Correct me if I’m wrong…

Recently I felt a strange sense of camaraderie with some unknown to me Vietnamese teachers. I was reading a case study (Ha & Murray, 2021) investigating how training can affect teachers’ beliefs and their practices regarding oral corrective feedback (OCF). It was comforting to realise that I was not alone: the questions I have had about OCF for years are also being investigated by others across contexts, cultures, and time. To date, research has, admittedly, provided some answers but there are still quite a few questions that have not been resolved. As OCF is, in most cases, an integral part of classroom practice, I thought I would share some of these questions, starting with the most obvious one: 

What is OCF?  

Lightbown and Spada, (2013) define oral corrective feedback as an “indication to a learner that his or her use of the target language is incorrect” (p.216). OCF can target a broad range of errors in learners’ production, can be immediate or delayed, and it can stem from different sources: the students themselves, their peers or their teacher (Loewen, 2013). It is also known as error correction or negative feedback. The latter can be contrasted to positive feedback, which is “an affirmation of the content or correctness of a learner utterance” (Nassaji, 2015, p. 11) intended to provide the learner with affective support.

Can we do without OCF in class? 

Well, OCF naturally occurs in class to varying degrees, whether intended or not, especially when meaning is negotiated (Loewen, 2013). Whether OCF should be provided or not depends on our beliefs (actual or stated) regarding language learning. From the nativist perspective, for which naturalistic exposure or modified input is the sole condition for learning, OCF is irrelevant; it hardly plays a role in the language acquisition process (Li, 2018). On the other hand, theoretical perspectives such as sociocultural theory, social constructivism and interactionist approaches claim OCF has an important role to play. For interactionists, especially, providing opportunities for learners to produce output and to receive feedback are core components (Mackey et. al., 2013). Experimental studies to date, mostly from the interactionist perspective, demonstrate beyond contention that OCF can facilitate L2 development (Li & Vuono, 2019).

From the pedagogical perspective, OCF was not considered significant in a range of different teaching methods, from humanistic approaches to the Natural Approach and to early versions of Communicative Language Teaching . In later versions of Communicative Language Teaching and in Task-Based Language Teaching, though, OCF was reinstated, as it was considered important for both accuracy and fluency (Ellis, 2017). Teacher guides (e.g., Ur, 1996; Scrivener, 2005) argue in favour of a “preventive pedagogy” (Lightbown, 1998, p. 189) that leaves little room for errors, but do advise in favour of OCF and even prescribe how and when OCF should be provided, so as not to interfere with communication and negatively impact affective factors.

Do all the errors that occur need to be addressed?

Research advocates focused OCF, i.e. addressing only some of the errors made (Ellis, 2009). Finding out about this research was a great relief for me, because growing up in a red-pen culture I felt overwhelmed with the number of errors corrected. It is not just that it was demotivating, it was also that after a while one simply does not pay attention anymore. A less-is-more approach is also what my experience as a teacher has taught me: As a novice teacher I was pressured into correcting everything explicitly to satisfy a demand for total accuracy that often equated teaching with testing. Learners, of course, due to cognitive and affective reasons, were not able to notice all the corrections in a process that was, it turns out, not conducive to learning. 

So, the subsequent question is if not all, then which errors need to be addressed. This, in theory at least, is an easy question to answer according to teacher guides (Ur, 1996; Scrivener, 2005): Teachers should prioritise those errors that can lead to communication breakdown or affect larger stretches of discourse (impeding, global errors) and those due to lack of knowledge over the ones that do not affect comprehensibility of whole utterances (local) and those that are not systematic (known as performance errors or mistakes). However, research does not fully support this convenient recipe: there is evidence in favour of correcting local errors as well as non-systematic mistakes in addition to impeding, global errors. In any case, it can be difficult for teachers, especially less experienced ones, to make distinctions between global and local errors or errors and mistakes in the short time afforded to OCF (Ellis, 2009; 2017).

When is the best time to provide OCF? 

