I mentioned in one of my recent DELTA diaries that there was a lot of information about error correction and feedback and this was an area that I wanted to reflect on more. After writing for a while (and reaching 2000 words) I realised that one post wasn’t enough so I will have more posts coming soon. But even this might be a bit too much for you to read in one sitting. If that’s the case then why don’t you bookmark it now or put it to the side till you do have time to read it.
Why is error correction important?
To many people “error correction” is seen as the essential function of a teacher. It’s rare to find a student who wouldn’t want/expect some form of error correction and even rarer to find someone who would want no feedback at all (I’m not saying it doesn’t happen but it’s rare).
A students can read the rules of a grammar book, memorise a list of vocab, practice reading different texts and listening to audio texts via the internet and generally find great resources for learning…but even if students have the “answer sheet“, error correction and feedback is something that they won’t necessaries get. Good feedback will help them identify their strengths and weaknesses, why they got something wrong or what they did well and help them progress to the next stage.
Of course unconstructive (or destructive) feedback can hinder or even reverse progress so this is an important area and one we need to think about, research and put good practices into…er…practice.
The general “rule” of error correction and feedback
The classic CELTA rule for feedback goes something like “during more controlled tasks where students are restricted in their options and focusing on a particular language item, errors should be corrected straight away so that students don’t incorrectly learn a language item. In freer tasks feedback should be delayed so that communication isn’t inhibited and the more general goals of communication in realistic tasks can be achieved.” or in quicker and fewer words
“In the Practice stage correct straight away, in the production stage delay feedback.”
There is a lot of truth in this general rule but it’s certainly not the end of the discussion.
Some inspiration from Audiolingualism
In Audiolingualism errors were seen as very negative. If a student reproduced an error too much they would possibly embedded the error into their memory rather than the correct form! If you are a student of a foreign language (and you should be) I’m sure you can think of a word or two where the mistake is deeply entrenched in your memory. For me, in Russian, it’s the different between the words for “Pity” and “Hot” which can lead to some awkward conversations.
Later investigation into language development changed the perception of errors to that of experimentation and part of natural learning (a key example is irregular past tense verbs where students always add -ed endings. Some of my favourites that I’ve heard recently from a young bilingual child are “I dided it” and “I swamed.” As such mistakes and experimentation were encouraged. [This post by Luiz Barros is great on subject verb agreements and their natural development]
However, I find the response from the Audiolingualist was very interesting. They used very restricted situations so that students “couldn’t” make mistakes and only practice the target language. Having studied with a method based of audiolingualism I can say that this certainly helped reinforce my confidence by knowing I wasn’t making many mistakes. However, restrictive tasks can be very boring and don’t really help prepare you for “real tasks.”
Real speech can go off in many directions away from the original topic, by restricting things too much we don’t prepare students for spoken discourse.
Some inspiration from The Natural Approach
I remember one day in Russian class having a bad day. The teacher seemed to be correcting almost every sentence I made and I was getting fed up. In the end I just shut up. I couldn’t face making more mistakes and being corrected more so I let others speak. Krashen pointed out that sometimes when we correct students they will just shut up and stop trying to express themselves or at best reduce to using simpler constructions rather than experimenting.
The Natural Approach suggested that taking a more relaxed approach to errors might help encourage that experimentation and had further merit as some errors will naturally be corrected over time and exposure to the language. [See the previous mentioned article by Luiz]
Error correction won’t always result in a student just shutting up or using easier language constructions but we need to avoid this from happening in class.
In general the trend in ELT has been from a strong focus on error correction and feedback to more relaxed attitude which of course has inspired some others to take a different approach. The demand higher movement has pointed out that maybe we are now too soft on our students and aren’t pushing them hard enough. We accept their single word answers and don’t take the opportunity to work on their pronunciation, or extract a full sentence from them.
I used to get very annoyed by students who would do “short cut answer” saying only “1, A” for a matching activity or something similar. How dare they be so lazy and couldn’t they see that this would actually be further practice for them! However, I realised they were doing the most natural thing in the situation and if you asked any native speaker what the answer to number X was they would do the same thing to. It’s not exactly a natural speaking task. So a better form or task to check answers is more important in my opinion rather than forcing the students to do the unnatural fuller construction
However, the point remains that when students are challenge and find a task more difficult they value the task more (and skill more). Getting the balance right is important…so what factors do we need to consider?
