Correcting written work

Coming back to work after two weeks of freedom and cake eating is always difficult. Coming back to work to face the pile of corrections you put off is a trauma. Instead of getting stuck into it, I've found myself thinking a lot about correcting written work over the last few days - the value and the best way to go about doing it. It can be a bit of a depressing rabbit hole to disappear down - the what is the best approach conundrum - but I'll do anything to put off correcting.

This article by Jay Schwartz argues the case for using correction codes (e.g. Sp = spelling mistake). I'm not a fan. The idea seems good but it never seems that workable. For it to work, you have to spend a lot of time indoctrinating students into the code. And even then, it can be confusing or unhelpful. Sp is fine - grab a dictionary - but if you write WW (wrong word), are the students really going to figure out what the right word is? Surely when they were writing, they thought the word they chose was the right one. Is a WW going to be all that helpful? How do they know the new word is better than the old one? If they show the correction to you, do you then write another WW if it isn't the word you had in mind? And are students really motivated by these codes? I wonder.

I believe Truscott kicked things off by calling for error correction to be abandoned - that it doesn't work and is demotivating. This article  by Ronald Gray offers a very thorough overview of research into the effectiveness of grammar correction, most of which finds that specific grammar correction (be it highlighting mistakes, using correction codes or specifically identifying and correcting the mistake) doesn't really work. But like most teachers, it can be hard to put down the red pen. How will they know its a mistake if I don't tell them? Sure, I can point out stuff about paragraphing, topic sentences, etc., but can I just stand by and watch a verb be conjugated like that?

In this video, Jeremy Harmer discusses the sensitivity of correcting errors; the judgements teachers need to make in deciding whether to correct or not. This talk relates to error correction of spoken English, but the sensitivity around correction equally applies to giving feedback on written work.

Scott Thornbury makes interesting points here in relation to general error correction (as well as leading a discussion here on what constitute errors and how to deal with them). One specific suggestion he makes which could be directly applied to written error correction is to...

"Recast it.  Reformulate the learners’ interlanguage productions into a more target-like form. This is not the same as correction. It is simply a way of indicating “I know what you’re trying to say; this is how I would say it”."

I like this. But sometimes, I am not sure what they are trying to say.

In short, I find correcting stressful because I am not sure of the best approach; because I have a lot to do and little time to do it; and because I have to balance the approach I take with the students' expectations. Saying that, I find it helpful to think less about correction and more about goals.

• Am I correcting simply to establish a grade or give a student an idea of their level? 
• Am I correcting for the purpose of helping them to improve their language skills? 
• Or am I correcting because another human being has tried to communicate something to me and rather than correcting, my goal is to offer a human response to what they have written?

Of course, really, I should be first asking myself, what was the student's goal in writing this for me. This then informs the way in which I respond to what they have written. If a student writes looking for a grade, and I don't give one, that could be demotivating. If the student wants feedback, and I don't give any, that could be demotivating. But if a student wants some form of human reaction, if their goal was to communicate, and I give a cold assessment of their grammar mistakes, that is arguably the most discouraging response I could offer.

So generally, I find the safest option is to give a bit of all three. Give them a grade. Point out an error or two (perhaps using the recasting method Thornbury suggests). But most importantly, respond to it. If there is an idea that I like, I say so. If there is something that I disagree with, I say why. If I don't understand something, I (kindly) ask them to explain to me. If I think of a related topic they might be interested in, I suggest they check it out.

I'd be very interested to hear how other people go about correcting/giving feedback.

Happy New Year.