Abstract
This chapter will discuss the empirical findings for each research question in separate subchapters. Follow-up questions for each media channel will be separated by inline paragraph headings in bold. The analyses for Sections 6.1 and 6.2 were performed in SPSS.
You have full access to this open access chapter, Download chapter PDF
This chapter will discuss the empirical findings for each research question in separate subchapters. Follow-up questions for each media channel will be separated by inline paragraph headings in bold. The analyses for Sections 6.1 and 6.2 were performed in SPSS. The analysis for Section 6.3 was performed in MPlus. A summary and discussion of the findings can be found in Chapter 7.
6.1 Forms and Frequency of Media-related Extramural English Contacts Among Adolescents In Germany and Switzerland
Following the literature and empirical review in Chapter 2, it was proposed that adolescents in Germany and Switzerland have regular extramural English contact via numerous media channels (H1). This chapter will look at the frequency with which students in the MEWS study reported contact with English-language media content in different media channels. Findings for each media channel will be presented separately, starting with the most popular one. For brevity, non-significant differences between students from Germany and Switzerland will not be reported in detail but can be found in the electronic supplementary material.
A first look at the questions about technical equipment reveals that almost all students have the technical tools necessary to access online media content in English (Table 6.1). There are only small differences between Germany and Switzerland: Significantly fewer German students have access to their own laptop (ϕ = .08), but more German students live in a household with a gaming console than Swiss students (ϕ = −.12) and have their own smartphone (ϕ = −.01), although this last effect is negligible in size. These results show that the technical requirements for regular online media-related extramural English contact are overall given in both countries.
A first look at the overall mean index for frequency of media-related extramural English contacts shows that almost all students, on average, engaged in extramural contacts at least several times a month (Figure 6.1). Students from Germany and Switzerland did not differ significantly in their overall extramural contact. The individual media channels will be analyzed in more detail in the following subchapters.
Frequency of English media content overall (in %)
6.1.1 Frequency of Listening to Music, Radio, and Audiobooks
Listening to English-language music is by far the most used media category. As Figure 6.2 shows, most students (85.9%) listen to English music every day, 9% multiple times per week, and only 5.1% do so less frequently. In contrast, listening to English radio programs and podcasts seems to be less common, as only 4.8% of students do so on a daily basis, while 57.2% do not listen to them at all. English audiobooks are even less popular among the participants: less than 4% of students listen to them more than a few times a month, and 77% state they never listen to them at all. Swiss students show a slightly higher frequency of listening to audiobooks (\(\overline{x}\)G = 1.39, SD = 0.80, \(\overline{x}\)CH = 1.30, SD = 0.73; t(1419.1) = 2.59, p = .01, d = .119).
Overall, music seems to be equally distributed as a source of extramural English contact within the sample. This is consistent with findings from other studies and can be explained by the dominance of the American and British music industry worldwide. Unfortunately, the data does not distinguish between active listening and passive listening. Students might passively listen to music while concentrating on other activities or actively listen to the music and try to understand the lyrics.
Frequency of listening to English media through music, radio, and audiobooks (in %)
6.1.2 Frequency of Surfing Online
Browsing on English-language websites and watching online videos in English are tied for second place among students’ most preferred free time media categories (Figure 6.3). This is not surprising, given the overlap between the two activities, as one needs to use the internet to watch videos online. Both categories consequently also show a high correlation (r = .61).
Frequency of engagement with English-language content online (in %)
Only 13.3% of students state to rarely ever surf on English-language websites, and even fewer students (9.5%) (almost) never watch online videos in English. There is no significant difference between German and Swiss students for the frequency of surfing activities in general but Swiss students tend to watch English-language online videos more frequently (\(\overline{x}\)G = 4.01, SD = 1.16, \(\overline{x}\)CH = 4.12, SD = 1.08; t(2016) = 2.05, p = .04, d = .097), yet the difference is small.
For the follow-up questions, students who visit English-language websites or watch English-language online videos at least a couple of times per week were categorized as surfers (n = 1903).
Hours spent surfing per day. Most surfers (84%) spent one to two hours per day engaging in English-language content online (Table 6.2). Only two percent spent 4.5 to 5 or even more than five hours online. German students spent significantly more hours online per day than Swiss students (\(\overline{x}\)G = 1.98, SD = 1.33, \(\overline{x}\)CH = 1.60, SD = 0.97; t(936.53) = 6.16, p ≤ .001; d = .335).
Popular Websites. Asked about websites in English they like to visit, almost all surfers named watching online videos in English on video-sharing platforms such as YouTube (Table 6.3). Again, this is not surprising and supports the high correlation between watching online videos and surfing reported above. The second most named website categories were social media platforms and search engines. Almost half of the students also named messaging apps like WhatsApp to communicate in English and websites for streaming movies and TV series in English online (for more details on movies, TV series, and TV shows, see also Section 6.1.3).
Students in Switzerland and Germany do not differ for most categories, but Swiss students are more likely to use free streaming websites while German students are more likely to use legal streaming services (ϕ = .28). Apart from streaming services, German students also indicated using fan-fiction websites (ϕ = −.09) more often and are more involved in online gaming (ϕ = −.13). Significantly more Swiss than German students prefer travel and shopping sites (ϕ = .06). Overall, the differences between the two countries ranges from small to moderate in size.
In addition to the closed-ended categories, students were also provided with an open-ended category, to name any other website. Six additional categories could be extracted from the answers: Seven students (0.4%) regularly read books/poems/literature online, four students (0.2%) visit humor sites, five students (0.3%) browse sports-related websites, three students (0.2%) listen to music streaming sites and three students (0.2%) regularly use the internet for scientific research activities. Seven students (0.4% − all but one male) even reported that everything they do online is in English. Although the total numbers for these categories are small, they still provide further evidence for the omnipresence of online English-language media content in the everyday lives of some students.
Popular influencer and Content creators. Surfers were also asked which English-speaking influencers and content creators they were following online. The recording of the answers revealed that the average student follows several content creators, channels, and celebrities online and does so across multiple social media platforms. Many students stated that they could not possibly name all of them. Instead, they limited their answers to some examples. From n = 1,557 valid answers, six main categories were extracted from the data: influencer, channels, celebrities, politics, sport, and TV shows (see Section 5.3.2 for a detailed definition for each category). Answers varied widely and represented the diversity of the industry. For brevity, only names that were mentioned more than ten times will be reported. It should be mentioned that the total number of entries, even for these names, does usually not surpass a few dozen students. The most frequently mentioned name was mentioned 115 times. The results can therefore only be seen as tendencies for students’ preferences. Figure 6.4 presents the most popular content creators as a word cloud. The more frequently a name was mentioned, the larger the font in the word cloud.
