e.g. to gather information on the immigration policy and the job market of the destination country (Al-Rawi 2017;Hiller, Franz 2004).Social media platforms and instant messaging software also allow migrants to establish new relationships with locals and other migrants in the country of destination, and maintain existing relationships after migration with friends and family members in the country of origin (Collin 2014;Dekker, Engebersen 2013;Lášticová 2014).
Although the relationship between media use and (im)mobility has been hypothesised in the literature (Jansson 2018), it still needs to be empirically tested.Previous research (i.e.Al-Rawi 2017;Glukhov 2017;Dekker, Belabas, Scholten 2015;Schrooten 2012;Wenjing 2005) has examined the media use among migrants, but has turned less attention to the media use patterns of immobile groups, and has not analysed the potential associations between media use and (im)mobility quantitatively.Some studies (Masso, Silm, Ahas 2019) have found that the practiced cross-border immobility of younger generations has increased, but the question whether it can be explained by a change in the (im)mobility potential has been left unanswered.
In this study we strove to fill this gap through mapping similarities and distinctions between potentially mobile and immobile groups in regard to the use of media technologies and channels, practices and preferences.Accordingly, answers to the following research questions were sought: 1) How have the emigration wishes of Estonia's population changed over time, in regard to mediatization? 2) What kind of media use patterns describe individuals with different emigration wishes?The study used data from a representative survey, conducted in autumn 2014 among members of the Estonian population aged 15-74 years, and then implemented quantitative analysis methods to answer these research questions.The intensifying entwinement of mobility and new media can be understood as a form of mediatization (Couldry, Hepp 2017;Jansson 2018;Krotz 2017).Mediatization's liberal scope and long duration makes it similar to other metaprocesses, e.g.globalisation, individualisation, commercialisation, rationalisation, democratisation, and urbanisation (Lunt, Livingstone 2015).Mediatization manifests itself in many facets of mobility experience, as media innovations, i.e. online news sites, social media platforms and instant messaging programs, which have enabled people to be constantly available for communication with others from different geographical locations (Agger 2011).Due to these innovations, migration has become easier, and also more acceptable to some groups to whom it was not acceptable before, e.g.mothers, who can use video calls to converse with their children living in their country of origin (Madianou 2014).
Previous research (e.g.Dekker, Engebersen 2013;Schrooten 2012;Tabor, Milfont 2013) indicates that participating in social media communities of people with similar ethnic background may facilitate migration from one country to another by providing a means of acquiring social capital and gathering practical information necessary to adapt to the conditions in the host country.Social media platforms may serve various purposes for migrants: creating social ties with other migrants (Al-Rawi 2017); maintaining existing ties in the country of origin (Kang 2011); gathering information (Oh 2016); expressing emotional attachment to the country of origin or destination (Marino 2015); shaping identity (Schrooten 2012); expressing an opinion and participating in a debate (Al-Rawi 2017); and providing practical and emotional support (Oh 2016).These purposes are achieved by using a variety of data (e.g.written texts, photos, videos, links) and communication channels, as well as by active participation in a dialogue or passively tracking content posted by others (Komito 2011).Particular communication practices are chosen according to contextual circumstances, such as the type of contacts (e.g.other migrants, friends in the country of origin) (Lášticová 2014) and the current phase of migration (e.g.pre-or post-migration phase) (Hiller, Franz 2004).For example, information gathering about job vacancies and accommodation may prevail before migration, but keeping in touch with family members and friends in the country of origin may become more important after migration (Hiller, Franz 2004).
The wide array of communicative practices needed in the context of migration suggest that the activeness and versatility of media use, and the diversity of social media contacts have, in general, become prerequisites to spatial (im)mobility.

Emigration wishes as an indicator of potential (im)mobility
Geographical mobility has been widely analysed by previous studies within various disciplines.From the globalisation perspective, mobility and the intensification of individuals' interactions (Giddens 1990) are phenomena natural to late modern societies.Concepts like spaces of flows (Castells 1998) and mobile lives (Bauman 1995) are often used for characterising the effect of globalisation on individual lives.
Unlike the neutral term of mobility, migration is often seen as reflecting the power-relations of a particular society and as such being analysed from conflicting and policy-centered perspectives (Castles 2010).This approach derives primarily from the argument that some wealthier countries in the Global North experience very extensive inflows of migrants, while some poorer countries in the Global South suffer massive outflows (Castles 2013;2014).New media is often perceived as a facilitator of migration (e.g.Al-Rawi 2017; Marino 2015), and, therefore, may indirectly participate in the reinforcing or restructuring of the aforementioned power-relations.