The error, the student and the situation
There are three important factors in deciding whether or not to correct the students error and if you think back to the statement at the start it is only really considering the situation (Is the goal practising and remembering the language item or is it communication.) But the other two factors are equally important in deciding whether to correct or not.
Often we focus on the errors that we hear the most and that have been taught longest ago (Basically third person s, I’m willing to bet that almost every class someone makes a third person s mistake, in fact I’d bet in most conversations a Native speaker will correct a third person s mistake). But are they really the best criteria for correction? Especially when we consider that some errors seem to just take time for a student to correct, perhaps we shouldn’t necessarily focus on most common most.
But what criteria should we use?
Is this an error that is related to this task or is this an error where the student is trying to operate above/beyond the task? Is this something they haven’t learnt yet but are experimenting in? Basically the more relevant to the task the more it should be corrected.
Is this something that makes communication more difficult or impossible or is it just annoying (every time a third person s is omitted, somewhere a teachers eye twitches). This is made more complicated by issues of L1 interference where a fellow student who speaks the same L1 might understand the false cognate or the literal translation of the phrase but someone from a different speech community would not. In Such an example, communication wouldn’t be hindered but normally it would be.
Slips vs errors
Is this a slip, where nine times out of ten the student will say the correct thing but on this occasion or this day they just aren’t (due to tiredness, stress, etc) or is this an error where they still don’t really understand the language or they are trying to balance what they learnt before with some new information they learnt. If it is a slip then you should be more lenient, if it is an error then tougher.
Certain personality types will react to different types of correction in different ways. Let’s consider two straw men extremes
Dave is a shy and unconfident student, he rarely speaks in class and as such is (or appear to be) a bit weaker than other students, especially when it comes to communication tasks. When he is corrected he usually spends time thinking to himself and trying to internalise the knowledge, he doesn’t usually talk afterwards
Anna is very talkative and appears to be very confident. She usually quickly rushes through the practice tasks and always tries to turn every task into a communicative task, She doesn’t always listen to the teachers error correction and frequently makes exactly the same mistake straight afterwards.
Let’s be honest with these two students we should take a different approach to error correction for each of them. With Dave we will probably try to be more subtle, distant the error from him, delay the correction till after communication is complete but perhaps we should also make sure to include very controlled practice to help provide him with the confidence for the latter communicative tasks.
In Anna’s case it might appear that we also need to be more direct with her and this is probably true, her responses suggest that she will carry on communicating no matter what. However, Anna’s confidence might actually be an act. I should mention that Anna is based on a real student with details changed. She frequently tried to communicate in L1 as well and I suspect that she was ignoring the errors more as a way to ignore that she had made a mistake (which she didn’t like)
But what about non PPP frame works
Most of what I have been looking at has been from a PPP (or ECRIF) perspective but not all lessons follow this framework. Sometimes one to one, business and Dogme lessons follow a framework more akin to Produce, present, practice framework to work on material the students bring to the class. In such cases the factors for interfering are still the same.
The same can be said of TBL where during analysis of a model task or the first attempt it can be viewed similar to a “practice” stage and when working on the task it can be viewed similar to the “production” stage.
Ongoing feedback and correction
So far everything has been looking at error correction within the class and not really ongoing feedback or correction but at over “000 words I feel it would be best to save that for another post.
As such I am currently left with a few practical teaching implications and considerations to try to balance during a lesson:
1. What is most important in the situation? Practising the Language item or practising communication?
If it is the language item then try to focus on only the language item (but still don’t expect this to lead to perfection), if it is communication then allow for more errors that don’t hinder communication and delay feedback.
2. What is the students personality like?
How will they react to your correction? Are they going to brush it off and ignore it or take it to heart? Try to adjust the method and timing of correction in light of this.
3. What is the students general ability?
Are they a strong or weak student? Is this a normal error for them or not?
4. What is the error?
Is this an error that they should/shouldn’t be making? Are the consistently making it or was this just a slip of the tongue? Is there some L1 interference or new knowledge of a new language item? Are they overgeneralising a grammar rule? Are they attempting a language item that they haven’t learnt yet?
A conclusion and the next part?
So those are my thoughts for the moment, I’d love your input and advice, I’d love to look at a couple of extra parts such as ongoing feedback, a closer look at feedback as opposed to error correction and tactics and activities for error correction but at over 2000 words I feel I should stop for today.
Do you stick to the traditional error correction framework or do you have different guiding principles?