Language Learning through Social Networks: Perceptions and Reality

Language Learning Social Network Sites (LLSNS) combine tutorial software and opportunities to interact with others in order to improve one’s foreign language competence. However, details on how learners use these platforms and what they learn from them are rarely found in the research literature. This survey study by Chin-Hsi, Warschauer and Blake (2016) investigated learners’ attitudes towards, usage of, and progress made on Livemocha, a popular LLSNS. The results show the potential of these sites for language learning, but also some key questions that need to be answered before LLSNS-use can bring about real success.


Users of LLSNS, such as Livemocha, iTalki, Lang-8, Hello-Hello, Duolingo, or Palabea, have generally optimistic views about such communities (especially those that promise interaction with native speakers!), though some studies have found decreases in use over time. Also, the research literature does not provide much insight into the actual progress language learners make in their L2 by participating in a LLSNS. This study thus set up the following research questions:

1. What were users’ attitudes toward L2 learning on a large LLSNS?
2. What patterns of usage emerged from LLSNS participation?
3. How much did LLSNS users think they learned? What actual L2 improvement occurred?

This study focused on Livemocha (now owned by Rosetta Stone), which had over 16 million users worldwide in 2013 (Livemocha, 2013). The site provides over 160 hours’ worth of free instructional materials in 38 languages, aimed at beginner and intermediate learners. Users complete practice exercises and share their answers for review by other users. By completing exercises and reviewing others users’ work, users gain points which they can use to ‘unlock’ further practice materials. Livemocha also provides premium courses which users have to pay for.

Research Method

For this study,  4,174 users of Livemocha were surveyed and the contributions of 20 case-study participants were analysed. The survey, which was posted on the Livemocha site, included 23 questions and was available in  English, Chinese, Spanish, and Portuguese—languages spoken by 84% of the participants on the site. The case studies focused on Chinese learners of English, who were all using the free resources on the site. These case study participants were interviewed individually, and a total of 253 writing and 275 speaking exercises they had completed were analysed. All of this data included users’ responses to questions on personal details, L2 proficiency, attitudes towards the site, L2 learning goals, learning hours, usage of the site, perceived progress, and, regarding the case study participants, actual progress based on the language accuracy and syntactic complexity of their written work.


What were users’ attitudes toward L2 learning on a large LLSNS?

85% of participants found using Livemocha motivated them to spend more time learning a language on the site.

86% agreed that using Livemocha increased their self-confidence in communicating in their L2.

71% reported feeling more comfortable communicating with native speakers on Livemocha than face-to-face.

Some participants said they felt discouraged by negative feedback from other users, but overall perceptions on this point tended to be neutral or positive, and generally appreciative of the peer-feedback system.

What patterns of usage emerged from LLSNS participation?

Data on usage over time was collected by analysing the contributions of case study participants. These users submitted an average of  26.4 exercises between their registration on the site and 31st December 2011. Over 37% of these exercises were submitted within one month of registration, with a much smaller number of exercises being submitted two months after registration. By eleven months after registration, none of the case study participants submitted an exercise more frequently than once every 3 months; the earliest ‘drop-out’ point was after just four months. There were some individual patterns in usage, with some participants posting again after period of inactivity, but overall only 50% of the case study participants were still active after five months, and only 25% after six months.

Chin-Hsi et al name two key factors which were shown to have a statistically significant impact on extending the period of active use of Livemocha: increased age and more ambitious learning goals.

How much did LLSNS users think they learned? What actual L2 improvement occurred?

Just over half of the surveyed users of Livemocha perceived “a lot of” improvement in their L2 competence, and a further 37% “some” improvement, across the four skills. Survey responses indicated that users perceived the least improvement in grammar, and would wish for more grammar instruction on the site.

Analyses of language accuracy and syntactic complexity were conducted on case study participants’ written exercises to investigate actual improvement.

Participants’ submissions averaged 27.8 words in length and 6.41 words per sentence. The number of errors per T-unit* averaged 2.0, and the number of error-free T-units  3.4 per submission. Detailed analyses showed that participants’ written submissions contained more errors, the longer they used the platform, though the number of error-free T-units was not affected by this. Still, the syntactic complexity of sentences in the users’ submissions did increase, the longer they continued to complete exercises on the platform. Participants who used Livemocha for more hours per week on average made more errors per T-unit and had a lower average number of error-free T-units. The distribution of learning hours between Livemocha and other practice activities did not affect syntactic complexity.

The number of error-free units was not significantly different between participants who self-reported higher and lower levels of proficiency. Indeed, participants who self-reported higher proficiency actually made more errors per T-unit on average. Self-reported improvement in proficiency over time was not reflected in language accuracy or syntactic complexity.

The language learning goals participants had set for themselves correlated with the average number of errors per T-unit; those with more ambitious goals made fewer errors and decreased their errors per T-unit more quickly, though the syntactic complexity of their written work also decreased over time. The average number of error-free T-units was not affected by the participants’ goals.