Most popular English influencers online
As the word cloud shows, the most popular English-speaking influencers at the time of the data collection were PewDiePie, Casey Neistat, Zoella, and Liza Koshy. PewDiePie was mentioned the most (n = 115), followed by Zoella (n = 69). These three names coincidentally represent four main topics for influencers and content creators in general: gaming, hair/make-up/fashion, vlogging, and comedy. The research showed that all four influencers are active on YouTube and Instagram and, in the case of PewDiePie, the gaming platform Twitch. All of them have also branched out into other businesses (e.g., makeup line), written books, or have started acting.
For channels, Buzzfeed was the most mentioned name (n = 64) (Figure 6.5). Buzzfeed is a professional media company that produces a variety of content via multiple channels for a variety of entertainment topics. They also produce newspaper articles and investigative pieces and maintain a popular YouTube Channel (Buzzfeed). Twenty-one students also regularly watch videos on the channel TedTalks on YouTube, and eleven students report watching the educational channel In a nutshell. These last names underline the fact that YouTube and other social media sites not only provide entertainment but can also serve as a source for information and educational content for adolescents.
Most popular English online channels
Several students also named YouTube channels from popular TV shows (Figure 6.6). The most popular ones are The Ellen DeGeneres Show (n = 39), The Late Late Show (n = 37), The Tonight Show (n = 36), and Last Week Tonight (n = 31). The channels do not provide the complete tapings of shows but rather share clips ranging from interviews with celebrities to commentary and analysis of political topics. While it cannot be said which clips students prefer, these results again point towards the role of online content for education and information among adolescents.
Most popular TV shows via video platforms
The only celebrity mentioned more than ten times in the present sample was Jon Olsson (n = 20), a skier who is very active on social media and had a YouTube channel with 1.5 million subscribers (as of November 2020). For the categories of sports and politics, no name was mentioned more than ten times.
Active and passive online behavior. Asked about their activities, most surfers said they regularly watch videos online and read short comments and posts on social media sites in English (Table 6.4). Only a quarter/a third said they read longer stories, posts, and comments as online activities.
Even fewer surfers named active online activities. While almost all surfers named watching videos online, only three percent said they record and upload videos themselves. Similarly, not even 10% of surfers said they write longer texts and upload them to the internet. Furthermore, only half of the surfers reported communicating with others via social media, and only 40.4% said they regularly exchange emails in English.
Surfers from Germany and Switzerland show similar results. However, more German surfers named reading longer posts or comments in online forums (ϕ = −.09) and writing longer texts and comments or posts online (ϕ = −.08). By contrast, more Swiss students in the sample said they record and upload English videos (ϕ = .05). The differences between the countries are again mostly small in size.
All in all, extramural English contact through online content seems to be dominated by input rather than output production or interaction for most students. This is in line with results from Fraillon et al. (2014), which showed that actively writing and uploading content was less popular among young people than passively consuming input online.
6.1.3 Frequency of Watching Movies, Tv Series, and Tv Shows
Watching movies and TV series is also very popular among students: 73.8% of students engage in it at least 1–3 times a month, 46.9% even watch them multiple times a week or (almost) daily (Figure 6.7). By contrast, English TV shows (e.g., game shows or talk shows) are not as popular, as almost half of the students never watch them. Swiss students watch significantly more TV series and movies in English (\(\overline{x}\)G = 3.16, SD = 1.26, \(\overline{x}\)CH = 3.35, SD = 1.18; t(2011) = 3.24, p ≤ .001, d = .153). However, there is no difference in watching TV shows.
Frequencies of watching English movies, TV series, and TV shows (in %)
From the sample, n = 1,566 students qualified as watchers, as they indicated they watch movies/TV series or TV shows at least 1–3 times per month. These students were shown follow-up questions to determine the time they spent watching audio-visual content.
Hours spent watching movies and TV series. The two open-ended questions for hours spent watching in one sitting and hours spent watching per week were recoded into categorical variables for the analysis. Unrealistic answers (e.g., 100 hours per day) were set to missing.
When sitting down to watch movies and TV series, most watchers watch one to two hours in a row (87%) (Table 6.5). German students on average watch slightly more hours in one sitting than Swiss students (\(\overline{x}\)G = 1.75, SD = 1.04, \(\overline{x}\)CH = 1.60, SD = 0.81; t(610.21) = 2.45, p = 0.02, d = .161).
When asked about the hours they usually watch per week, 72.5% stated they watch up to four hours per week (Table 6.6). German and Swiss students do not differ significantly (\(\overline{x}\)G = 3.71, SD = 2.98, \(\overline{x}\)CH = 3.46, SD = 2.72; t(874.22) = 1.57, p = .118).
6.1.4 Frequency of Reading Books, Magazines, and Newspapers
Figure 6.8 shows that only 39.3% of the students in the sample read books in English at least 1–3 times per month or more. Magazines and newspapers in English seem to be slightly more popular with the students, as 47.2% read them at least 1–3 times per month or more. However, only 7.8% out of these students read magazines every day. Since the questionnaire does not distinguish among genres, it is unclear whether students read an entire issue or only selected articles. However, it seems likely that the shorter overall length of newspaper and magazine articles in English makes them more suitable for a quick read than books written in English.
Swiss adolescents showed a slightly higher frequency for reading books in English (\(\overline{x}\)G = 2.15, SD = 1.14, \(\overline{x}\)CH = 2.47, SD = 1.09; t(1998) = 6.07, p ≤ .001, d = .286) and reading magazines and newspapers in English (\(\overline{x}\)G = 2.29, SD = 1.25, \(\overline{x}\)CH = 2.53, SD = 1.32; t(1388.19) = 3.93, p ≤ .001, d = .182).
Frequencies of reading English books, magazines, and newspapers (in %)
Number of English books read per year. Since a book is usually not read within one sitting, it must be assumed that students will need multiple reading sessions for each book. The n = 1,507 regular readers were asked about the number of books they read per year. Table 6.7 shows that almost half of the readers indicate that they read two to three books per year, and almost 20% even read up to five books.
Swiss students, on average, read more books than German students, who are more likely to only read one English book or less per year. Yet, German students also surpass Swiss students who read seven or more English books (\(\overline{x}\)G = 2.24, SD = 1.23, \(\overline{x}\)CH = 2.45, SD = 1.13; t(1505) = −3.159, p ≤ .002; d = .179).
6.1.5 Frequency of Gaming
Gaming in English is the least frequented category by students overall. Only 27.4% of students state that they engage in the activity at least 1–3 times per month, and only 19.7% do so every day (Figure 6.9). As a result, only n = 965 students qualify as regular gamers. This is surprising given that the questionnaire also included smartphone gaming apps, which are designed to be played on the go. Consequently, a higher frequency might have been expected.
