Exploring L2 TV mode preferences and perceptions of learning

This study explores the effects of extensive exposure to L2 TV series, describing perceptions of learning from this type of input. A total of 136 university learners of English participated in a pre-/post-test design study which included extensive watching of ten full-length episodes of an English TV series. There were three groups: captions, textually enhanced captions, and no captions. Learners’ viewing mode preference and feeling of learning from L2 media were explored through questionnaires on out-of-class exposure to English media distributed at the beginning and end of the classroom intervention. The study results suggest a shift in the preferred mode of exposure to L2 media after the intervention and provide insights on the students’ impressions of and reactions to language learning from L2 TV series.


Introduction
Extensive exposure to L2 media has positive effects on several aspects of language, such as vocabulary (e.g. Suárez & Gesa, 2019), comprehension (e.g. Rodgers & Webb, 2017), and grammar constructions (Pattemore & Muñoz, 2020). It is, however, somewhat unknown how foreign language learners perceive these positive effects of L2 TV viewing. Captions (L2) and subtitles (L1) have been found to promote the feeling of learning for vocabulary, expressions, spelling, and accent comprehension (Vanderplank, 1988). Likewise, Montero Perez, Peters, Clarebout, and Desmet (2014) demonstrated feeling of vocabulary learning from the availability of captions and textually enhanced captions (bolding, highlighting, capitalizing). However, Sydorenko (2010) reported participants' feeling they had learned less due to lack of feedback. This might be affected by the viewing mode students are exposed to, as in Pujadas (2019), where students exposed to captioned TV series noted better vocabulary learning than the L1 subtitles group.
Preference for captions over no captions has been found for better content and listening comprehension (Montero Perez, Van Den, Noortgate, & Desmet, 2013); although, students might still prefer L1 subtitles in their daily exposure to L2 TV (Kusyk & Sockett, 2012

Method
The study involved 136 Catalan/Spanish bilingual undergraduate students from four intact classes. The participants' mean age was 19 and their proficiency varied from A1 to C2, with a mean of B2 according to Common European Framework of Reference for languages (CEFR) levels. The classes were randomly assigned to three different viewing conditions: with captions (n=71), without captions (n=27), and with enhanced captions (n=38).
The students watched ten episodes of The Good Place TV series (Schur, 2016) across five weeks. Before watching, the participants took a pre-test with grammar constructions (e.g. I wish I had -see Pattemore & Muñoz, 2020) and a viewing habits questionnaire on their out-of-class L2 media viewing preferences and feeling of learning. After watching ten episodes (227 minutes), they took an immediate post-test and completed a follow-up questionnaire on feeling of learning from The Good Place as well as the viewing habits questionnaire (the same as at the beginning of the intervention).
The questions explored in this paper are:  Figure 1 displays the distribution of answers to the first question above in percentages; the participants could choose more than one affirmative answer. Students reported learning mostly expressions and vocabulary, but not so much grammar or pronunciation, and 19% responded they did not know if they actually learned anything.

Figure 1. All participants' feeling of learning
As for the second research question, also expressed in percentages, students perceived learning vocabulary and expressions more than grammar and pronunciation regardless of the in-class viewing mode. The captions group were most likely to report not perceiving any learning (see Figure 2).

Figure 2. Feeling of learning per treatment group
Concerning the third research question, the students' preferred mode of watching changed after the intervention (see Figure 3). While there was little change in viewing with L1 subtitles (56.45% before the intervention, 47.54% afterwards), there was a noticeable drop in the use of L2 captions (from 72.58% to 11.47%) and a remarkable increase in the preference for watching L2 media without captions (from 27.41% to 59.01%).

Figure 3. All participants' viewing mode preference
Finally, regarding the fourth research question, the results showed that both the captions and enhanced captions groups sharply decreased their use of L2 captions and increased their preference for watching without captions. Likewise, the no captions group greatly decreased their preference for use of L2 captions, but only slightly increased watching without captions (see Figure 4).

Figure 4. Attitudes toward viewing mode before and after the intervention
These results suggest that the students mostly perceived learning expressions and vocabulary. Perhaps expressions and vocabulary are easier to notice than grammar and pronunciation. Also, as in Kusyk and Sockett's (2012), and Vanderplank's (1988) studies, these are the linguistic aspects learners tend to pay more attention to in out-of-class exposure, so it is unsurprising they are also the ones perceived as most learned in this study. Grammar is the aspect felt to be learned least, although the analysis of the language outcomes of the intervention showed significant pre-/post-test grammar gains (Pattemore & Muñoz, 2020). A tentative explanation for this incongruity is that incidental grammar learning is less noticeable without additional feedback or practice, as suggested by Sydorenko (2010). Lastly, the enhanced captions group perceived more learning than the others, including grammar. This is likely due to the increased salience provided by caption enhancement.
Regarding students' viewing mode preferences, students who preferred L1 subtitles before the classroom exposure continued watching with L1 subtitles after the intervention. They may have faced some difficulties using L2 captions or no captions in the classroom and therefore L1 subtitles were a more convenient way of watching L2 TV outside of the classroom. On the other hand, those who were originally watching with L2 captions tended to switch to no captions regardless of their in-class viewing mode. This echoes Vanderplank (2019), where participants turned captions off after a period of time. Our participants in the L2 captions and enhanced captions groups might have become comfortable with L2 audio-visual input and captions could have felt redundant for them. The without captions group also decreased their use of L2 captions in their out-of-school L2 media watching. As this group was exposed to a challenging viewing mode over an extensive period, students may have adapted to watching without any textual support and felt confident enough to view L2 TV in this mode.

Conclusions
This classroom intervention can be considered an overall positive experience as most students felt they were learning from this extensive exposure to L2 TV series. Both expressions and vocabulary were perceived to be learned the most, probably because they were more noticeable, especially for the enhanced captions group who had the strongest feeling of learning. This extensive viewing experience might have fostered the students' confidence as active viewers and foreign language learners given the general switch from L2 captions to no captions.
For further research, we hypothesize that the learner behaviors presented here might depend on learners' proficiency levels, as well as the use of learner viewing strategies.