To cite this article: Sturt-Schmidt, James (2020). Don’t Make the Owl Angry: Assessing Duolingo’s Potential for Improving Student Motivation in MFL, The Millennial [Online], Available at: (this page URL).
This research project aims to explore the potential of the language-learning platform, Duolingo, for improving student motivation in Modern Foreign Languages (MFL) through the lens of expectancy-value theory. The research question it addresses is: Can Duolingo be used to increase student motivation in MFL? It responds through quantitative surveys, qualitative, semi-structured interviews, teacher observation and analysis of online data of students’ activity on Duolingo. The study was carried out in two phases and was adapted to respond to the difficulties and limitations of the first. The findings of this study are positive but non-generalisable, and this research should be read as a pilot study whose findings and limitations may inform more thorough and accurate research on this topic in the future.
This research was conducted in a bilateral, mixed community school where students appear to have low interest in, and motivation for, foreign language learning. Notably there are few MFL students at Key Stage 4 (GCSE), and none at Key Stage 5 (A Level). Located in a coastal town in South West England, the school’s catchment area ranked as the most deprived local authority in 2019 (Smith, 2019) and 32.88% of the entire school population are Pupil Premium. This study focuses on motivation since low student motivation is arguably one of the greatest challenges currently facing MFL teaching. Its consequences are low student numbers at GCSE, and subsequently at A Level and university, leading to a deficit in able linguists in Britain (Coleman et al., 2007; Parrish and Lanvers, 2018). Furthermore, due to the implications of language learning on social mobility, addressing low student motivation in MFL in disadvantaged areas can be considered essential in addressing social inequality and ‘challenging the gap’.
I started using Duolingo’s mobile application in October 2019, and quickly realised its potential for use with students of MFL. I began learning German ab-initio Duolingo and found the sequencing of content, repetition of vocabulary, and its gamification made for a language-learning platform that was both effective and fun to use. I became a Duolingo Certified Educator and started experimenting with the Duolingo for Schools website, which led me to consider using the app in some of my classes as part of a research project, to assess its potential for improving student motivation and decide whether it would have a place in my own, and the department’s, future practice. Although I personally find the app useful for my own language learning, I had no desired and so carried out the research as reflexively and objectively as possible.
The Context of Student Motivation in UK MFL Teaching
One of the greatest challenges facing MFL teaching is low student motivation. Coleman et al. (2007) state, “motivation is one of the most significant predictors of success in foreign language learning” (p.245). According to Parrish and Lanvers (2018), “studies on MFL motivation in the UK tend to show that students are generally poorly motivated” (Coleman, Galaczi and Astruc 2007; Lanvers 2017a; Williams, Burden and Lanvers 2002, cited in Parrish and Lanvers, 2018: 281). Furthermore, students’ enjoyment of language lessons can determine their motivation for the subject, and students often “enjoy the [MFL] lessons less than in other subjects” (Graham, Macfadyen and Richards 2012, cited in Parrish and Lanvers, 2018: 281). Studies have also found that English MFL students are less motivated than students of MFL in other countries (Bartram 2006, cited in Parrish and Lanvers, 2018: 281).
Recent developments have undoubtedly given MFL renewed importance, for example the British Academy’s (2019) declaration that monolingualism is “the illiteracy of the 21st Century” (Roberts et al, cited in The British Academy, 2019: 3) and the inclusion of foreign languages as a requirement to pass the English Baccalaureate (EBacc). However, numbers remain lower than they were before the removal of MFL as a compulsory subject at GCSE (Dobson, 2018). The British Council’s study (Tinsley & Doležal, 2018; Tinsley, 2019) shows there is much to be done to improve the status of languages and their take-up beyond Key Stage 3 in the United Kingdom. Given these findings, it is not surprising that Parrish and Lanvers (2018) discovered external factors such as gaining the EBacc were not strong motivational factors for students’ choice to study a language, whereas internal factors such as their enjoyment and perceptions of the subject’s usefulness guided their decisions. For this reason, changes need to happen at classroom level, not just at government level, to address the issue of low student motivation.
A key finding on the topic of student motivation in MFL is that “offering a free choice to students or making languages compulsory are […] likely to yield better student motivation than, for example, selecting students based on achievement, ability or other factors” (Parrish and Lanvers’, 2018: 293). According to Parrish and Lanvers’ (2018), “students who were given free choice, or no choice at all – rather than students with higher attainment – demonstrated higher levels of intrinsic motivation towards language learning, and overall higher autonomous regulation” (p. 292). These findings suggest that the best way to address low student motivation in MFL is to return to the previous model of ‘Languages for All’, or to implement a completely free choice for all. That said, this research project is concerned only with changes that can be made by individual MFL teachers and departments to address low student motivation, and therefore will not focus on factors which require change on a governmental level.