German students tend to engage in gaming activities in English slightly more often than Swiss students (\(\overline{x}\)G = 2.84, SD = 1.65, \(\overline{x}\)CH = 2.48, SD = 1.56; t(1265) = 4.7, p ≤ .001, d = .225).
Frequencies of gaming in English (in %)
Hours spent gaming. The two open-ended follow-up questions about time spent gaming in one sitting and hours spent gaming per week were recoded into two categorical variables for the analysis. The majority of regular gamers (84.8%) spent up to two hours gaming in one sitting (Table 6.8). German students show a slightly higher average of hours played in one sitting (\(\overline{x}\)G = 1.88, SD = 1.29, \(\overline{x}\)CH = 1.58, SD = 1.01; t(623.76) = 3.84, p ≤ .001, d = .271).
When looking at the hours spent gaming per week, the data shows that half of the gamers spent one to two hours per week with the activity; however, high-frequency users spent 10 hours or more per week. No other category shows such a high percentage of high-frequency users (Table 6.9). While the two countries are similar in their percentages of low-frequency users, the data shows a significantly higher percentage of high-frequency users in Germany, resulting in a significant difference between the two countries (\(\overline{x}\)G = 4.87, SD = 4.06, \(\overline{x}\)CH = 3.41, SD = 3.42; t(669.49) = 5.71, p ≤ .001, d = .390).
Popular gaming genres. As another follow-up question, gamers were presented with 12 gaming genres and asked to name the genres they regularly engage in (see Table 6.10). The most named genres were action games and first-person shooters. Additionally, roughly a third of the students named adventure games, multiplayer online role-playing games, and strategy games. The least named category was rhythmic games. As these games often require special equipment, one reason for their low popularity might be their high purchase threshold.
Students in both countries only differed in four of fifteen genre categories: German students overall showed a higher preference for adventure (ϕ = .12), multi-player-online- role-play (ϕ = .11), simulation (ϕ = .14) and horror games (ϕ = .09).
Active and passive gaming behavior. Online games allow for different levels of interaction with other players. Opportunities for interactions range from single-player mode without any interaction, simple written chatrooms, to audio and even video chats. Some of the most interactive online games are organized as community games in which players form long-lasting groups with others (multiplayer online role-play). Therefore, gamers were asked if they only passively engage in gaming or make use of communication channels to talk to other gamers during the games. The results show that both passive and active gaming behavior in English is only carried out by a third of the gamers in Germany and Switzerland (Table 6.11). More German students stated that they communicate with others in English (ϕ = −.08) and have set their gaming menu to English (ϕ = .01), yet the latter is negligible in size.
6.1.6 Reason for Extramural English Contact
Students who reported regular extramural English contact at least 1–3 times per month were asked why they engaged in these contacts. The question was provided in an open-ended format. Overall, 1,857 valid answers were recorded and coded into nine main categories using content analysis. Students can fall into multiple categories at once, as they could name as many reasons as they wished. Table 6.12 reports the frequencies for each category and subcategory, starting with the most frequent one.
Almost half of the students cited language learning as a reason to engage in English-language content. This category summarizes all statements regarding the possibility or the wish to increase language competences through media-related extramural English contact. This was also the most frequently named reason in the sample and can be illustrated by the following exemplary quote:
“So that I can improve my English. Watching TV series in English is great because you relax and at the same time you passively learn English. [Damit ich mein English verbessern kann. Serien auf English zu schauen ist toll, denn dabei entspannt man sich und zugleich lernt man passiv die Sprache Englisch mit.]” (Case 285)
Following the results from Stecher (2005), it could have been expected that students assign different learning potentials to different media categories. However, even though books were often named in terms of language learning, other media outlets were not explicitly excluded. Indeed, movies and TV series were often named as a good source for vocabulary training, coming into contact with different accents and dialects, and subsequently increasing listening comprehension.
Despite the awareness of learning opportunities, the resulting language learning is most likely still incidental in nature. As discussed in Chapter 4, incidental language learning, while mostly an unconscious process, does not mean that students must be unaware of possible learning processes, yet the emphasis of the activity is on decoding the message or participating in an interaction.
That this is most likely the case also becomes evident by the fact that while these students are aware of the possibility of language learning through media contact, it is almost never the sole reason for the contact. Most of the students in this category particularly stress that they appreciate the opportunity to learn or maintain their language skills while still engaging in a fun and voluntary activity. Since they can choose the media content they want, it is reasonable to assume that their choice of media content is driven by interest and personal relevance. Therefore the choice to engage with English-language media content seems in part to be driven by the desire to combine fun with learning opportunities.
Four students also specifically stressed the opportunity of coming into contact with dialects other than the standard English usually taught in school. While these represent only isolated opinions, they illustrate the variety of English content available for students outside the classroom.
In addition, some students also stated that extramural English contacts have even helped them more than their classroom instructions since they provided more input and additional information. Again, these are isolated statements, yet they illustrate the rich learning opportunities through exposure from extramural contact.
Some students also explicitly stress the importance of being proficient in English for their future and their desire to practice the language as much as possible outside of school:
“Since I live in Germany, I have no other daily contact with the English language. And since language is something living, I try to integrate it into my everyday life so that I stay fit. I think that English as a world language is an important language, which one should know at least a little bit. [Da ich in Deutschland lebe, habe ich sonst keinen täglichen Umgang mit der englischen Sprache. Und da Sprache etwas Lebendiges ist, versuche ich sie in meinen Alltag einzugliedern, damit ich darin fit bleibe. Denn ich denke, dass Englisch als Weltsprache eine wichtige Sprache ist, die man heutzutage mindestens in Ansätzen beherrschen sollte.]” (Case 246).
The second most frequent category was quality. Almost half the students stated that the original audio track provided a better or more authentic experience. Many of the students considered the dubbed German audio track to be of inferior quality and to disturb the viewing experience. Students also stated that meaning was lost during translation in terms of humor, emotion, and atmosphere. Some students mentioned similar problems with translated books or computer games. The following quote illustrates these findings:
“Movies/series/books are often more authentic in English (the original language); a lot gets lost in translation, e.g., jokes that have to do with the culture and especially word plays. [Filme/Serien/Bücher sind oft authentischer auf Englisch (Originalsprache); viel geht verloren z.B. Witze, die mit der Kultur zu tun haben und vor allem auch Wortspiele]” (Case 620)
In addition, some of the students said that English media sources had better quality and provided more information. This was mainly geared towards the English influencer community and online content.
Almost half of the students also mentioned exclusivity as a reason for extramural English contact, as the content they are interested in is only available in English. This seems to be especially true regarding content created by influencers. Since every creator produces a specific type of content, often tied to their personal life, this is not surprising. Videos and posts are not always available in a dubbed or translated version, and an English-speaking influencer is not simply replaceable with a German or Swiss influencer. As a result, their content is unique and exclusively linked to the language it was produced in. The same seems to be true for many websites, posts, and message boards. In addition, students also stated that since there was so much more English content available online, it was easier to find information and material.