The boosts to motivation and self-confidence participants in this study reported regarding their language learning highlights the potential of LLSNS like Livemocha in creating opportunities for richer L2 socialisation and engagement than traditional language lessons. Further potential of such platforms can be found in the increased opportunity for authentic langauge production, and more opportunities to monitor one’s own language production than in a traditional classroom. Nonetheless, none of the case study participants here continued to submit exercises to the site after eleven months, with some ceasing to use the platform much earlier. The reasons for this would be an interesting focus for future work in the area. This finding shows, though, that merely having access to learning resources is not a guarantee that learners will use them to their full potential – this is applicable to LLSNS as well as other CALL provisions or MOOCs. Guidance and support from language teachers may be key to maintaining learners’ engagement with such resources over time. An interesting discussion is whether this type of support could be provided within such LLSNS platforms.

This study’s results show worrying decreases in langauge competence by using Livemocha over time. This may be due to the peer-peer nature of much of the feedback received or the non-standard nature of a lot of langauge produced online, or may be less worryingly attributed to learners attempting to use more complex langauge and thus making mistakes in new vocabulary or grammar. Again, these findings warrant further investigation; especially in comparison to users’ perceptions of improvements.

In general, it seems that the main benefit of LLSNS lies in the opportunities for using the L2 in meaningful ways to interact with other users of the site, and the motivation to continue learning that this can bring about. The combination of targeted practice and freer communication reflects and in some ways extends the langauge teaching of a traditional classroom, though the expected improvements in L2 competence were not confirmed by this study. Still, the findings here provide valuable insight for language teachers, and also highlight some interesting and important directions for future research.


*A T-unit is “one main clause with all subordinate clauses attached to it” (Hunt 1965:20).



Hunt, K. (1965). ‘Grammatical structures written at three grade levels’. NCTE Research report No. 3. Champaign, IL, USA: NCTE.

Lin, Chin-Hsi, M. Warschauer & R. Blake. (2016). ‘Language Learning through Social Networks: Perceptions and Reality’. Language Learning & Technology. 20/1. pp. 124-147.

Livemocha. (2013). Livemocha and Rosetta Stone join forces: A letter from Livemocha CEO, Michael Schutzler. Retrieved from

Clare Maas on TwitterClare Maas on Wordpress
Clare Maas
Lecturer in EFL and EAP at Trier University (Germany)
Clare holds post-graduate qualifications from the University of Wales and Trinity College London. Before moving into tertiary education, she taught English at German grammar schools, and English for Specific Purposes at several language academies in the UK and Germany. Her professional interests include EAP materials development and CPD for teachers. She also blogs at

2 thoughts on “Language Learning through Social Networks: Perceptions and Reality”

  1. Hi Philip,
    Thanks for taking the time to comment.

    You’re right – there is lots of information that would be interesting to find out to add some more depth to the findings of this study.
    My feeling is that when people pay for something, they’re more likely to commit more and for longer. Then again, how many people have unused gym memberships…?!

    For me, the conclusion points to some involvement of teachers in encouraging and guiding learners’ use of such apps. but that would then require teachers to test them all and see which ones they believe will be most beneficial for their learners.

    Still, I liked the article in tht it prompts us (teachers) to think about apps and the like, to consider how helpful they might be, and how/whether we could integrate the use of them in our teaching. From what I’ve read here, trying to learn a language just using these apps seems unlikely to be successful in the long run, so at least it makes the case for teachers and gives us hope that we won’t be out of jobs too soon! 🙂


  2. Thanks, Clare.
    The article is freely available online:
    I found the findings about attrition rate particularly interesting. 85% of participants said that using the site helped their motivation. This figure refers to people who have already actively engaged with the site by clicking on the banner advertising the survey. Presumably the number of people whose first interaction with the site was not motivational is higher than 15%, as we would not expect them to respond to the survey call so much.
    Still, 85% found it motivating and submitted a fair amount of work in the first month. By the second month, the overall amount of work being submitted had fallen by a whopping 70%. In the third month, those who were still there maintained their effort. But by the fourth month, there was a further 40% drop. By the fifth month, most had given up, or only worked very sporadically thereafter.
    No one who has been involved in the development of language learning apps will be surprised by this attrition. In recent years, many app developers (like LiveMocha) have tried to keep their users on board with a combination of gamification and social networks. While this may help a little, it doesn’t seem to help a lot.
    What I would like to know is:
    1. whether attrition rates would be lower if users had had to pay for the app
    2. how many and what other free language learning apps these users had tried before and after experimenting with LiveMocha, how long they stuck with these and what their perceptions of them were
    3. whether engagement with LiveMocha had any impact on the users’ language learning behaviours outside of using the app
    4. what users thought about their experience with LiveMocha a year (or more) afterwards
    In many ways, it’s a frustrating article. Its conclusion (“if online education is to play a positive role in the teaching and learning of English and other languages, learners will need support, guidance, and well-structured activities to ensure the kinds of participation and linguistic interaction that can lead to success”) doesn’t exactly tell us anything we didn’t already know!
    But, thanks, again,

Leave a Reply