The Theoretical Underpinnings of Student Motivation
There is an incredibly rich and varied body of research literature on student motivation. Early theories, constructed through means of psychological research, studied factors that influenced and defined behaviour in situations where individuals were trying to achieve an outcome, known as expectancy-value models (see Atkinson, 1958; Lewin, et al., 1944). Whilst early research focused on arbitrary tasks in laboratory settings (Hulleman et al., 2016), later studies were conducted in a more contextualised environment. For example, Weiner (1980) began studying what he called achievement motivation, referring to the extent to which students engage in the tasks and activities required in an environment where achievement is highly valued. Other research carried out in the early 1980s sought to reconceptualise expectancy-value motivation, focusing on students’ achievement choices within educational contexts (Eccles et al., 1983). Now, research-led teaching is increasingly popular, and subsequently many studies of student motivation are conducted by educators themselves. There are innumerable theories around motivation in an educational context, however this study’s primary focus will be on the key themes and findings that informed and shaped this study, all of which come from the field of expectancy-value theory.
The first key concept is conceptualised by Bandura (1997) as self-efficacy, meaning one’s belief in one’s skills or ability to carry out certain actions. According to Bandura (ibid.), self-efficacy beliefs are determined by previous performance, vicarious learning (observing someone else successfully complete a task), verbal encouragement from others and physiological reactions to a situation or task. In context, this could mean a students’ perception of self-efficacy in French is defined by their previous performance, seeing the teacher or peers successfully speak or write in French, being encouraged by the teacher with praise and recognition, and feeling a positive ‘buzz’ after understanding something or completing a task. In this scenario, the student would likely be a motivated learner of French. However, should they have a negative perception of their previous performance (e.g. they never do well on vocab tests), watch others struggle around them, receive little encouragement or negative feedback, and feel anxiety rather than excitement about being asked to produce French, they are likely to have low motivation. As Hulleman et al. (2016) explain, positive physiological reactions like excitement can accompany higher levels of self-efficacy, whereas anxious reactions can accompany lower levels of self-efficacy, and subsequently lower levels of motivation.
The second key concept is expectancies, defined as an individual’s perception of their ability to successfully complete a task Eccles et al. (1983). This may appear to be a very similar concept to self-efficacy, however Eccles & Wigfield (2002) suggest two forms of expectancies: ability beliefs and expectancies for success. The former refers to one’s perceived ability, in the moment, to successfully complete a task, whilst the latter refers to their longer-term perception of their ability to successfully complete such tasks in the future. In the minds of students of French, for example, it can be argued that their expectancy to be able to complete a reading task at 2:15pm on a Tuesday afternoon, and their expectancy that they will be able to respond simultaneously to questions in their French oral exam in three years’ time, may both influence their overall expectancies in French, and thus their levels of motivation. Indeed, as Hulleman et al. (2016) explain, Bandura’s (1997) concept of self-efficacy relates to specific tasks which seek a specific result, whilst expectancies relate to broader domains. Both concepts have informed this research project, which assesses Duolingo’s potential to improve students’ self-efficacy and expectancies alike, thus improving motivation.
A third, expectancy-related theory, is self-concept. According to Hulleman et al. (2016), self-concept can be understood as a person’s perception of themselves (Shavelson et al., 1976, cited in Hulleman et al. 2016). In the academic domain, a strong relationship has been found between positive academic self-concept and higher grades and test scores (Marsh, Byrne, & Yeung, 1999, cited in Hulleman et al. 2016). Furthermore, according to Marsh and Martin (2011), this relationship is self-perpetuating; positive self-concepts can lead to greater academic success, which in turn feeds a more positive academic self-concept. Therefore, a platform such as Duolingo, which allows learners to feel successful at language learning, arguably has the potential to boost learners’ perceptions of their self-efficacy, increase their expectancies, and improve their academic self-concept, thus improving their levels of motivation and subsequently their educational outcomes.
According to Cognitive Evaluation Theory, an early form and sub-theory of Self-Determination Theory (Deci, 1971), externally mediated rewards can have profound effects on students’ motivation. Hagger and Chatzisarantis (2015) posit that the externally mediated rewards can ‘shift’ the “‘origin’ of the behaviour […] from ‘within’ the person (e.g., as if emanating from the self) to ‘outside’ the person (i.e. caused by external events).” (p. 4). Conceptualising Duolingo as a form of – or vessel for – ‘externally mediated reward’ in language learning, we can consider the potential benefits of using the platform to boost student motivation. Duolingo arguably has potential to create extrinsic motivation in students where neither intrinsic nor extrinsic motivation is particularly high. It does arguably also have potential to boost their perceptions on their own abilities in MFL, subsequently creating greater intrinsic motivation.