Both immediate OCF, provided immediately after an error occurs, and delayed OCF, provided at a later point in time, have been found to be beneficial (Ellis, 2017). Immediate correction can have the benefit of incorporating the correct form into learners’ attempts to communicate and, thus, not compromising form over meaning in communicative tasks (Li, 2018). There are valid theoretical arguments in favour of delayed OCF, too. Quinn and Nakata (2017), besides affective reasons, note from the theoretical perspective of cognitive psychology that longer intervals between error and OCF lead to better long-term retention (p.37). Empirical research has not been conclusive as to when is the best timing for OCF (Quinn & Nakata, 2017). Teacher guides (e.g., Ur, 1996; Scrivener, 2005) suggest immediate correction in accuracy work and recommend delaying it in fluency work. Research, though, does not support the avoidance of OCF in fluency work, as it has been found to assist acquisition (Ellis 2017). 

Who should be the one to provide OCF? 

According to teacher guides, correction should be attempted in a very specific order; self, peer, and finally teacher (Ur, 1996; Scrivener, 2005). However, this suggested good practice regarding OCF is inconsistent with learners’ beliefs and expectations: learners generally have positive attitudes towards OCF, prefer more explicit correction of all errors if possible, and expect to be corrected by the teacher rather than their peers (Li & Vuono, 2019). My own experience of observing teachers in various contexts also indicates that most teachers don’t seem to follow the suggested practice anyway.

How should OCF be provided? 

This is a difficult question to answer since the contextual, individual, and linguistic features affecting the delivery of OCF are intertwined (Yu et al, 2018). Teachers are often expected to make an on-the-spot decision about how to correct, which will inevitably be influenced by their own beliefs. Teachers’ beliefs regarding the provision of OCF, shaped by education, training, and experience both as learners and as teachers, are often rigid and can even contrast with research evidence (Junqueira & Kim, 2013). For many years, I advocated in my practice as a teacher providing focused (i.e., selective) OCF in an implicit way, so as to have minimal impact on fluency, taking into account affective factors as well. I believed it was best not to spoon-feed corrections and that self-discovery was both cognitively and affectively the optimal choice. I consciously avoided using any explicit OCF strategy, although I was aware of empirical evidence attributing merit to explicit correction, including those corrections which make use of metalinguistic clues (Ellis, 2017). By following closely pedagogical suggestions, I was in fact being dismissive of research and effectively limiting the range of my OCF repertoire. 

I remember once I employed an elicitation strategy for a grammatical error that occurred during a short dialogue (I not like dogs in parks) in an adult elementary class. After all learners finished their contributions, I initiated an elicitation process as I had done successfully before. This time however, it ended up being a tiresome, awkward, ambiguous, and ultimately ineffective elicitation process. It would have been best to address this error with an explicit correction after the contribution was concluded. It would have been concise, closer to the occurrence, and since the omission of the auxiliary is a common error with low level learners, it could have provided others with an accurate example of the form to use in their contributions. Apparently, even in the most accommodating of contexts, even well-informed teachers’ beliefs do not always lead to efficient OCF practices. 

How can a teacher decide on a recipe for OCF that works? 

Unfortunately, as is true with many things in life, there is no such thing as a single foolproof recipe for OCF across contexts. I have found this to be the case in many of my previous learning experiences as a learner and as a teacher, as my example above illustrates. In fact, there are no definite answers provided either by research or by teacher guides (Ellis, 2017). Perhaps the search for a single recipe is misguided as “the constellation of moderating variables” (Russell & Spada, 2006, p.156) cannot always be simultaneously satisfied even in the same context.

What are teachers supposed to do? 

As Ellis (2017, p.14) notes “the best advice that can be given to teachers is to deploy a variety of strategies”. This is good advice, but it presupposes that teachers are knowledgeable, have a good grasp of classroom complexities, and are willing to challenge their beliefs. For those less experienced and/or less knowledgeable what emerges as an invaluable step towards more efficient delivery of OCF, is teacher education and training (Borg, 2011). Delivery of OCF in the L2 classroom is demanding and difficult for all teachers, even those with a lot of experience and knowledge. In my experience, it is not always easy to adapt successfully to the challenges posed by the dynamic process of OCF. Consideration of the profile of your class and a comprehensive language analysis in lesson planning help to identify potential problems and subsequently elect OCF strategies more appropriate to the context. Voice, posture, gestures, choice of words are important elements to consider especially since these can enhance saliency of the OCF provided.

