This Delta Notes series has come about because I am packing up all my stuff to move out of my flat and have found my Delta notebooks. I don’t want to put them in a box (got plenty to store as it is plus it’s pointless…) and let them gather dust, so thought I’d write up the notes I’m interested in keeping and get rid of the notebooks instead! I will also add some reflections at the end of each set of notes. Feel free to share opinions, add ideas, argue against any ideas you disagree with etc by commenting using the comment box beneath the posts. (These are just some of my notes from Delta input sessions – I may have misunderstood or missed something: there was a lot of information flying around that semester!)
Here are my (written up) notes from a Module 2 input session on error correction:
Errors are evidence of learner development and are made for a variety of reasons. They are something we, as teachers, have to deal with on a regular basis. To do this effectively, it helps to have a clear understanding of why errors might be made and what can be done with them.
If a learner makes a slip, they have the requisite knowledge, e.g. that in the third person present simple, we add –s or –es, but do not produce the item correctly. In this case, they are likely to be able to self-correct quickly. Errors can also provide evidence of learners’ systems – if a learner produces the same error consistently, it is systematic. Learners may also make attempts to say something that they have not learnt how to say, and not quite manage. This provides information about what they are ready for – what they can do and what gaps there are in their knowledge.
From the teacher’s point of view, some errors are covert i.e. learners produce something correct but it wasn’t what they wanted to say and this isn’t obvious to the teacher, while some are overt, i.e. obvious.
Errors can be caused by incorrect L1 transfer. However, it is worth remembering that transfer can often also be positive. Errors can also be intra-lingual, developmental and systematic. These refer to learners’ current awareness of the language and can be a result of over-generalisation or incomplete application of rules. They could also be a result of mis-teaching, where there is lack of clarity, or over-teaching, where some language feature, e.g. –ing, gets stuck in their head!
A breakdown of different types of errors:
- Suprasegmental – word or sentence level mistakes e.g. incorrect intonation or stress.
- Segmental – sound-level mistakes e.g. mixing up consonant sounds /p/ and /b/
- Combinatorial – mistakes relating to how sounds are linked e.g. producing consonant clusters incorrectly.
- Incorrect selection of a word/phrase
- Inventing a word/phrase
- Transferring words/phrases from L1 incorrectly
- Distortions of words e.g. kitchen v chicken
- Words could be too specific or too general for a given purpose
- Use of a superordinate instead of a more appropriate hyponym
- Use of the wrong collocation
- Production of an incorrect form
- Wrong level of formality
- Unintended connotation
- Covert: a correct form but not the intended form
- Morphological (but this can be a pronunciation error rather than a grammatical error e.g. not pronouncing the final ‘s’ rather than not using plural)
- Confusion regarding function e.g. Is this ‘Can’ for ability or request – requires interpretation of language in context.
- Literal meaning could be different from use e.g. “It’s cold in here” literally means the temperature in here is quite low, but it can be used as a request to close a window/put on a heater etc
- Taboo subjects
- Learners may mis-process input and give the wrong response.
How can we deal with learner errors?
If they are overt, we can deal with them instantly or wait till a more appropriate moment.
If we decide to deal with learner errors instantly, how can we go about this?
This very much depends on the error type and on various contextual factors (what learners are used to, the focus of the lesson phase, how much time is available etc)
One way of dealing with errors:
Ask for repetition: this signals you aren’t sure of what the learner has said and gives them the opportunity to self-correct if it is a slip. It also gives you thinking time! (I.e. time to decide how to deal with the error)
Ask for self-correction: learners may have missed your previous cue or attempted to self-correct but not corrected the error.
Ask the rest of the class to try and help: this engages all learners in what started as a one-to-one interaction and maximizes on the different developmental stages and sub-levels that are present within a single class.
If nobody can help: either give up and provide the answer or give prompts that may help learners to reach the answer. (Worth remembering that you can’t elicit what learners don’t know and considering whether the benefits of laboring over a particular error balance out the amount of time spent.)