Research literature on how to best measure student motivation is abundant with many measurement scales based on different conceptual proposals. Similarly, this research project has based its enquiry on a variety of scales and theories. Originating from the influential work of Deci and Ryan (1985), the field of Self-Determination Theory (SDT) “considers that motivation can be expressed through a continuum of increasing self-determination with three fundamental positions reflecting the degree of autonomy on which behaviors are based” (Stover et al., 2012). These three factors are amotivation, extrinsic motivation and intrinsic motivation. Alongside the other aforementioned key concepts, this study has also aimed to consider these specific factors, particularly extrinsic and intrinsic motivation. This will be discussed further in the Methods section.
Research on Duolingo
Whilst no research has specifically addressed Duolingo in relation to student motivation in a classroom context (a gap which this research project aims to explore), there is nonetheless a significant amount of academic literature about Duolingo which has informed this study. Extensive research has found Duolingo to be “a promising supplementary tool for language learning” (Savvani, 2019: 139). Research conducted by Crowther, Kim and Loewen (2017) aimed to evaluate the effectiveness of Duolingo as a language-learning tool, specifically looking at “how effectively instructed second language acquisition (ISLA) research and theory has been implemented by Duolingo programmers” (p. 20). Their findings suggest “the benefit of Duolingo is more likely as a learning support app than as the sole tool for autonomous learning” (ibid.). Indeed, in this study, Duolingo is used only as a ‘learning support app’, therefore Kim and Loewen’s (2017) research offers promising findings.
The owners of Duolingo have funded research into the platform’s effectiveness, discovering that for native English speakers with no previous knowledge of Spanish, an average of 34 hours of Duolingo usage would give similar results to an 11-week university semester of study (Vesselinov and Grego, 2012). Furthermore, an analysis of the effect of Duolingo’s ‘winning streak’ (defined by a user’s consecutive days of use) on users’ motivation and engagement with the platform found that “the winning streak can boost up [sic.] learners’ motivation and attention to complete their goals” (Hiroyuki Iida, 2017: 23). It is important to note that this analysis looked at users’ motivation and goals using the platform, not in relation to school-based study. As Nushi and Eqbali (2017) explain, Duolingo utilises five different types of exercise: translations, matching exercises, pairing exercises, listening exercises and speaking exercises. Each ‘lesson’ includes a mixture of 10 to 15 activities with these exercises. It is clear from the literature that Duolingo is a well-designed, varied language-learning tool that shows promise as an aid to school-based learning.
Due to Duolingo’s design, users are arguably made to feel successful in their experience of language-learning from their very first time using the platform. It makes the user feel good about their performance at the time of use through real-time praise (e.g. “you’re doing great!”), and past performance through weekly reports about how many words one has learnt and how many days one has been active. Duolingo also seems very effective at eliciting positive physiological – and arguably psychological – reactions from the learner, thus arguably having the potential to boost students’ perceptions of self-efficacy, and subsequently their motivation. There is no doubt that Duolingo is extremely effective at motivating its users to continue using Duolingo, and thus to learn a language. With good reason it has over 300 million active users who collectively complete over 7 billion language exercises every month (Smith, 2020). However, what remains to be seen, is whether that motivation can be harnessed in the classroom and fed into MFL activities not using Duolingo, which is precisely what this research project aims to determine.
Use of Games and Gamification for Motivation in an Educational Context
In a research article on the potential for games to motivate students to learn grammar, particularly focused on the use of Kahoot in the classroom, Zarzycka-Piskorz (2016) argues “there is strong evidence that shows the relationship between game playing and increased motivation” (p.17). The research also highlights that intrinsic motivation is the main factor in engagement in gamified (made into a game) learning. In conclusion of the research findings, Zarzycka-Piskorz (2016) states:
“In the online game context intrinsic motivation is enhanced by the perspective of winning and/or getting a reward. The win as a drive to play a language game cannot be underestimated. It is the factor which allows a learning class environment to be conditioned and shaped according to the needs of the students, the learning process, or the requirements of a course. […] Implementing language games into the learning process will bring variety, break monotony, enliven classes, and motivate students to work. Rewards, points, levels are forms of extrinsic motivators, but the whole gaming experience touches significantly the intrinsic motivation aspects” (p.33-24).
These findings offer a clear reason to explore the possibilities of implementing gamified learning into foreign-language courses, particularly as a method of boosting student motivation. However, Zarzycka-Piskorz (2016) also raises many questions for researchers to consider, amongst them ‘At what point, after numerous games, would the students become bored and disinterested?’ and ‘How effective can this method of gaining knowledge and improving skills be in the long run?’. These are questions which cannot be explored by this research project, given its short-term nature, but should be considered by the reader as they engage with this research.