The traditional approaches both to mobility and migration studies have often been criticised (Cadwallader 1992) because they are unable to analyse the complex character of mobility including both micro-, meso-and macro-level indicators.There have been some attempts for conceptual improvement of the migration theory (Castles 2010), offering general social theory as a framework in general, and concentrating on migration as a social transition process in particular.The social transformation approach is one of the theoretical frameworks offered for analysing geographical mobility (Castles 2010) and used as a framework in this article.
For this study, we understood mobility/migration as a way of cultural adaptation (Sztompka 1994) with social changes.We assumed that geographical mobility may be seen as a (lack of) cultural resource at the level of the individual that supports them when changing the place of residence.We assumed that migration is not only an expression of economic and demographic social changes, but rather a cultural process, expressing the transition of temporary mobility habits to permanent practices and expressing the social and cultural connectedness to both the home and host countries.
Emigration wishes, from this perspective, indicated individual potential for mobility and an early stage of cultural adaptation.Such wishes may or may not be fulfilled as practiced spatial mobility, depending on the resources available to potential migrants (Flamm, Kaufmann 2006;Carling, Schewel 2018).Previous research (Dubois et al. 2015) has focused more on the relationship between mobility potential and these resources, including access to transport infrastructure, and personal skills (e.g.language proficiency, map reading skills), but less on media use.

Data and methods
The representative survey "Me.The World.The Media", carried out by the Institute of Social Studies, University of Tartu, and Saar Poll market research company at the end of 2014 provided the data used in this research article.The survey covered the Estonian population aged 15-74 years.For comparison, we used data collected in 2002, 2005, 2008, and 2011 within the same research project.The data enabled us to map similarities and differences between spatially mobile and immobile groups, in regard to the use of media technologies and channels, practices and attitudes.
The total sample size was 1,503 (1,028 respondents completed the questionnaire in Estonian and 475 in Russian).To alleviate the differences between the representative population model (based on the demographic statistics) and the sampling outcome, the data were weighted by the main sociodemographic attributes (gender, age, ethnicity, and place of residence) (Vihalemm, Masso 2017).
The survey concerned attitudes towards changes in Estonia during the past ten years and covered topics like personal values, political and civic participation, usage of time, media use, life-styles and life conditions.The survey was a combination of a self-administered questionnaire and interviews.The questionnaire was created at the Institute of Social Studies, University of Tartu by a research group that included one of the authors of this paper, Anu Masso.Some of the questions were formulated from those posed in our previous qualitative works (e.g.Soll, Salvet, Masso 2015;2014;Salvet 2014).
The data used in this article enabled the mapping of similarities and differences between spatially mobile and immobile groups, in regard to the use of media technologies and platforms, practices and attitudes.We focused on analyzing three groups of variables: emigration wishes, media use (composite indicators), and general sociodemographic background variables.
To study emigration wishes as an indicator of potential mobility, the following question was posed: Would you like to leave Estonia, permanently move abroad?
Respondents had four answer choices: "Certainly not" -1; "Hard to tell, maybe in the future" -2; "For some time, but not permanently" -3; "Yes, I would like to leave, permanently" -4.For the analysis of variance, multiple linear regression, and correlation analysis, the answers were recoded: "Certainly not" -1 as "Wish to stay" -1; "Hard to tell, maybe in the future" -2 as "Do not know" -2; "For some time, but not permanently" -3 and "Yes, I would like to leave, permanently" -4 were merged into one code "Wish to leave" -3.
As the mediatization of mobility experiences assumes that mobility is more strongly integrated with new than more traditional media use (Couldry, Hepp 2017;Jansson 2018), we analysed the use of both old and new media to test this assumption.