Lastly, a small percentage of the students stated earlier access to movies and TV series, but also to books, as an important criterion for engaging in extramural content. Since authentic material does not have to be translated prior to release, it is often available for download or purchase earlier than the translated version. Some students stated that they had started a TV series in German but then got impatient waiting for the release of the next season in German.
While it is reasonable to assume that all students who engage in extramural contact appreciate the English language somehow, more than a tenth explicitly stated that English is ‘fun’ and that they liked to incorporate it into their everyday life (appreciation for English). Some students simply expressed their wish to see the original versions, often after or in addition to the dubbed or translated versions.
Some students reported being encouraged by outside factors to engage in extramural English contacts (external influence). Most of these students reported coming into contact with English outside of school and started to engage in English media contacts while spending time abroad. Some students have English-speaking friends or were introduced to English media content by their parents, family members, or teachers. However, overall, the numbers for these categories are small. It is thus reasonable to assume that this is not necessarily the case for most students.
Since the follow-up question was shown to all students who regularly engaged in any form of media-related extramural contact, respondents to the open-ended follow-up question do not frequent all media categories alike. Seventy students explicitly reported to only read books in English for homework, yet engage in contact with other media categories voluntarily. This might be due to the fact that reading a whole book is a demanding activity, especially for beginners, as could be shown in other studies (Huckin & Coady, 1999; Peters, 2018; Sylvén & Sundqvist, 2015). In addition, nine students reported getting into some forms of out-of-school contact for school-related reasons while also engaging in extramural contact voluntarily, without naming specific media categories.
In summary, the results indicate that students engage in extramural English contact because they appreciate the undubbed or untranslated versions of media content such as books or movies. This is mainly due to the perceived lack of quality of the German translation and because they genuinely like English. Students stressed the omnipresence of English, especially online, and the fact that some content is only available in English or that the English content is of higher quality or more informative. Students also see English as an important factor in today’s world and hope to increase their language competences while engaging in fun leisure time activities.
6.1.7 Patterns of Media-Related Extramural English Contact
As the results in the last section have shown, most students do not only have extramural English contact through one media channel. Instead, students tend to engage in various media activities, best serving their needs and interests. It was thus interesting to see if any patterns emerge from the data.
With the exception of surfing on English-language websites and watching English-language videos online, most media categories only show a small to medium correlation with each other (Table 6.13). As discussed above, the high correlation between these two online activities is not surprising: one has to use the internet to watch videos online.
The low to medium correlation for the other media channels is probably due to the fact that students show a variety of use patterns and preferences. In addition, some media categories can be engaged in simultaneously (e.g., listening to music while reading a book or surfing the internet), and some media categories are popular with almost all students (e.g., music), while others are used by almost no one (e.g., audiobooks).
To investigate if the variables can nevertheless be reduced to a smaller number of underlying latent components, an explorative principal component analysis (PCA) was conducted (see Table 6.14). The results from the PCA show three principal components with an eigenvalue bigger than 1. Component 1 explains 34% of the variance, component 2 explains 13.83%, and component 3 explains 10.19% of the total variance.
The first component could be interpreted as the use of traditional media channels. It captures a considerable part of the variance for listening to audio content (audiobooks and radio), reading books, magazines, and newspapers, as well as watching movies, TV series, and TV shows. However, watching TV shows also shows a substantial cross-loading on the second component, and watching movies shows cross-loading for component 3. In the case of cross-loading, it is helpful to take into account theoretical considerations (Research Methods Consulting University of Zurich, 2018). For the present study, it seems more logical to see watching movies and TV shows as part of component 1, given the closeness of the categories to watching TV series.
Component 2 could be seen as the use of online-based media channels, as it explains most of the variance for online-based extramural English contact via surfing, watching videos online, or gaming.Footnote 1 However, surfing and watching online videos also shows cross-loading on component 3.
Last, component 3 could be categorized as the use of music because music is the only media category with a high factor loading.
While these components might show a tendency for certain usage patterns, the frequent and substantial cross-loading points towards a lack of discriminatory power between the components. Given these non-conclusive findings, the results should be interpreted with caution (Research Methods Consulting University of Zurich, 2018), as it seems there are no clearly distinctive patterns of media usage for the present study.
6.2 Media-related Extramural English Contacts and the Digital Divide
The theoretical and empirical literature presented in Chapter 3 provided evidence that investigating media behavior patterns while ignoring the social structures in which the behavior is embedded would fail to capture the unique social conditions under which such patterns emerge. The present study included two social factors of interest: socio-economic background and gender. In this chapter, students’ frequencies and forms of media-related extramural English contact will be analyzed considering these two factors.
6.2.1 Socio-economic Background and Media-related Extramural English Contacts
For the present study, students from the penultimate year of baccalaureate schools (Gymnasium), the highest secondary educational track, were selected. Studies have repeatedly proven the significantly lower probability for children from lower socio-economic families to receive a recommendation for this highest educational track in both Germany and Switzerland. These differences persist, even after controlling for grades or test performance (see, for example, Angelone & Ramseier, 2012; Buchmann et al., 2016; Frank & Sliwka, 2016; Hußmann et al., 2017; Klemm, 2016; Konsortium PISA.ch, 2019; Kuhl et al., 2013; OECD, 2016, 2020; Solga & Dombrowski, 2009).
Since educational trajectory and a probability of transitioning to the highest educational track thus already depend on the socio-economic family background, it could be assumed that the sample for this study would also be highly selective. Indeed, the data shows that, on average, the parents’ highest level of education in the sample is at least a post-secondary non-tertiary education, either vocational or academic (Table 6.15). Parents from Switzerland have a significantly higher educational level than parents in Germany (t(1826.77) = −9.87, p ≤ .001, d = .421). This is most likely due to the even higher selectivity of the Swiss educational system, in which the entrance to the highest educational track is even more restrictive than in Germany (Keller et al., 2020).
Similar results can be seen for the average number of books at home (indicating cultural and educational capital) in both countries. The data shows that families on average own 100 to 250 books (Overall: n = 2234, \(\overline{x} =\) 5.25, SD = 1.37; G: n = 781, \(\overline{x}\) = 5.30, SD = 1.29; CH: n = 1453, \(\overline{x}\) = 5.22, SD = 1.41). There was no significant difference between families from Germany and Switzerland (t(1724.81) = 1.26, p ≤ .207).
In addition to these two structural factors, the study also measured the use of the English language at home, parents’ English competences, and parents’ attitudes towards English as an important investment into their children’s futures. These process factors can be seen as the operationalization of a family’s incorporated cultural capital concerning English as a foreign language (Rolff et al., 2008).