The research question this study aims to answer is, Can Duolingo be used to increase student motivation in MFL?
A mixed-methods approach was used so the research would be as inductive as possible. The data-collection methods used were quantitative surveys with 18 students, qualitative (semi-structured) interviews with 8 students, analysis of online data which measured 70 students’ activity using Duolingo, and teacher observation. Qualitative, semi-structured interviews were used to allow students to respond in their own words without being limited to the questionnaire statements. Despite the disadvantage of a small sample, this qualitative aproach allows participants to respond to questions with greater detail, giving them freedom to answer in their own words and with authority on issues of personal significance (Polkinghorne, 2005).
Sampling Strategy: Phase 1
A mixture of convenience sampling and judgement (a.k.a. ‘purposeful’) sampling was used. In the first instance, for ease of access and to save time, the target population was reduced only to the eight classes with whom I was timetabled and thus had contact with either as their main teacher or as an observer. From these eight classes, the Head of MFL and my Principal Subject Tutor (PST), also an MFL teacher, considered which class would yield the most interesting results. A Year 8, ‘bottom-set’, mixed ability class (8F) were considered most suitable due to their reportedly low levels of interest in and motivation for French.
Sampling Strategy: Phase 2
Given the positive initial results of implementing Duolingo with 8F, I was given permission to use the platform with other classes. To assess the potential of using Duolingo with a potentially more motivated and higher-ability class, the second class chosen was 8S, a high-ability, Year 8 Spanish class, with whom I had been teaching short episodes and some whole lessons. The third and final class chosen was a Year 10 French class (10F) of 14 students with predicted GCSE grades ranging from 3-9, reflecting their very mixed levels of ability and motivation. I was 10F’s main teacher which facilitated the implementation considerably.
The ethical approval for this study was granted by the University of Exeter Graduate School of Education. During the research design process, steps were also taken to ensure the research would meet the British Sociological Association’s (2002) Statement of Ethical Practice, ensuring the personal safety of both the participants and myself, and taking all possible precautions to guarantee the confidentiality of the research participants and anonymity of the data. It should be noted that all participant names and class codes used are pseudonyms.
Survey questions aimed to discern participants’ levels of motivation for learning French and to gather information about their perceptions of the subject. The questionnaire design was based on three surveys commonly used to measure motivation: the Academic Motivation Scale (AMS) (Stover et al., 2012), the Motivated Strategies of Learning Questionnaire (MSLQ) (Pintrich et al, 1991) and the Intrinsic Motivation Inventory (IMI) survey (Ryan, 1982). Student answers consist of four categories: strongly agree, slightly agree, slightly disagree and strongly disagree. This was drawn from Stover et al.’s (2012) adaptation of the AMS as it was simpler than the seven-category, Likert-type scale employed by other surveys. A neutral or ‘indifferent’ option was not included to encourage respondents to think and reflect more on the statements, and to remove the option to ‘opt-out’. The survey was completed twice: before implementation of Duolingo, and four weeks after.
Following Procter’s (2008) guidance, many statements were included that are “essentially the same item twice, but with reversed meaning” (p.218), in order to test the reliability of the responses. Once responses were removed that showed significant inconsistencies (suggesting the respondent had either not understood or not thought about the statements), along with respondents who were absent for either of the surveys, 18 pairs of responses were left. Whilst a full-scale study with less limitations on time and resources could have analysed the data using factor analysis and SPSS, and the statistical significance of the findings could be calculated, for this small-scale study the findings are presented in a series of simple charts.
Qualitative interviews were conducted during lesson time in an adjacent classroom so the research would fall into ‘normal lesson time’, posing less burden for the students (i.e. not requiring them to sacrifice free time) and avoiding any ethical complications (such as requiring parental consent to keep them after school). Notes were taken live (non-verbatim) to avoid audio recording and subsequent transcription, avoiding the need for secure electronic storage of audio files and saving time. All students were asked the same questions and given the opportunity to answer freely. Prompts and probes were used to encourage students to elaborate or expand on their answers, and closed questions were avoided. At each interview, students were given an opportunity to ask questions and their consent was given verbally. Interviews were conducted only with two classes, 8F (from Phase 1) and 10F (from Phase 2).
Activity Data from Duolingo and Teacher Observation
As Hine (2008) argues, “the internet offer rich possibilities for collecting data” (p.305). In the case of Duolingo, simply setting up a Duolingo for Schools account and creating a virtual classroom allows one to add students to the ‘classroom’ and track their activity. With a few clicks of the mouse, one can see who has completed ‘assignments’ the teacher has set, consult precisely when they have used the platform and see what content they were learning. Detailed summaries of students’ activity can show the teacher how many days they have been active over the past week, month, or specified date range, how many ‘lessons’ they have done and how many points they have gained. Further to this online data, Duolingo also sends the teacher weekly reports for each ‘classroom’ by email. This data, combined with teacher observation, can be considered supplementary data to add context and meaning to the primary data-collection.