New media use was examined, based on the next composite indicators: following online media and news portals (the frequency of using the online editions of Estonian Pobrane z czasopisma Mediatizations Studies http://mediatization.umcs.plData: 30/09/2023 15:30:44 U M C S newspapers, Estonian, and international news portals), computer use activity and variety (time daily spent on computer use for: work, studying, following media, communication, playing), variety of used Internet technologies (the use of: desktop computer, laptop, e-reader, tablet, mobile phone, smartphone), using Internet with tablet and smartphones (time daily spent on smartphone/tablet use for: work, studying, following media, communication, playing), use of various social media channels (the frequency of using: Facebook, Twitter, Snapchat, Instagram or other photo-sharing networks, Foursquare or other location-sharing services, YouTube, LinkedIn, World of Warcraft or other gaming communities, Geni.com), functional variety of social media use (sharing information, changing information, asking for support, sharing news, discussing about TV/radio programs, discussing political issues, commenting or asking questions regarding topics related to health and cultural events, suggesting books, movies or music, sharing information about products or services, following social media pages, inviting friends for participating events, giving feedback to public institutions, giving an opinion on products and services), and variety of friends on Facebook (peer family members or relatives, family members or relatives who are older than me, family members or relatives who are younger than me, colleagues and studymates, peer friends and acquaintances, friends and acquaintances who are older than me, friends and acquaintances who are younger than me, former acquaintances with whom I do not communicate anymore, strangers, public figures, friends and acquaintances who have emigrated from Estonia, foreigners).
Traditional media use was analysed, based on the following composite indicators: the activity and diversity of reading newspapers (the frequency of reading: national, provincial, city, county, Estonian Russian-language, Russian, and English, German, Finnish, etc. newspapers), the activity and diversity of listening to the radio (the frequency of listening to national talk and music, local, Estonian Russian-language, and international radio channels), the activity and diversity of watching TV (the frequency of watching Estonian public broadcasting and private channels, foreign channels with Estonian subtitles, Russian channels, international news channels, other foreign channels), and following TV news (the frequency of watching international news channels, Estonian news programs, Internet-streamed catch-up TV news).All of the composite indicators of traditional and new media use were constructed by additive aggregation of single variables.The scales of composite indicators were shortened so that their values would range from 1 to 5 (Vihalemm, Masso 2017).
We used the sociodemographic indicators of language, income, age, gender, region, education, and the social status of respondents to test for confounding variables that could explain potential associations between other analysed indicators.We used general descriptive univariate analysis to explain the temporal tendencies in the spatial mobility desires and described the structural peculiarities of potentially mobile and immobile groups, based on multivariate analysis by calculating the association coefficient Cramer's V.The values of Cramer's V vary between 0 and 1, indicating Pobrane z czasopisma Mediatizations Studies http://mediatization.umcs.plData: 30/09/2023 15:30:44 U M C S how strongly two categorical variables are associated: .000means no association, .000-.200 -weak, .200-.300 -moderate, .300-.500 -strong, and over .500-redundant association.We conducted the analysis of variance (ANOVA) to evaluate the statistical significance of differences in media use between potentially mobile and immobile groups.The ANOVA employs the F-test to compare the factors of the total deviation.Stronger and statistically significant associations were selected for multiple linear regression analysis, which allows the inclusion of all of the selected variables in one statistical model and to compare the extent to which each of them contributes to the variation of the (im)mobility potential.Additionally, the correlation structure of the model was examined to explain the results.

Results
The (im)mobility potential of the Estonian population In the current paper, we used the wish to emigrate from Estonia as an indicator of (im)mobility.The respondents were asked, whether they would like to leave Estonia and permanently settle abroad.
The proportion of respondents who wish to emigrate (including temporary and permanent emigration) has slightly fluctuated from year to year by 6-10%.According to the survey data, the percentage of respondents with aspirations to emigrate was the lowest in 2014 -only 18% (compared to 32% in 2005, 22% in 2008, and 28% in 2011) (Figure 1).As the proportion of respondents who wish to permanently move abroad has been relatively stable (around 5-8%), this fluctuation mainly reflects the changing numbers of respondents who wish to temporarily live abroad.
In Table 1, we have summarised the association analysis between the aspirations to emigrate from Estonia, and structural background variables.We have calculated the values of Cramer's V (measure of association), and marked the statistical significance of these associations with asterisks -the more asterisks added, the lower the alpha value.
The results indicate that the differences based on survey language have decreased slightly over time.Although, in 2008 and 2011, a moderate association between survey language and emigration aspirations could be detected, the association was weak in 2014.It means that, before 2014, there were slightly more Russian-speaking respondents who wished to emigrate from Estonia than Estonian-speaking respondents who wished to leave.
There has also been a moderate association between emigration aspiration and age throughout all of the analysed years.The age-specific differences increased to a peak in 2014.While younger respondents were more likely to wish to emigrate from Estonia, or were hesitant about their future plans, the majority of respondents over the age of 40 expressed their wish to stay in Estonia.  1, was that emigration aspirations cannot only be explained through economic factors, such as income, education correlated to income, or social status.Although the associations between emigration aspirations and these variables were statistically significant, they were very weak.There were not any statistically significant associations between emigration aspirations and gender.