The way teachers provide OCF can also serve as a model for peer correction. Learners need to be trained how to provide peer correction and, as with teachers, this is a process that requires time, effort, and opportunity. Planning, acting, reflecting, observing, sharing experiences of practice, are key towards a more reflective practice.

The journey of reflective practice does not only help manage even unplanned OCF more efficiently but certainly improves all aspects of our teaching (Borg, 2011). The necessary condition, though, for all the above is that the teacher should challenge their beliefs, be ready to try new recipes. A well-guided, experiential training course does exactly that: it provides the tools to challenge the ways we do things in class including the way we deliver OCF. At least, this is what training offered me: a bold new perspective on practice and a re-evaluation of my beliefs. 


Borg, S. (2011). The impact of in-service teacher education on language teachers’ beliefs. System39(3), 370-380.

Ellis, R. (2009). Corrective feedback and teacher development. L2 Journal1(1).

Ellis, R. (2017). Oral corrective feedback in L2 classrooms: What we know so far. In H. Nassaji & E. Kartchava (Eds.), Corrective feedback in second language teaching and learning: research, theory, applications, implications (pp. 3–18). Routledge.

Ha, X. V., & Murray, J. C. (2021). The impact of a professional development program on EFL teachers’ beliefs about corrective feedback. System96, 102405.

Junqueira, L., & Kim, Y. (2013). Exploring the relationship between training, beliefs, and teachers’ corrective feedback practices: A case study of a novice and an experienced ESL teacher. Canadian Modern Language Review69(2), 181-206.

Li, S. (2018). Corrective feedback in L2 speech production. The TESOL Encyclopedia of English Language Teaching, 1-9.

Li, S., & Vuono, A. (2019). Twenty-five years of research on oral and written corrective feedback in system. System84, 93-109.

Lightbown, P. M. (1998). The importance of timing in focus on form. In C. Doughty & J. Williams (Eds.), Focus on form in classroom second language acquisition (p. 177–196). Cambridge University Press.

Lightbown, P. M., & Spada, N. (2013). How languages are learned – Oxford handbooks for language teachers (4th ed.). Oxford University Press.

Loewen, S. (2013). The role of feedback. In S. M. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition (pp. 24-40). Routledge.

Mackey, A., Abbuhl, R., & Gass, S. M. (2013). Interactionist approach. In S. M. Gass & A. Mackey (Eds.), The Routledge handbook of second language acquisition (pp. 24-40). Routledge.

Nassaji, H. (2015). The Interactional feedback dimension in instructed second language learning: Linking theory, research, and practice. Bloomsbury Publishing.

Quinn, P. G., & Nakata, T. (2017). The timing of oral corrective feedback. In H. Nassaji & E. Kartchava (Eds.), Corrective feedback in second language teaching and learning: Research, theory, applications, implications (pp.35-47. Routledge

Russell, J., & Spada, N. (2006). The effectiveness of corrective feedback for the acquisition of L2 grammar: A meta-analysis of the research. In J. M. Norris & L. Ortega (Eds.), Synthesizing research on language learning and teaching (p. 133–164). John Benjamins Publishing.

Scrivener, J. (2005). Learning teaching: A guidebook for English language teachers. Macmillan Education.

Ur, P. (1996). A course in language teaching: Practice and theory. Cambridge University Press. 

Elisavet Kostaki - Psoma
Elisavet Kostaki – Psoma

Elisavet has been involved in various aspects of TESOL for more than two decades: as a teacher, an examiner, a mentor and a teacher trainer. For the last two years she has also worked on various CELTA courses as an approved tutor.

Can a leopard change its spots?

Stalking Aditya Singh—Flickr RF/Getty Images

The interview

It was one of the first few times I was interviewing candidates for a teacher training course. I remember being quite nervous myself trying to make complete strangers feel comfortable enough to maybe get a glimpse of their motives by asking for their reasons for taking the course and their expectations. And there was that young confident teacher, ticking all the right boxes, up until she looked straight at me and declared in brutal honesty: “I don’t think I’ll become a better teacher or a good teacher. I only care about getting a better job”. Hm. Sure. Fine. A good qualification helps in that direction I said, smiled, and moved on. 