If somebody can help: Ask them to repeat their correct form. Get everyone to say the correct form. Then ask the learner who originally made the error to repeat the correct utterance – this reinstates the class as it was, but with the correct form. (Very often, there is no need for a “teacher model”, except for pronunciation – and even with pronunciation, learners will often repeat better from a learner model.)
It is important to show awareness of errors: If you are not correcting errors, it is important to be explicit about why you are not correcting errors. This might relate to the focus of the lesson phase (i.e. you might be focusing on fluency development and so may be less worried about accuracy at that point) or your plan (i.e. you might plan to do a delayed error correction feedback phase after an activity rather than correct during the activity). However, it is also very important to respond to what learners say, not only focus on how they are saying it.
When a learner produces language, ask yourself:
- Is this adequate?
- Can I get more?
- Do I want more?
Here are some of my reflections on error correction:
Error correction is, I think, one of the minefields of ELT. Learners desperately want it, and may feel they are being short-changed if it doesn’t happen. Teachers may have good reasons for not doing it, or may be doing it in such a way that learners are not explicitly aware that they are being corrected. Teachers might also get into the habit of always using the same narrow selection of error correction techniques, which may not be effective for some of the learners in the class. Of course, what constitutes effective is another can of worms! I think there’s a lot to be said for variety and experimentation, where error correction technique is concerned: Different techniques will be better suited to certain error types, different learner preferences and so on. Experimentation – and, of course, post-lesson reflection on this experimentation – can enable a teacher to build up a range of techniques that he or she will be able to draw on when the need arises.
(For this, I recommend having a look at Classroom Management Techniques by Jim Scrivener, which contains many practical ideas, and the reasoning behind them, to try out: Though it is not specifically about error correction, there is a useful chapter on eliciting (p139 -145), which is applicable. Also have a look at his Learning Teaching book, specifically chapter 14 “Toolkit 2: focusing on language 1. Error Correction” p298 -302. (NB link and page numbers refer to second edition, which I have, but I gather there is a third one now…) Finally, there is a very good chapter in Harmer’s The Practice of English Language Teaching, Chapter 8: “Mistakes and Feedback” p137-152, which I’d recommend reading. In terms of the Delta, as far as I understand it, you are supposed to try and demonstrate that you are able to draw on a wide range of techniques, within an assessed lesson, so all the more reason to have a read and get experimenting if you are a Delta trainee – or a prospective one!)
Errors can be a great source of further learning, but only if they are used as such. For example, if you are doing whole class feedback on a listening exercise, and a learner provides an incorrect answer, merely providing the correct answer will probably not result in much learning. However, if you involve other learners and explore the cause of the misunderstanding, then learning opportunities increase. Some errors may, of course, not be worth spending too much class time on – this comes down to teacher judgement and may be influenced by factors such as the aim of the activity, how it fits into the sequence of activities that make up the lesson, whether you think the error is something that learners should already know/be able to produce correctly and so on.
Finally, I think it can be valuable to involve learners in negotiating how and when error correction should take place. For example, if you are going to do a speaking activity, ask them if they want to be corrected during the activity or to be given feedback once they have finished speaking. Depending on the activity goal, your preference may be the former or the latter, the learners may (think they) want the opposite. Correct those learners who request it while they speak, correct those who request delayed feedback when they have finished, then once the activity has been completed and all feedback given, briefly discuss the pros and cons of each method with them. Elicit their ideas before giving yours, and explain your choice of method isn’t arbitrary but based on what you think will benefit them the most for any given activity. When you experiment with new techniques, involve the learners by explaining what you are doing and asking for their feedback afterwards. Hopefully this kind of discussion and learner involvement will also increase learners’ trust in you, and what you are doing with them, as well as giving you extra evidence to reflect on after the lesson.
Lightbown and Spada (2006:125-128) “Corrective feedback in the classroom” in How languages are learned (third edition) Oxford University Press, Oxford.
(Usefully describes different types of error correction – explicit correction, recasts, clarification requests, metalinguistic feedback, elicitation and repetition, giving examples of each)