Great care was taken to limit the researcher’s impact on how participants answered questions or responded to the survey statements; such as by being mindful of misdirected probing and prompting, and by phrasing questions in an open, non-leading manner (Fielding and Thomas, 2008). To address the potential impact of the researcher’s status as a trainee teacher (a position of authority), it was made clear to students that the value of their answers was in their honesty, that there were no right or wrong answers and there would be no consequences for answering in any given way. The vast number of ‘strongly agree’ responses to statements like ‘I find French lessons boring’ suggests this was not an issue.
Analysis of the questionnaire data produced some interesting findings, although the validity of some findings should be carefully considered. Data suggested a significant drop in extrinsic motivation and an increase in the perceived value of learning a foreign language, after the implementation of Duolingo. This may be deceiving since, between the two questionnaires, students received academic reports, had parents’ evening and attended assemblies that discussed their upcoming ‘options evening’, where they would choose which subjects to take at GCSE. It is not surprising that many students gave more importance to showing their ability to family and receiving a good report from their teacher, just before reports were due to be released and shortly before parents’ evening.
The rise in the value or importance they give to speaking a foreign language could also have been influenced by an assembly they had where the English Baccalaureate was explained, including the importance of choosing a language to gain this. Overall, students had more negative opinions of French in the second questionnaire. This could similarly have been affected by their decisions regarding their GCSE options, as well as how they felt about the content of their French reports.
In terms of students’ perceptions about self-efficacy and expectancies, questionnaire data showed only minor changes and some polarisation (responses moving into the ‘strongly (dis)agree’ zone). Without statistical analysis, this is inconclusive. The data from the questions about students’ learning preferences suggested the responses to the questionnaires were valid, since they were comparatively very similar, as one would expect from a control variable. Data from only one question shows a significant change, but this is only in the split between ‘slightly agree’ and ‘strongly agree’, and so can be considered a minor anomaly.
Further questions pertaining to self-efficacy and student’s perceived confidence in French showed clear improvements, most significantly in the decrease in their anxiety when asked to respond to a question in class and their confidence volunteering answers. This could suggest using Duolingo as part of their homework (and independently) aided their confidence in class. However, those that received positive feedback in their report and at parents’ evening could also have received a ‘confidence boost’ from this.
Phase 1 Interviews: Year 8 French
Data collected from interviews suggests Duolingo had a positive effect on student motivation. Whilst Amy felt that using the platform helped her understanding in class, it did not change her perception of French. Ben found it “fun and immersive” and was excited that “you get points and lingots which can be used to dress up the owl [Duolingo mascot] and get power-ups”. He also reported he enjoyed French more as a result of using Duolingo. It would seem that the gamification element of the platform was a key factor in his increased motivation. This is echoed by Oscar, who said, “it seems more fun that you can play a game”. Oscar also liked that “you can get streaks and achievements and XP to show you’re the best”, suggesting the element of competition is also an important feature. Whilst there is a lot of encouraging data from Oscar to consider Duolingo a positive influence for his overall motivation for language learning, it does bring to mind the danger that “the individual does not, therefore, engage in the behaviour for its own sake and the inherent satisfaction and interest derived from the behaviour itself but instead engages in it to obtain the reward or externally-referenced contingency” (Hagger and Chatzisarantis, 2015: 4.).
It is feasible that Oscar was motivated by his desire to be “the best”, in which case it should questioned whether this motivation can translate into motivation in French lessons. That said, one can also consider whether this is necessary. Arguably, using Duolingo based on a desire to gain points has the secondary effect that one acquires vocabulary and revises grammar, thus aiding them in lessons which, in turn, may affect their intrinsic motivation by improving their perceptions of self-efficacy (Bandura, 1997). Indeed, there was evidence to support this from Oscar’s interview. It is important to note that Oscar is an elective mute, which presents particular difficulties in French lessons where speaking is an important part of class activities. In the interview he said, “if I get something wrong, everyone will look at me and it makes my heart beat really fast and I get scared”. As a result of using Duolingo, he reported feeling more confident in class, which could arguably be a prerequisite of being more motivated in class.