Social media use by (im)mobility potential
To analyse the distinctions in media use between groups with varying (im)mobility potentials, we have calculated the expected values of composite indicators describing the traditional and social media use of respondents, F-, and p-values.
Table 2 demonstrates that, in general, respondents who wished to leave Estonia (either permanently or temporarily) were less active and less diverse users of traditional media, such as newspapers and television -but surprisingly not radio, as they actively listened to various stations.However, these differences in the diversity and activeness of traditional media use could be considered rather small and statistically significant only at the significance level of p ≤ 0.05.
The differences in the use of new media were larger and statistically significant at the significance level of p ≤ 0.001.People who desired to move abroad followed more online media and news portals, used a greater variety of devices (e.g.computers, tablets, smartphones, etc.) to access the Internet, used their devices more multifunctionally, used a wider variety of social media platforms for more diverse purposes, and had a wider variety of friends on Facebook (including friends living abroad).In other words, they were more active and diverse group of users of new media.
Respondents who preferred to stay in Estonia, by contrast, were the most active and diverse group of followers of traditional media -particularly newspapers and television, but not radio.They were also the least active and diverse group of users of new media, based on their interest in online media and news portals, used devices, social media platforms, and variety of social media contacts, as well as variety of purposes in using digital media.
Those respondents who were hesitant about their future plans followed traditional media (newspapers and television) more than respondents who wished to leave Estonia, but less than respondents who plan to stay.Yet, they were more active and versatile group of users of digital media than people who see Estonia as their future country of residence, but less than respondents who wanted to move to another country.In general, hesitant respondents were a bit more similar to people who aspired to emigrate from Estonia.Based on the results in Tables 1 and 2, those variables with stronger and statistically significant associations with emigration wishes or bigger variations in their means across different emigration wishes were chosen for the ensuing multiple linear regression analysis.Those variables included age, and the composite variables describing new media use.Multiple linear regression analysis allowed us to combine all of the aforementioned variables into one statistical model and to compare to what extent each variable contributed to the variance of (im)mobility potential.A multiple linear model, where the wish to emigrate was the dependent variable and other factors were independent variables, was evaluated for the total sample.Separate models were also constructed for respondents who filled in the questionnaire in either Estonian or Russian.
All three models were statistically significant (p Total ≤ .001,p Estonian ≤ .001,p Russian ≤ .001).The model for the total sample explains 22.8% of the emigration wishes' variance (adjusted coefficient of determination R 2 Total = 0.284), the model for Estonian respondents explains 23.3% (adjusted coefficient of determination R 2 Estonian = 0.233), and the model for Russian-speaking respondents -22.9% (adjusted coefficient of determination R 2 Russian = 0.229) of the variance.Table 3 shows that the association between emigration wishes and age was the strongest, and this association was similar among both Estonian-and Russian-speaking respondents.Furthermore, the association with age was the only association in the multiple linear regression analysis that was statistically significant at the significance level of p ≤ .001.
The associations with new media use indicators were weaker and statistically significant only at levels of p ≤ .05 or p ≤ .1.The relationship with functional diversity of social media use was the most notable of these associations in the total sample, but it was still rather weak, and evident among Russian, but not Estonian respondents.The association with the variety of used Internet technologies was even weaker in the total sample, and it was characteristic to Estonian-, but not to Russian-speaking respondents.The association with the use of tablets and smartphones was detected only in the total sample, but not separately among Estonian-and Russian-speak-Pobrane z czasopisma Mediatizations Studies http://mediatization.umcs.plData: 30/09/2023 15:30:44 U M C S ing respondents.Other variables, such as following online media and news portals, computer use activity and variety, the use of various social media channels, and the variety of friends on Facebook, do not have statistically significant associations with emigration wishes.
The association with age was considerably stronger than other relationships, and was the only association that was persistent in both languages.This result implied that age may be a confounding variable which explained the distinctions in media use of potentially mobile and immobile groups previously found by ANOVA (see Table 2).This assumption needs to be further analysed by examining the correlation structure of the multiple linear regression model for the total sample.Table 4 gives an overview of the correlations between variables that were used in the previous multiple linear regression analysis.All the variables were correlated to each other at the significance level of p ≤ .001.Rather strong associations between new media use indicators were very expectable, as they were measuring the different aspects of the same construct (e.g. the use of various social media channels, and the functional variety of social media use: r = .89).However, it can be seen that the correlations between emigration wishes and age (r = -.47), as well as, age and new media use indicators (e.g.age, and the variety of friends on Facebook: r = -.60) were much stronger than associations between emigration wishes and the indicators of new media use (e.g.emigration wishes, and the variety of friends on Facebook: r = .37).This finding supports the assumption that the distinctions in media use between groups with different (im)mobility potential can indeed be explained by age.In other words, younger generations had a greater preference to emigrate from Estonia, or were more hesitant about their future plans, while older generations had a greater preference to stay in the country of origin.At the same time, younger respondents used new media more actively and diversely than older generations.