The data show that most students indicated that English is not used regularly at home. However, most students would agree that their parents have good English competences and see English as an important investment into their children’s futures (Table 6.16). Swiss and German parents only differ significantly in their English competence, with Swiss students indicating a slightly higher competence level for their parents than German students. This might be due to the higher overall education of Swiss parents in the sample.
Process factors and structural factors show a significant correlation with each other (Table 6.17), which is not surprising, as the process factors are themselves already influenced by the level of educational and monetary capital within a family (Rolff et al., 2008). However, the parents' perceived value of English for their children’s future shows only a small correlation with the parents’ highest level of education and objectified cultural capital. This shows that, at least for the MEWS sample, the value of English as an important investment into a child’s future does not vary as much across social class boundaries.
Overall, the results show that students within the sample disproportionally come from families with a higher level of objectified and institutionalized cultural capital and educational and economic resources.
Since the highest level of education (HISCED) and cultural capital within a family are also indications of the monetary situation, it could be assumed that these factors also influence a family’s possibility to afford state-of-the-art technical equipment. It is reasonable to assume that while it might have become more affordable for families to provide each child with a smartphone and provide at least one computer per family, providing each child with a personal computer or providing stable internet access at home might not be possible for every family. In addition, educational and cultural capital might influence the technical equipment a family is willing to invest in. However, logistic regression models only showed significant effects for the probability of a student owning a gaming console (Χ2 (2) = 59.44, p ≤ .001, R2 = .047, n = 1793) for both highest educational level (OR = 0.86, 95%CI [0.799 – 0.934]) and objectified cultural capital (OR = 0,805, 95%CI [0.740 – 0.877]). Here, children from families with a higher educational and cultural capital have a lower probability of indicating ownership of a console.
This result can be read as an indication of the diminished importance of educational and financial capital for students’ access to technical equipment and the internet. The negative effect for gaming consoles could be interpreted as the result of a less favorable attitude towards gaming as a free time activity for children from families with a higher socio-economic level, even if the monetary resources would allow them to buy a console. This can, in turn, be seen as the first support for the hypothesis of the existence of a class-specific media habitus.
To determine the effects of the socio-economic background factors on extramural English contacts through media, regression analyses were used for each individual media category and the overall media index (Table 6.18). Since Rolff et al. (2008) could show that the effect of the structural factors are most likely mediated by the process factors, a stepwise regression analysis was chosen. For each dependent variable, the first model (M1) included only the two structural factors, parents’ highest educational level and objectified cultural capital (operationalized through the number of books at home). In a second step (M2), the three process factors use of English within the family, parents’ English competence, and parents’ perceived importance of English as a school subject were included in the analysis. If findings from Rolff et al. (2008) hold true for the present sample, the two structural factors should become non-significant after introducing the process factors into the regression model.
The results show that students’ socio-economic background does indeed influence the frequency of extramural contact through most media categories as well as their overall frequency of media-related extramural English contact. Similar to findings in Rolff et al. (2008), the structural factors do not show a significant effect for most media categories after the three process factors are included in the model. The only exception can be found for the level of objectified cultural capital on the frequency of reading books in English. Here, a small to a medium positive effect of objectified cultural capital on the frequency of reading books was found, even after controlling for all three process factors. The effect is, however, not surprising since the number of books at home can also be seen as an indicator of parents’ reading habits: Children from families in which literacy plays an important role are also more likely to enjoy reading in English.
For gaming, the first model shows a significant effect for the parents’ highest educational level and is therefore in line with other empirical findings indicating a more restrictive parenting style from parents with higher educational backgrounds, as gaming is usually seen as a negative leisure time activity (Graham, n.d.). However, surprisingly the second model did not show any significant effects. Overall, the data from the present study show little to no effect for socio-economic background factors on extramural English contacts through gaming. This might be partly influenced by the low number of frequent gamers in the sample.
Neither structural nor process factors showed a significant effect for listening to English music in either model. This was to be expected, given that the data shows little to no variation, as almost all students listen to English songs (almost) daily.
Among the process factors, English use within the family shows the most consistent and strongest effects on students’ frequency of extramural English contact across most media categories. It positively influences students’ frequency of listening to English radio programs or audiobooks, reading books, reading newspapers and magazines, watching movies, TV series, and TV shows, surfing on English-language websites, and watching English videos online. This finding is again in line with hypothesis H2.3 and illustrates the important influence of cultural praxis and the media habitus in regard to English media content within a family.
Parents’ perceived importance of English as an investment in their children’s future also shows small positive effects on students’ frequency of extramural contact via radio, podcasts, books, newspapers, magazines, movies, TV series, surfing, and watching online videos. Surprisingly, parents’ English competence has a small negative effect on students’ frequency of listening to English radio or podcasts. However, without a more detailed follow-up question, it is hard to determine the exact nature of this effect.
For the overall index, English use at home and parents’ perceived importance of English also shows a significant positive effect. This again supports the notion that parents serve as role models in shaping their children’s aesthetic taste and relationship with foreign media sources.
As discussed above, the Swiss sample, on average, has a slightly higher socio-economic background level, with students reporting slightly higher educational and cultural capital at home, as well as a higher English competence among parents. Therefore, it was interesting to see if the differences in extramural English contact via certain media channels between Germany and Switzerland reported in Section 6.1 still hold after controlling for students’ socio-economic background. To this end, regression analysis for each media category and the media index was performed using country as a dichotomous dependent variable while controlling for all five socio-economic factors (Table 6.19). However, it should also be noted that including the country variable at the individual level is not without risks (Snijders & Bosker, 2012). The results should therefore be interpreted with caution.
Results showed that even after controlling for students’ family background, both countries showed the same significant differences in the average amount of extramural English contact via books, audiobooks, newspapers/magazines, movies/TV series, online videos, and gaming. Thus, the differences between the two countries for these media categories cannot solely be attributed to differences in the sample's socio-economic background factors.
Including both socio-economic background factors and country of residence into one model also resulted in small differences in the effects of the background variables. After controlling for country, parents’ English competence now shows a significant effect for (audio-)books, newspaper/magazines, movies/TV series, TV shows, online videos, surfing online, gaming, and the overall media index.
Despite these small changes, the results still confirm the fact that language habits within the family seem to influence students’ media-related extramural English contacts, even after controlling for differences between countries. The weight of this influence might differ in different national contexts.
All in all, the results show that the habitus towards the use of English as a foreign language within the family has a significant influence on students’ extramural English contacts. The results thus underline the importance of parents as role models for their children in shaping their media habitus. The effects are apparent not only for traditional media contact via books, but also for newer forms of audio-visual contact and interactive contact via online platforms.