Phase 2 Interviews: Year 10
Harry found the app “very helpful” and particularly liked the “button to go over the words [which] gets them into your head” and the fact that “you have to put it into the right tense”. Susie, one of several students in 10F who already used Duolingo, found the app helped her to translate sentences more easily and said with time she had become “faster and faster at translations”, which would suggest it was helping with vocabulary recall. Lianne said, “it’s very useful for going over things and I like how it asks you the same questions in different ways”. She also said the app was “good for consolidation” of what she learns in class. Similarly, Miley found “it teaches you new words and gives you examples”. She also said, “you learn tenses in class and then they come up in Duolingo so you get to practice them and make links”. Susie, Miley and Lianne all said the competition makes them do it every day, to keep their ‘day streaks’, to move up in the ‘leagues’ and to compete amongst their friendship group. This echoes the finding from Oscar’s interview, that the competitionelement was an important motivator for their efforts using Duolingo.
The only negative feedback given in interviews, came from Annie, a German exchange student in 10F. Annie had a very advanced level of French, comparable with that of a high-ability Year 12 student in the UK. She found the app was “not so useful for in class because I already know it” and was not sure about the fact that “it gives you random phrases”. Since Annie is an example of a highly motivated and high-attaining student, this may suggest Duolingo has much less to offer students like her. Nonetheless, since this study aims to improvestudent motivation, particularly where it is low, this finding is not problematic provided Duolingo has no adverse effect on the pre-existing motivation of students like Annie.
Activity data from Duolingo
With 8S, the only method of data collection used was consultation of students’ activity via the Duolingo for Schools website. It is clear from this information that due to their teacher’s strict ‘policing’ of Duolingo homework, all students in the class completed this. There were also many students who completed far more than was required. Over a 16-day period, they were required to complete 4 lessons, equating to approximately 40XP (the points one gains for each completed lesson). Interestingly, one student completed a total of 75 lessons and gained 1065XP. Indeed, of the 31 students in the class, only one student completed the minimum amount required of them.
Activity Data from 8F and 10F show a similar effect, though with less compliance from the students with homework which was an existing and ongoing issue. Oscar in 8F, for example, completed his homework which was to gain 20XP, equating to 10 minutes’ work. However, over the period of this study, he gained a total of 2254XP, equating to 14 hours and 47 minutes of French study at home or in the library. In 10F, 13 out of 14 students did more than was required to complete the homework. During this study, the group were assigned three pieces of homework (I was their main teacher which made it easier to control this), requiring them to gain 60XP. However, the average amount of XP gained by 10F students over this period was 662, over ten times the required amount. The highest amount was 2292XP, equating to over 18 hours of independent French study. These findings would seem to echo findings from the other data collected, that Duolingo can be incredibly effective at motivating students to continue using Duolingo. Arguably, this can positively affect motivation in class by improving students’ perceptions of their self-efficacy (Bandura, 1997), their expectancies(Eccles et al., 1983) and their academic self-concept by helping them to achieve higher grades and test scores (Marsh, Byrne, & Yeung, 1999, cited in Hulleman et al. 2016), creating a more positive self-fulfilling prophecy (Marsh and Martin, 2011).
After doing a seemingly super-human amount of Duolingo at home in the first week, Oscar began only periodically using it. When I asked him about this, he reported he knew he was so far ahead of the rest of his class that it would take them a long time to catch up, thereby his initial motivation had a de-motivating effect in the long-run, at least in terms of his motivation to learn French using the platform. To address this, I began using a different approach with the three classes who were using Duolingo. Whereas before I showed classes an ‘all-time leader board’ (a slide I included in lesson presentations), I began showing a ‘weekly leader board’ which named students who had gained the most XP (points) over the last 7 days. From observation and consultation of the data available on my Duolingo for Schools account, this appeared to have a considerable effect on maintaining the motivation of those who initially did the most, giving them an incentive to continue each week.
In terms of observable differences in motivation, Harry (10F) was the clearest example that Duolingo had had a positive effect. He was late to start using the app due to absence, but once he began using it there was a noticeable difference in his confidence giving answers in lessons, and innumerable times he would give a correct answer then proudly say, “I learnt that on Duolingo, sir”.
Teaching episodes and whole lessons with 8S, I observed that sharing the Duolingo weekly leader board was a great motivator for students’ work on Duolingo, that they were proud to appear on it and also that it seemed to be the source of kudos amongst the class.
Class teacher impact on study
A significant shortcoming of this study was that the teacher of the primary class being studied, 8F, did not consistently support the implementation of Duolingo. At the beginning of the study, the teacher was happy to see the immediate success of Duolingo as some students were using it very frequently, setting the class a Duolingo ‘assignment’ as part of their homework and celebrating their effort in class. However, undoubtedly with other priorities such as writing reports, marking work, and preparing for parents’ evening, this enthusiasm was short lived. The result was that the very promising start to the Duolingo trial quickly faded and very few students continued using the platform. It was only set once for homework, and when I informed the class teacher of those who had not completed it, it was not followed up with much rigour. Unfortunately, students who did not complete their homework did not have the necessary experience using Duolingo to draw conclusions about its effect on their motivation.