It got me thinking, though. The part about becoming a good teacher was perplexing. It was said in a way that made it look like being a good teacher is an unattainable goal and good teachers are only short of mythical creatures. Surely, there is no simple answer to the question ‘what makes a good teacher’. I was more concerned, though, with this teacher’s attitude towards the course, which I fear is shared by many others: almost convict-like, being resigned to do the time; keeping their heads down, avoiding trouble, reading and doing what they are told, and then finally being released into the world of ELT with a professional competitive qualification. But why take for granted that no substantial change will come of it? Why not challenge existing beliefs about teaching and learning? There are of course some valid reasons for the whys that started piling up. But still, I kept missing the point. The real question, the one that really mattered from where I was sitting, was how these beliefs could be challenged on a teacher training course. 

A new set of questions 

So, a new set of questions arose with the first thing to clarify being what beliefs are in the teaching training context.

There has been a lot of work from psychological and philosophical perspectives aiming to define beliefs in general. Borg (2011), based on that work, describes beliefs in our teaching training context as ideas or opinions individuals hold to be true, often tacit, with a strong evaluative and affective component, used as a basis for action and resistant to change (pp. 370 -371). Beliefs then affect teachers’ learning and their practices; so much so that the importance attributed to them in teacher education has reached the point of describing them as possibly “the clearest measure of a teacher’s professional growth” (Kagan, 1992, p. 85).

That much is clear then. For teacher education to have an impact on what teachers do it has to impact their beliefs, too. Obviously, the tricky part is how to impact those beliefs which are resistant to change. For us teachers the part of resistance is even more difficult to challenge because our initial beliefs about second language teaching (the ideas that we come to teacher education courses with) are to a great extent based on our previous experiences as language learners, or what Borg (2003) calls the “apprenticeship of observation”. This apprenticeship is a powerful thing that has been found to withstand training and not to waver even as teaching experience accumulates (Junqueira & Kim, 2013). The cards are stuck against us right from the start, it seems. 

Research to the rescue 

The good news is that research into the matter, and there has been plenty since the 1970s (see Borg 2009 for a comprehensive account), points to the direction that not only are there changes in the behaviours of trainee teachers but also changes in their thinking and beliefs as a result of teacher education courses (Borg, 2009). Research has also had an impact on the design of teacher training courses which acknowledge the importance of teachers’ beliefs, such as for example the DELTA, which aims at developing “candidates’ ability to reflect critically on their own beliefs about teaching and learning” (Delta Syllabus specifications, 2019, p.4). 

Some suggestions

Reflecting critically on one’s beliefs is no easy task, though, especially for many teachers for whom it can be a novel experience that requires support and examples of how to start this process by asking the right questions. Teachers need to be able to distinguish beliefs from practices and theoretical knowledge; they also need to be given an explanation as to why they are encouraged to examine their beliefs. Providing a safe space, ways, and opportunities to make their beliefs explicit, is as important as it is to be able to share this process with peers. For some, questioning their beliefs about learning and teaching leads to the verification of pre-existing beliefs, for others it opens new prospects. Either way, teachers have been introduced to a roadmap of how to reflect upon their praxis (knowledge enacted into a practical and active process), an integral part of which is to challenge their beliefs, to view them as fluid rather than set in stone.


Teacher training courses are so much more than a welcome addition to a resumé. On top of everything else, a training course is a transformative time that equips us with tools to reflect on our practice and our beliefs. Learning how to challenge our beliefs is perhaps the secret to becoming a unicorn, the mythical unattainable creature, a good teacher. As Prodromou (2020) eloquently frames it “Good teachers are not experts with answers; they are practitioners who ask questions” (p. 309). So, keep on asking. The teacher from the interview certainly did.  


Borg, S. (2003). Teacher cognition in language teaching: A review of research on what language teachers think, know, believe, and do. Language Teaching36(2), 81-109.

Borg, S. (2009). Teacher cognition and language education: Research and practice. Bloomsbury Publishing.

Borg, S. (2011). The impact of in-service teacher education on language teachers’ beliefs. System39(3), 370-380.

DELTA Syllabus specifications. (2019). Cambridge English.