The follow-up questions also further support these results. For the follow-up questions, multiple linear regression models and logistic regression models were calculated to determine the effect of the structural and process factors. Since the previous analysis only showed small to marginal differences between the two countries and the analysis in Table 6.19 showed that including country in the analysis does not seem to dramatically change the overall picture of the relationship between socio-economic background and overall frequency of media-related extramural English contacts, the analysis for the follow-up questions will not differentiate between the German and Swiss subsample.
Number of English books read per year. Even after including the three process factors in the regression analysis, the results again show a significant effect for the objectified cultural capital in the form of books at home (Table 6.20). In addition, the use of English within the family again also shows a positive effect on the number of English books read per year. These findings are in line with the findings for the overall frequency of reading reported above.
Hours spent with extramural activities. Additional stepwise regression analysis showed almost no influence of socio-economic background factors on students' actual amount of time with most media categories (Table 6.21). The perceived value of English as an important investment into a child’s future and the use of English within the family both show small significant positive effects on students’ hours surfing on English-language websites. This once again supports the hypothesis that parents are role models for their children’s behavior when it comes to extramural English contacts.
Parents’ highest level of education and language competence show negative effects on students’ hours spent surfing per week. The same is true for the negative effect of English use within the family on the hours spent with extramural gaming activities. Again, this probably represents the negative effect of a higher educational background in general on surfing and gaming that was also found in other studies (e.g., Graham, n.d.; MPFS, 2017; Waller et al., 2016). However, without additional information on parental restrictions on children’s gaming and surfing activities, this cannot be conclusively proven with the given dataset.
Popular websites. Separate logistic regression models for types of preferred websites only showed significant effects for students’ prevalence for eleven website types (Table 6.22). Once again, English use within the family had the most consistent effect on all categories. Apart from gaming, the language habitus within the family does significantly increase students’ probability to engage in extramural English contact through these eleven website categories. The same is true for parents’ perceived importance of English for information websites, shopping websites, and communicating via messaging apps. This finding is in line with hypothesis H2.3.
Objectified cultural capital also showed a significant effect on students’ chances to engage in literacy-based activities, again supporting the hypothesis that cultural capital fosters literacy-related contacts above and beyond the indirect effect via the three process factors.
Parents’ English competence again shows negative effects for some of the website categories while showing no significant effect for others.
Popular gaming genres. Only eight of the fifteen gaming genres showed significant effects for at least one socio-economic background factor. Except for puzzles, quizzes, and strategy games, a higher socio-economic background decreases gamers’ likelihood of engaging in each gaming genre (Table 6.23). These results confirm findings from other studies, which show that children from higher socio-economic families usually show less frequent gaming activities. However, the exceptions for puzzles, quizzes, and strategy games indicate that a high educational family background might motivate students to choose games with a level of educational purpose or cognitive challenge.
Frequency of active and passive media use. Table 6.24 shows the results for active and passive extramural English contacts in regard to the five socio-economic background factors. As for other follow-up questions, socio-economic factors seem to play only a minor role in students’ choice to engage passively or actively in extramural contacts. Parents’ English competence again shows a negative effect on some of the media activities. In contrast, the use of English within a family and parents’ perceived importance of English tends to positively affect most categories, increasing students’ chances of engaging in active and passive extramural activities. An exception is, once again, the probability of engaging in communicative behavior while gaming online.
Summary. Overall, the results can be read as an indication of the importance of the family environment for children’s use of the language outside of school. If the presence or usage of English is a regular occurrence within a household and is frequently used by parents and children alike to communicate or engage in leisure-time activity related to EFL, children will grow accustomed to a daily life in which English plays a significant role. As a result, they are more likely to engage in numerous extramural activities themselves.
6.2.2 Gender and Media-related Extramural English Contacts
As discussed in Chapter 2, female adolescents have been shown to be less active online, watch fewer online videos, and spend less time playing computer games. At the same time, male students tend to read less and use the internet for less communicative purposes.
However, studies have also shown that female adolescents have caught up in terms of technology use and ownership. In the present study, male and female students also do not differ much regarding ownership or access to the internet and technical equipment (Table 6.25). Nevertheless, a slightly higher percentage of female students owned a personal computer to study (ϕ = .07), while more male students had access to a gaming console (ϕ = –.097).
Looking at the frequency of extramural English contacts, Figure 6.10 and Table 6.26 show that male and female students do not differ significantly in their frequency of extramural contact through music, radio/podcasts, audiobooks, and movies/TV series. However, male students can be shown to be more active gamers and be more engaged with English content online. On the other hand, female students read English books at least a few times per month (45.6%), while only 30.1% of male students indicated the same.
(Note: Significant differences between male and female students indicated by asterisk; p ≤ .05)
Frequency of extramural English contact through media by gender (in %).
Female students also stated that they read more English books per year in the follow-up question (\(\overline{x}\)M = 2.15, SDM = 1.04, \(\overline{x}\)F = 2.52, SDF = 1.21; t(1248.72) = −6.31, p ≤ .001, d = .321). By contrast, male students show a higher average for reading magazines and newspapers in English and watching TV shows. Male students were also shown to be more active online, surfing on English-language websites and watching online videos.
In line with other empirical findings, male students also showed a higher frequency of extramural English contact through computer games than female students. The effect size for this category is the strongest of all media categories. This difference becomes even clearer when looking at gamers who play multiple times per week or daily (frequent gamers): while 58.4% of the male students in the sample are frequent gamers, only 17.7% of female students fall into the same category; the difference is significant and again shows a strong effect size (ϕ = −.42, p ≤ .001).
Given the higher involvement of male students in six out of ten media categories, it is no surprise that the t-test for the overall frequency of media-related extramural English contact also shows a significant difference between male and female students, with male students reporting a significantly higher contact frequency. The effect size is medium. So far, the results do support hypothesis H3.1 and H3.2. The follow-up questions further support the hypothesis of gender stereotypical media behavior in regard to English-language media content.
Hours spent with extramural activities. Follow-up questions revealed that male students spend more total hours online than female students (\(\overline{x}\)M = 1.81, SD = 1.15, \(\overline{x}\)CH = 1.67, SD = 1.09; t(1838) = 2.66, p = .008, d = .125), play computer games for a longer period of time in each sitting (\(\overline{x}\)M = 1.90, SDM = 1.24, \(\overline{x}\)F = 1.30, SDF = 0.77; t(907) = 9.14, p ≤ .001, d = .530), and spend more hours per week with these games (\(\overline{x}\)M = 4.94, SDM = 3.98, \(\overline{x}\)F = 2.20, SDF = 2.41; t(934.97) = 13.19, p ≤ .001, d = .733).