Issues with first contact
Another issue with the first phase of this study (class 8F) was their first contact with Duolingo. Part of a lesson was dedicated to setting students up with Duolingo accounts in a computer room. The aim was for them to access the accounts and to spend the rest of the lesson using Duolingo. Due to the short amount of time given for this, many students merely accessed their accounts and completed very few ‘lessons’ (or none at all), limiting their exposure to the rewards and incentives which arguably make Duolingo appealing and effective. Indeed, it later became clear that this was an important factor, as those students who did the most during the initial set-up lesson continued to use the platform, completed their Duolingo homework, and gave positive feedback about Duolingo in the second questionnaire. For these reasons, I amended the way I introduced it to other classes, ensuring they were given enough time to experience using the platform for the second phase of the study.
Issues with timetabling
Another important limitation of the first phase of this study was the fortnightly timetabling of the 8F. The group received French lessons on three consecutive days in week A, and only once – on a Monday – in Week B. With Half Term taking place in the middle of the study, 8F had no French classes for fifteen days, accounting for over a third of the period over which this study was conducted. Combined with the fact that they were only set Duolingo homework once, with limited follow-up, and that they received little encouragement from their class teacher, these issues represent a considerable limitation on the extent to which this phase of the study can provide valid findings. Furthermore, as discussed in the Analysis section, the effects of reports, parents’ evening and options evening should also be considered limitations of this study.
This study has made clear the potentially motivation-boosting impact competition and gamification can have on language-learning, echoing Zarzycka-Piskorz’s (2016) finding that “gamification can add motivation to learning activities and as such should not be underestimated” (p. 33). It is clear that Duolingo can be incredibly effective at motivating students to continue using Duolingo, though this raises the question of whether this motivation feeds into motivation in class. Whilst the direct link between Duolingo and increased motivation in class remains uncertain, it can be argued that consistently using Duolingo (for example as part of students’ homework) can positively affect their motivation in class by improving perceptions of their self-efficacy(Bandura, 1997), their expectancies (Eccles et al., 1983) and their academic self-concept by helping them to achieve higher grades and test scores (Marsh, Byrne, & Yeung, 1999, cited in Hulleman et al. 2016). Whilst use of Duolingo implies learning and learning creates better outcomes in lessons, it would appear that this upwards cycle can create a more positive self-fulfilling prophecy (Marsh and Martin, 2011) and ultimately cultivate greater intrinsic motivation.
The implications this has for educators, including myself, is that Duolingo should be considered as an effective, fun and free tool available to schools to increase student motivation in MFL. The secondary findings of this study demonstrate the importance of how Duolingo is presented to students, how it is used for homework, and how the ‘competitive energy’ it creates can be effectively harnessed to maximise results. These secondary findings should be considered by educators looking to implement Duolingo, or similar platforms, themselves.
I would like to thank my university tutor, Chris Wakely, for his support and for allowing me to start collecting data sooner than programmed in order to improve the quality of research. I would also like to thank the MFL department at my first PGCE placement school, who allowed me to trial Duolingo with several classes. Finally, I thank my students for their participation and their honesty.
For a copy of the complete research project with all appendices and graphs, please contact me by email.
- Atkinson, J.W. (1958). Towards experimental analysis of human motivation in terms of motives, expectancies, and incentives. In J.W. Atkinson (Ed.), Motives in fantasy, action, and society (pp. 288-305). Princeton, NJ: D. Van Nostrand Company, Inc.
- Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman.
- British Sociological Association. (2002) Statement of Ethical Practice for the British Sociological Association [online]. Available from: https://www.britsoc.co.uk/equality-diversity/statement-of-ethical-practice/#_anon [Accessed 08/02/2020].
- Busse, V. and Williams, M. (2010) Why German? Motivation of students studying German at English universities, Language Learning Journal, 38(1): 67-85.
- Coleman, J. A. (2009) Why the British do not learn languages: myths and motivation in the United Kingdom, Language Learning Journal, 37(1): 111-127.
- Coleman, J. A., Galaczi, A. and Astruc, L. (2007) Motivation of UK school pupils towards foreign languages: a large-scale survey at Key Stage 3, Language Learning Journal, 35(2): 245–281.
- Crowther, D., Kim, K. M. and Loewen, S. (2017) The Implementation of ISLA in MALL Technology: An Investigation into the Potential Effectiveness of Duolingo, MSU Working Papers in SLS, Vol. 8.
- Deci E. L. and Ryan R. M. (1985) Intrinsic Motivation and Self-determination in Human Behavior. New York, NY: Plenum Press.