Junqueira, L., & Kim, Y. (2013). Exploring the relationship between training, beliefs, and teachers’ corrective feedback practices: A case study of a novice and an experienced ESL teacher. Canadian Modern Language Review69(2), 181-206.

Kagan, D.M., 1992. Implications of research on teacher belief. Educational Psychologist, 27, 65-90.

Prodromou, L. (2020). Epilogue. Good Language Teachers – What Do They Do? What Do They Know?. In C. Griffiths & Z. Tajeddin (Eds.), Lessons from Good Language Teachers (pp. 309-312). Cambridge University Press.

Elisavet Kostaki - Psoma
Elisavet Kostaki – Psoma

Elisavet has been involved in various aspects of TESOL for more than two decades: as a teacher, an examiner, a mentor and a teacher trainer. For the last two years she has also worked on various CELTA courses as an approved tutor. It is our pleasure to welcome her to ACE TEFL!

Teachers and Corpora

Do you use corpora?

If you think you don’t use corpora in your planning and teaching, think again! Because we teachers actually do use corpora all the time, perhaps unknowingly, for example whenever we google a word or an expression, looking for examples of its use, or when we go through texts online to find the one with the required linguistic elements for our class.

What is a corpus?

Corpus (plural corpora), the Latin word for body, has come to mean, in corpus linguistics, a “collection of electronic, naturally-occurring texts (written or spoken) which are selected to be representative of a particular language or language variety” (McEnery et al, 2006, p. 5). Corpora usually include millions, if not billions of words, not only as used by native speakers but also by L2 speakers. They can provide evidence of language use in different genres and registers and reveal patterns in language.

Accessing all these becomes feasible with the use of corpus software tools. For example, have a look at the three figures below, which illustrate how I used Sketch Engine to produce frequency lists for the word teacher (figures 1 and 2) and to create a concordance for the same word, based on data from the British National Corpus (BNC).

Figure 1. Visualisation of a frequency list for ‘teacher’ generated by Sketch Engine
Figure 2. Different view of the same search of a frequency list for  ᾽teacher᾽ generated by Sketch Engine.
Figure 3. List of concordance lines for the lemma ᾽teacher´ generated by Sketch Engine. 
KWIC stands for keyword in context

Frequency of linguistic items can inform curricular decisions about the order these items should be taught in, materials to be used, as well as test development. Concordances, i.e. all the instances of a linguistic item listed in their immediate context, are not to be confused with collocations, ie.the way individual lexical items are used together regularly. Collocations can be traced in concordance lines, along with other information on language use in context which can be utilised in teaching.

Why is it a good idea to use corpora?

So, corpora can provide a wealth of information about authentic language use of different registers and genres, and corpus software tools can help explore them. But why should teachers use corpora? Most importantly, why should learners use them?

Teachers can use corpora to enhance their research skills and develop their language awareness to inform their teaching. Quite often teachers need to adapt or supplement published materials which cover to a limited extent desirable linguistic features, or in which natural speech instances are not presented, or need to be presented using a different approach (Walsh, 2010). Teachers can source instances of natural language from corpora to provide tailor-made activities to cater to specific learners’ needs considering level, context, system and skill. They can modify corpora or use them as is to demonstrate a language pattern in context or ask learners to notice features of a function or compare data from native and L2 corpora for error correction. Imagination is the limit. 

The use of language corpora as a resource holds significant benefits for the learners on many levels as well. Cheng (2010) states that it “has been shown to contribute to the acquisition of both implicit and explicit knowledge” (p.320). It can support exploratory and discovery learning, which is motivating for the learners and as they engage more actively probably retain knowledge for longer. Autonomous learning is encouraged and since learners acquire or practice essential learning skills such as noticing, making inferences, and reflecting, they can transfer all these to other fields of study (Gilquin & Granger, 2010).

The cognitive, pedagogical, and motivational benefits of the use of corpora in language learning in the form of Data-Driven Learning (DDL) have been discussed repeatedly in presentations in conferences of local and national TESOL associations, and the annual IATEFL conferences over the years. ELT Blogs often feature corpora and there is even a dedicated FB group promoting the use of corpora in language teaching. Why is it then that teachers are reluctant to apply DDL and mostly view the use of corpora as a tool limited to lexicographers, academics, researchers, material developers and test designers? 