Popular websites and gaming genres. The data shows significant gender differences for eleven of the fourteen website categories (Table 6.27). Significantly more female students indicated that they visit social media sites, messaging apps, fan fiction communities, and blogs regularly. This is in line with females' more communicative online behavior and their stronger interest in reading found in other studies. However, these findings might also be driven by the fact that girls and women are often excluded from certain male corners of the internet (Kommer, 2008; Tillmann, 2014).
Male students are once again found to be more active on gaming websites. They also preferred video platforms, magazine and newspaper websites, and they named forum and message boards more often than female students.
In regard to gaming genres, the data showed significantly more female students who indicated puzzles, quizzes, card and gambling games, and real-life simulations (e.g., The Sims) as preferred game categories (Table 6.28). Male students demonstrated a clear tendency for sports, action, and speed-driven games, as well as for first-person shooter games and multiplayer online role-playing games. These findings are again in line with findings from other studies and with H3.2
Frequency of active and passive media use. The data shows that significantly more male surfers indicate to engage in active content production online, such as recording and uploading their own videos in English and writing longer posts and comments in online communities (Table 6.29). However, in line with findings from other studies, female surfers are more likely to write longer creative texts, such as fan fiction or poems.
In addition, while half of the male gamers indicated to regularly communicate with others during a game session, only 6.6% of female gamers stated the same. This is most likely due to the fact that female students tend to favor non-communicative games, such as puzzles, quizzes, and simulations (see above). In contrast, male students showed a stronger preference for gaming genres that afford more opportunities for interactive gaming, such as multiplayer online role-playing. As a result, female students not only spent less time gaming overall; if they play, they show a preference for non-communicative gaming categories.
Summary. Overall, the results showed that adolescents in Germany and Switzerland show gender-stereotypical media behavior in regards to English media content. Male students show a higher overall frequency for extramural English contact and strongly prefer game and audio-visual-driven content. In contrast, female students show a more communicative media behavior but seem to be more reluctant to engage in real-time communication through gaming environments or upload their own video content. Together with the more active and communicative behavior, these results might indicate an advantage for male students in terms of incidental language learning opportunities through these extramural English contacts.
6.3 Media-related Extramural English Contacts and Language Competences
The third research question asked how students’ extramural English contacts might relate to and influence their English competences. Following the theoretical framework in Chapter 4, extramural contacts can be expected to be positively correlated to all three language skills and even have a positive causal effect on them (H4).
As discussed in Section 5.2, the present study will assume a multi-dimensional model for the three language skills (Jude et al., 2008; Schoonen, 2019). This also allows for a separate analysis of the effect of media-related extramural English contacts on students’ reading, writing, and listening skills. Table 6.30 shows the correlation for each media category as well as the overall media index with all three language skills at T2.
Music only shows a small significant correlation with all three skills. Given the small variation in the data set, this was expected, as almost all students listen to English music (almost) daily. Similarly, no significant correlation either for reading or listening skills was found with audiobooks and radio or with watching TV shows, probably because most students do not engage in these media channels. However, watching TV shows and listening to the radio showed a small but significant correlation with students’ writing skills.
Reading books and magazines, watching movies and TV series, watching online videos, surfing, and gaming all show a medium correlation with all three language skills. The same is true for the frequency of extramural English contacts overall.
Together, these correlations show a positive relationship between the amount of time students spent interacting with English via media content and their language competences. It could have been expected that reading English books might show a particularly high correlation with students’ reading skills. Interestingly the correlation with students' listening skills is of the same magnitude, and the correlation with writing is even stronger. The same applies to watching movies, TV series, or online videos. This finding will be discussed in more detail in Section 7.3.
In a next step, structural equation models (SEM) were calculated to further investigate the relationship between students’ language skills and their extramural English contacts. In addition, such a model also allows a further investigation of the causal direction of the relationship.
In contrast to linear regression models, SEM allows for the integration and simultaneous testing of multiple linear regression models. They thus also allow the integration of gender and socio-economic background factors into the model. In this way, it will be possible to integrate the results of the previous chapters and combine them with the analysis of possible learning effects. The proposed path model can be seen in Figure 6.11.
Proposed path model. (Note: Socio-economic factors summarized for brevity)
As the IRT scaling for the three language skills was performed separately, all variables in the following analysis are manifest. Separating the steps of IRT scaling for the test scores and SEM analysis is common in large-scale assessments (Jansen et al., 2016). Consequently, the models discussed below do not include measurement models.
All models were conducted in MPlus Version 8 (Muthén & Muthén, 1998–2017). Full information maximum likelihood (FIML) was used to estimate any missing information in the independent variables. The hierarchical structure, i.e., students clustered in classes, was accounted for by using the type = complex and cluster = class command. Controlling for class also helps to adjust for differences in class composition due to differences in educational track selectivity between the two countries. For potential drawbacks of this method, see Section 7.3.
For language competences, information from all 15 PV data sets was used by employing the type = imputation command in MPlus to average the final results (Rubin, 1987).
The first step was a baseline model (M0) with only gender and social background factors regressed on all three language skills. As already mentioned above, multiple studies have shown female students to reach significantly higher test results and better in-school performance in terms of foreign language learning (Hartig & Jude, 2008). Similarly, studies have also continuously shown that students from higher socio-economic backgrounds attain significantly higher test results in national and international studies (see, for example, Angelone & Ramseier, 2012; Buchmann et al., 2016; Frank & Sliwka, 2016; Hußmann et al., 2017; Klemm, 2016; Konsortium PISA.ch, 2019; Kuhl et al., 2013; OECD, 2016, 2020; Solga & Dombrowski, 2009). As has been discussed in Section 3.1, similar results were also evident for English as a school subject (Rolff et al., 2008). This can also be expected for the present sample.
The model fit statistics suggest an adequate fit of the model, although the Chi2 test for the model is significant (χ2/df = 64.798/ 5, p ≤ .001). As the null hypothesis for the Chi2 test assumes a perfect fit of the model in the population, a significant test suggests that the model does not fit perfectly. However, given the Chi2 test’s sensitivity to large sample sizes, this result was not surprising. Under such circumstances, researchers are recommended to take into account other measures of model fit in order to better assess their structural equating models (Geiser, 2010; Schermelleh-Engel et al., 2003). The other model fit statistics show an adequate fit for the baseline model (CFI/TLI = .984/ .892; RMSEA = .069; SRMR = .019), yet the results should be interpreted with caution. Path coefficients are reported in Figure 6.12. For readability, only significant effect paths are reported.