- Deci, E.L. (1971). Effects of externally mediated rewards on intrinsic motivation, Journal of Personality and Social Psychology, 18: 105-115.
- Dobson, A. (2018) Towards ‘MFL for all’ in England: a historical perspective, The Language Learning Journal, 46:1, 71-85.
- Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53, 109–132.
- Eccles, J. S., Adler, T. F., Futterman, R., Goff, S. B., Kaczala, C. M., Meece, J. L., & Midgley, C. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motivation (pp. 74-146). San Francisco, CA: W.H. Freeman.
- Hine, C. (2008). The Internet and Research Methods, in Gilbert, N. (Ed) Researching Social Life (2008). London: Sage.
- Hiroyuki Iida, D. H. (2017) An Analysis of Winning Streak’s Effects in Language Course Of “Duolingo“, Jurnal Teknologi Maklumat dan Multimedia Asia-Pasifik, Vol. 6: pp. 23-29.
- Hulleman, C. S., Barron, K. E, Kosovich, J. J. and Lazowski, R. A. (2016) Student Motivation: Current Theories, Constructs, and Interventions within an Expectancy-Value Framework, in Lipnevich, A., Preckel, F. and Roberts, R. (Eds.) Psychosocial Skills and School Systems in the 21st Century: Theory, Research, and Practice. New York: Springer.
- Lewin, K., Dembo, T., Festinger, L., & Sears, P.S., (1944). Level of aspiration. In J. M. Hunt (Ed.), Personality and behavior disorders (pp. 333-378). New York: The Ronald Press.
- Marsh, H. W., & Martin, A. J. (2011). Academic self‐concept and academic achievement: Relations and causal ordering. British Journal of Educational Psychology, 81(1), 59-77.
- Nushi, M. and Eqbali, M. H. (2017) Duolingo: a mobile application to assist second language learning, Teaching English with Technology, 17(1), 89-98.
- Parrish, A. and Lanvers, U. (2018) Student motivation, school policy choices and modern language study in England, The Language Learning Journal, 47(3), 281-298.
- Pazzaglia, A. M., Stafford, E. T. and Rodriguez, S. M. (2016) Survey Methods for Educators: Selecting Samples Administering Surveys (part 2 of 3), Applied Research Methods [Online] Available at: https://files.eric.ed.gov/fulltext/ED567752.pdf [Accessed 25/03/2020].
- Pintrich, P.R., Smith D.A.R, Garcia T. and McKeachie, W.J. (1991) A Manual for the use of the motivated strategies for learning questionnaire (MSLQ), Washington, DC: University of Michigan.
- Procter, M. (2008) Measuring Attitudes, in Gilbert, N. (Ed) Researching Social Life (2008). London: Sage.
- Ryan, R. M. (1982). Control and information in the intrapersonal sphere: An extension of cognitive evaluation theory. Journal of Personality and Social Psychology, 43, 450-461.
- Savvani, S. (2019) State-of-the-Art Duolingo Features and Applications, in Auer and Tsiatsos (Eds) The Challenges of the Digital Transformation in Education: Proceedings of the 21st International Conference on Interactive Collaborative Learning (ICL2018) – Volume 2. Cham: Springer Nature Switzerland AG.
- Smith, C. (2019) Torbay has a problem: The terrible statistics behind the English Riviera, Devon Live [Online] Available at: https://www.devonlive.com/news/devon-news/torbay-problem-terrible-statistics-behind-3334873 [Accessed 26/03/2019].
- Smith, C. (2020) 17 Amazing Duolingo Facts and Statistics [Online] Available at: https://expandedramblings.com/index.php/duolingo-facts-statistics/ [Accessed 06/04/2020].
- Stover, J. B., de la Iglesia, G., Rial Boubeta, A. and Fernández Liporace, M. (2012) Academic Motivation Scale: adaptation and psychometric analyses for high school and college students, Psychology Research and Behavior Management, 5(1): 71–83.
- The British Academy. (2019) Languages in the UK: A call for action. London: The British Academy.
- Tinsley, T. (2019) Language Trends 2019: Language Teaching in Primary and secondary Schools in England. London: British Council.
- Tinsley, T. and Doležal, N. (2018) Language Trends 2018: Language Teaching in Primary and secondary Schools in England. London: British Council.
- Vesselinov, R. and Grego, J. (2012) Duolingo Effectiveness Study: Final Report [Online] Available at: http://static.duolingo.com/s3/DuolingoReport_Final.pdf [Accessed 06/04/2020].
- Weiner, B. (1980). The role of affect in rational (attributional) approaches to human motivation. Educational Researcher, 9(7), 4-11.
- Zarzycka-Piskorz, E. (2016) Kahoot it or not? Can games be motivating in learning grammar?, Teaching English with Technology, 16(3), 17-36.