Discussion and conclusions
In the article, we quantitatively analysed: 1) the extent to which (im)mobility, specifically the wish to emigrate from or stay in the country of origin, is intertwined with media innovations; 2) how (im)mobility patterns have changed, in the context of this deepening intertwinement.
According to the results, a small and relatively stable proportion of respondents has aspired to emigrate permanently from Estonia between 2005-2014.The percentage of respondents who have planned to move abroad temporarily has been more unstable.One possible explanation is that while global economic changes have no effect on permanent emigration wishes, they do influence temporary emigration wishes.In 2008, the decreased interest in temporary emigration could have been the result of the Great Recession, which started that year.Similarly, in 2014, several destination countries that have been popular among Estonian emigrants, e.g.Finland and Australia, experienced growing unemployment rates (OECD 2018).It can be assumed that in unfavourable economic conditions people are less prone to give up their jobs in the country of origin for a short-term mobility experience.This means that the current study does not support the assumption that mediatization expands and intensifies mobility (Jansson 2018).However, it must be agreed that the timeframe (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)  Yet, migration cannot be explained only through economic factors, as it also is a social and cultural process (Sztompka 1994), in which the connectedness to countries of origin and destination via new media is a key feature.The results of this paper support the understanding that differences in media usage and in the access to media may accompany a variety of (im)mobility patterns.The active and diverse use of traditional media, especially newspapers and television, and the one-sided and infrequent use of new media is more characteristic to an immobile mindset and to the wish to stay permanently in the country of origin.By contrast, the more active and diverse use of new media is more likely accompanied by higher mobility potential and a definite plan to leave Estonia either temporarily or permanently.More moderate media use patterns, that lie somewhere in between these two options, tend to be characteristic to people who are uncertain whether they would like to stay in Estonia or move abroad.Consequently, these findings are consistent with previous research (e.g.Al-Rawi 2017; Marino 2015) that suggest that digital media can facilitate migration by supporting the wish to emigrate.Active and diverse use of traditional media, however, appears to have the opposite effect.This is consistent with the argument that mediatization deepens through media innovations -the associations between emigration wishes and new media use were stronger (compared to associations between emigration wishes and traditional media use).
However, these associations can be explained through age: younger respondents were more willing to leave Estonia, and they also were more active and diverse group of users of new media than older respondents.This may indicate that, at least in the realm of mobility experiences, generational change is an important component of mediatization.
In the context of studies (Masso, Silm, Ahas 2019), that have pointed out the increasing cross-border immobility of younger generations, it is important to note that, according to this paper, the (im)mobility potential of different age groups has been quite stable throughout the surveyed years.It may mean that (im)mobility potential is increasingly realised in other ways, e.g. as digital mobility, and that mediatization is leading to new ICT-based forms of mobility.This, however, needs to be analysed in future studies.Similarly, further research is necessary to map other tendencies that may accompany mobility and immobility in the context of mediatization.For example, discourse analysis, and sentiment analysis, combining different machine learning methods, can be used to analyse the attitudes towards migrants in social media, as possible facilitating or restrictive factors of integration of migratory groups.

Table 2 .
Emigration desires across traditional and new media use (mean values and F-test)

Table 3 .
Emigration wishes accompanied by age differences and new media use habits, based on linear regression models (results in total, and by questionnaire language)

Table 4 .
Spearman's correlations between emigration wishes, age, and media use patterns -emigration wishes, 2 -age, 3 -following online media and news portals, 4 -computer use activity and variety, 5 -variety of used Internet technologies, 6 -using Internet with tablet and smartphone, 7 -use of various social media channels, 8 -variety of friends on Facebook, 9 -functional variety of social media use.*** p ≤ .001Source: Authors' own study.
for the comparisons is probably too short to capture the long-term, even historical, nature of mediatization.Therefore, further research enabling a longer timespan is needed.