Path analysis: the effect of socio-economic background and gender on students’ language skills (M0). (Note: Averaged across all 15 PVs, n = 2,487; standardized coefficients are reported; only significant effect paths are shown (p ≤ .05))
Overall, the model explained 2%–4% of the variance in the three language skills. Results replicate findings from other studies, yet a few effects are worth mentioning: In line with findings from Rolff et al. (2008), the two process factors use of English within a family and parents’ English competence are influenced by the structural factors for educational and cultural capital (HISCED and number of books at home). As already discussed in Section 6.2.1., this was to be expected since parents who hold higher educational degrees are more likely to have higher English competences and are more likely to use English at home. Additionally, and also similar to results in Section 6.2.1., parents’ perceived importance of English was only significantly influenced by the number of books at home. This might be due to sampling. As students attending the Gymnasium are usually expected to go on to tertiary education, it is likely that all parents place a high value on English for their children’s future. It is therefore also not surprising that parents’ perceived importance of English shows a significant effect on all three language skills. By contrast, parents’ own educational background does not show a significant direct effect on students’ language skills. These results once again support findings from Rolff et al. (2008), which showed that while controlling for language practice at home, monetary resources and institutionalized educational titles did not significantly influence students’ language achievements. However, contrary to Rolff et al. (2008), the results showed a significant effect for the cultural capital on all three language skills.
Surprisingly, no effect could be found for parents’ English competence or the use of English at home on students’ language skills (with the exception of listening). This might be due to the selective sample in the present study and the resulting small variance in social background. The effect of English use at home on listening is most likely explained by the fact that the activities captured by the index focused more on listening activities than on reading and writing. Interestingly, there was also no significant effect for gender on any of the three language skills.
In a next step, students’ frequency of extramural contacts was included in the model (M1—Figure 6.13). Model fit statistics for M1 again showed an adequate fit (χ2/df = 63.944 / 5, p ≤ .001; CFI/TLI = .985/ .874; RMSEA = .069; SRMR = .018).
The model once again shows the expected effect of the two structural factors on the use of English within the family and parents’ English competences. The number of books at home also shows a significant positive effect on students’ test achievement. However, the highest level of education in the home does not. The model also shows a significant effect of parents’ perceived importance of English as an investment in their children’s future on writing, but not for the other two skills. By contrast, neither the parent’s own language competences nor English use at home show a significant effect on any of the language skills.
Path analysis: the effect of extramural English contact on students’ language skills (M1). (Note: Averaged across all 15 PVs, n= 2487; standardized coefficients are reported; only significant effect paths are shown (p ≤ .05))
After including extramural English contacts, the model showed the expected effect for gender. Female students receive significantly higher scores on all three language domains. The missing effect in M0 might thus be due to the higher frequency of extramural contacts of male students. After controlling for these contacts, female students show significantly higher achievements once again. However, the effects are small.
In terms of extramural English contacts, male students were shown to be more likely to engage in extramural contact than their female counterparts. This supports the findings from other empirical studies and is in line with hypothesis H3.1. In line with hypothesis H2.3, the results also showed that a conducive home environment and parents’ perceived importance of English as a school subject have a significant positive effect on the frequency of extramural contacts. As before, the small negative effect of parents’ English competence is probably due to the fact that parents with a higher educational level tend to be more restrictive when it comes to their children’s media use. As expected, the two structural factors only show an indirect effect mediated through the process factors.
Overall, the results support the hypothesis that language and media practices within the family are more important for students’ own media habits than institutionalized and objectified educational and cultural resources. Gender and family background combined explain 14% of the variance found in the frequency of extramural English contact.
In line with hypothesis H4, the frequency of extramural English contacts had a significant positive effect on students’ test scores in all three language domains. The more often students engage in voluntary contact with English outside of the classroom, the higher their test achievements in reading, writing, and listening. Including the frequency of extramural English contacts in the model significantly increased the size of the explained variance for the three language skills to 7%–11%.
In M1, the effect of the social background and gender is partly mediated through students’ extramural English contacts. Therefore, it can be argued that students from more conducive home environments and a higher educational background engage in more extramural English contacts, and as a result, they benefit more from the incidental learning processes activated through these contacts. Similarly, as male students engage in more extramural contact, they benefit from more frequent incidental learning processes outside of school.
A longitudinal research design was employed for the present study, with language competence measured at two time points. While the nature of the non-experimental research design and the limited number of time points prevent a definitive proof of causality, students’ language competencies at T1 could be included in a third step to control for prior knowledge (M2—Table 6.9).
As family background and gender preceded performance at T1, M2 models the direct effect of socio-economic background and gender on language competences at T1. Media-related extramural English contact was measured at T2. As questions about extramural language contacts were not specifically directed to the time between T1 and T2, the relationship between extramural contacts and language competences at T1 is conceptualized as a correlative effect in the model. The following results can thus be understood as the effect of extramural contacts on gains in all three language skills between T1 and T2.
Model fit statistics showed a good fit of the model to the data (χ2/df = 112.98/ 23, p ≤ .001; CFI/TLI = .989/ .964; RMSEA = .04; SRMR = .02). After controlling for language competences at T1 the model explains 30% to 53% of the variance for the language skills at T2. This is unsurprising, given the high correlation between test achievements at T1 and T2 (Figure 6.14).
(Note: Averaged across all 15 PVs, n= 2487; standardized coefficients are reported; only significant effect paths are shown (p ≤ .05))
Path analysis: the effect of extramural English contact on students’ language skills under the control of prior knowledge (M2).
Cultural capital again shows a positive effect on students’ test achievements in all three skills at T1. Educational background, however, does not, nor does gender. The effect of the two structural factors on students’ test achievement is again mediated through the three process factors. The perceived importance of English shows a significant positive effect on all three test scores at T1. In addition, the use of English within the family had an effect on both listening and writing skills at T1.
The model also finds significant effects for family background and gender on students’ frequency of extramural English contact. However, the results show the complex interconnectedness of these social dimensions in terms of media-related extramural English contact. Chapter 7 will discuss and interpret the results in more detail and address important limitations of the present study.
Regarding the effect of extramural English contacts on language competences, the results show that even after controlling for T1, extramural contacts still have a significant positive effect on language gains between T1 and T2. The effect size is now small to negligible. However, correlations with language skills at T1 show medium-sized effects.
Notes
- 1.
Even though, as we have demonstrated, online based platforms like Netflix have become increasingly important for movies, TV series and TV shows.
Author information
Authors and Affiliations
6.1 Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.
The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Copyright information
© 2023 The Author(s)
About this chapter
Cite this chapter
Krüger, M. (2023). Results. In: Media-Related Out-of-School Contact with English in Germany and Switzerland. Springer VS, Wiesbaden. https://doi.org/10.1007/978-3-658-42408-4_6
Download citation
DOI: https://doi.org/10.1007/978-3-658-42408-4_6
Published:
Publisher Name: Springer VS, Wiesbaden
Print ISBN: 978-3-658-42407-7
Online ISBN: 978-3-658-42408-4
eBook Packages: Education and Social Work (German Language)