Challenges associated with using corpora

McCarthy and O’Keeffe (2010) point out that teachers’ attitudes towards DDL are not favourable as “it turns the traditional order within the classroom upon its head. The corpus becomes the centre of knowledge, the students take on the role of questioner and the teacher is challenged to hand over control and facilitate learning” (p. 7). Attitudes aside, teachers’ limited time and workload, lack of computer skills and training on how to access and use corpora are major issues. There is also the cost to access corpora since many of them are not freely available. To that, schools will need to add the cost of the investment in computers, training of teachers and students, and software tools. And with empirical data on the effectiveness of language learning still limited it is not even a guaranteed investment for them to make (Gilquin & Granger, 2010)! Apart from teachers’ attitudes and logistics, learners’ attitudes towards DDL are an additional reason why interest in DDL is still limited. There are challenges for the learners, especially when little training is given, or they do not have appropriate proficiency level and technological knowledge (Soruç & Tekin, 2017). 

To address the challenges learners experience, time and effort need to be invvested in training in using corpus tools, reading and interpreting corpus results, and interpreting quantitative results may be needed (Sripicharn, 2010). The step preceding this, though, should be the training of the teachers. Corpus linguistics is more common nowadays in initial language teacher education but besides basic knowledge, development of skills is required in a way that would promote active engagement with corpora. 


Should we then abandon the use of corpora in non-academic teaching contexts? In my humble opinion, no. It is a tool that could contribute very positively not only to language learning but to teacher development and most importantly to learner autonomy in a way very fitting to today’s COVID-19 pandemic online teaching and learning challenges. Forced into a world which requires learners to be technologically adept and more independent provides the perfect opportunity to explore the world of DDL especially when more corpora can freely be accessed nowadays, and a quick online search produces results for corpus software tools that are not just free but also user friendly. 

A few corpus resources

For those interested in exploring corpora and their use further, some indicative sources follow:

General reads in corpus linguistics 


  • British National Corpus (BNC)
  • Corpus of Contemporary American English (COCA)
  • The International corpus of learner English (ICLE) (via CQPWeb)
  • Longman Learner Corpus (via CQPWeb)
  • Cambridge Learner Corpus (not freely available)

Classroom resources based on corpora 

Corpus software tools


Cheng, W. (2010). What can a corpus tell us about language teaching? In A. O’Keeffe & M. McCarthy(Eds.), The Routledge Handbook of Corpus Linguistics (pp. 319-332). Routledge.

Gilquin, G., & Granger, S. (2010). How can data-driven learning be used in language teaching? In A. O’Keeffe & M. McCarthy (Eds.), The Routledge Handbook of Corpus Linguistics (pp. 359-370). Routledge.

McCarthy, M., & O’Keeffe, A. (2010). Historical perspective. What are corpora and how have they evolved? In A. O’Keeffe & M. McCarthy (Eds.), The Routledge Handbook of Corpus Linguistics (pp. 319-332). Routledge.

McEnery, T., Xiao, R., & Tono,Y. (2006). Corpus-based language studies: An advanced resource book. Taylor & Francis. 

Soruç, A., & Tekin, B. (2017). Vocabulary Learning through Data-driven Learning in an English as a Second Language Setting. Educational sciences: theory & practice17(6), 1811–1832.

Sripicharn, P. (2010). How can we prepare learners for using language corpora? In A. O’Keeffe & M. McCarthy (Eds.), The Routledge Handbook of Corpus Linguistics (pp. 371-384). Routledge.

Walsh, S. (2010). What features of spoken and written corpora can be exploited in creating language teaching materials and syllabuses? In A. O’Keeffe & M. McCarthy (Eds.), The Routledge handbook of corpus linguistics (pp. 333-344). Routledge.

Elisavet Kostaki - Psoma
Elisavet Kostaki – Psoma

Elisavet has been involved in various aspects of TESOL for more than two decades: as a teacher, an examiner, a mentor and a teacher trainer. For the last two years she has also worked on various CELTA courses as an approved tutor. It is our pleasure to welcome her to ACE TEFL!