Wednesday, 23 February 2011

What's red and green and pointless? The endless, anxious debate about the colour of marking.

What's got two thumbs and couldn't give a damn what colour his marking pen was?
*points two thumbs at chest*

The meme de jour of horror flicks is to have the  final frame hinting at the imminent return of the hellish antagonist- see: Carrie; Saw; Freddy; Jason, ad nauseum. Well here's another teaching myth that I thought had been staked a long time ago, but apparently keeps rising from the grave with  the certainty of sunrise. Does it matter in what colour you mark students' books?

No it doesn't  
I only mention this because someone emailed me this question recently, and I had to rub my eyes and pinch myself (not easy) to check if I was dreaming. Are people still asking this? Apparently, yes. Dracula has returned. So it's time to dip my crossbow bolts in holy water and bless my silver candelabra, and get ready to knock the brains out of this one, although believe me, it won't take much.

When I started teaching, this was received wisdom; it was dogma; it was part of the catechisms of the Church of Progressive Teaching: do not mark in red ink, I was told. When I asked why, I was solemnly told that it was 'bad'. Which is great, because for a minute I thought they were going to be vague about it. I doubted it then, because it just seemed so counter-intuitive. How could it matter in any real sense what colour I used? But, like many axioms absorbed in the infancy of one's education, I complied, and dutifully stocked up on soothing, somehow more supportive shades of emerald. I wondered then which shade in particular was supposed to have the best effect? I think more work needs to be done.

But let's settle this. There is no research whatsoever to support the view that marking should be done in green ink, purple, vermilion, or a thousand other shades of the spectrum, as a preference to red. Let me repeat that: there is none. So why do so many people think it's true? I've investigated, so you don't have to. Jus' doing my job, ma'am.

Well, what there has been is an enormous amount of psychological investigation into colour preferences in an enormous number of groups. There's also been a lot of studies that have looked into the symbolic associations that people have towards certain colours. This is nothing new: people have always attached meaning to just about everything: black has connotations of mystery, magic, fear, fascism; blue connotes to calm, the ocean, etc. You can add your own. There are recurring themes that reflect the cultural expectations of the group investigated. Within each group there will be understandable variations ascribed to any particular colour- I find Sunflower Yellow irritating (mainly because I associate it with one of the companies I used to work for), whereas I'm told by SENCOs that autistic children apparently fall into soporific calm at the sight of it. I have no idea.

This kind of thing can be studied quite easily- get a sample, ask them what they associate with each colour, collate the data. Or perhaps be even more clever and do it a bit more 'blind'- don't tell them what you're investigating, and show them pictures of people wearing different shades and hues and ask them to express their thoughts towards them, that sort of thing. The point is that this kind of data is easy to gather, and assuredly marketing people have been doing this since dinosaurs walked the Earth and they wanted to know what colour of loincloth sold best to AB1 Hunter/ Gatherers.

Of course, the idea that the colour of your marking pen matters goes beyond even this: the idea seems to be that people associate the colour red with aggression, threat, danger and negativity. People- seemingly quite sensible ones- have expressed the opinion that they remember  their own books being 'covered with red ink,' in a process that they describe as traumatic. Ergo, we should stop using red ink and get busy with shades more arborial.

Show me the research

Well, quite. So what research is there to support this? I'm happy to report that the answer is 'bugger all'. Curious, I decided to do a bit of digging, and I was amazed to see that nothing on paper could explicitly support this theory at all. In fact, even the most recent research seems to struggle to prop it up, or barely offers any insight into the question except in the most tangential way. To give one example,  a 2010 study in the European Journal of Social Psychology by Rutchick, Slepian (of the gorgeously named Tufts University) and Ferris offers the cutting edge of research into this area, and frankly, it's a hoot. It shows, or attempts to, that using red ink can prime markers (not students) to tackle papers more aggressively, more critically. Its central premise is that just seeing red (ah, such a loaded term) is enough to make markers...well, see red. Those using red pens in the research seemed to grade papers lower, and notice a higher level of mistakes than the other, non-red-ink group. They also seemed to indicate that, when confronted with word stems to complete, they usually went with more negative, aggressive ones over the altogether more harmonious, Scandinavian alternatives.Let me give you some examples; see how you do. Finish these words:

MIN_ _.
WRO_ _.

Now, I don't know about you, but I went straight for 'fail', 'minus', 'flunk' and 'wrong'. And so did more of the red ink group. It's results like these that convinced the researchers that they were on to something with their red-ink bad hypothesis. The blueys, incidentally, had higher proportions of 'fair', 'minty', 'clunk' and 'wrote' apparently. You can't argue with science.

Except that you can certainly argue with this science. For a start, what's to say that the 'negative' connoting words weren't simply more common idiom? "Minty' isn't exactly something that trips off my tongue regularly, but then, I don't sell toothpaste. And what's to say that there weren't fewer toothpaste salespeople (for example) in this group than in the other? It's these kind of uncontrolled variables that knock the guts out of a piece of research, because without reassurances that one group isn't unfairly weighted with certain types of respondents there's no way of knowing if your analysis is statistically significant.

What about the other part of the research- that the red inkers were more stringent/ harsher/ more negative in the number of mistakes they corrected? Again, we need to know if there was any meaningful way of making sure that one group wasn't unfairly (or should that be unfailly?) loaded with pedants, grammar Nazis and careful essayists. I checked the paper to see how the group was selected. It says, and I quote:

'The current findings are qualified by additional limitations, primarily concerning the participants in the studies. Due to time constraints associated with conducting the experiments in a realistic setting, little was known about the participants beyond their presence in the university environment; their age, ethnic background, level of education, and other factors were not assessed. Several of these individual differences, such as verbal ability, educational background, and field of study, could influence participants’ ability to detect errors, their propensity to mark them, and the harshness with which they make evaluations.'

You don't say. So in essence, we have a bunch of people marking with red pens, and a bunch of people marking with blue pens, but we don't know if there are any factors in each group that could produce the results that we found. But we think it's got something to do with the pens. In fact, it's not just the pens, it's the fact that the 'we propose that the red pen effect is driven by increased accessibility of the concepts of errors, poor performance, and evaluative harshness.'

Or, if I can reformulate that again, 'we don't know why, out of all the possible differences between the two groups, there should be a statistical difference in such things as number of errors recorded etc, but we've decided it's the colour of the pen ink, because that's what we're investigating.' Ladies and gentlemen, it's research like this that makes your line manager tell you with an air of authority to ditch all your crimson ballpoints and go green in the classroom

So what do the authors think of this inability to account for the variations in the participants? 'However, these uncontrolled differences should manifest as random variability, and thus make it more difficult to detect the effects we report.' In other words, 'We don't know if they have any factors that could affect the experiment, but because we don't know we reckon it should all work out jess' fine, than 'ee.' Give me strength.

I often blow pretty hard about social science, for reasons just like this: although I don't necessarily believe that all social science should be circumscribed by the methodology of the natural sciences, I do think that whenever it makes a specific empirical claim about the way people think and routinely act, that there should be some kind of empirical method to show that this is the case, and that the experiment can be meaningfully reproduced in other situations, tested and confirmed. If all you want your paper to do is to start a conversation, or to add to a debate, then by all means keep it anecdotal, or unrepresentative, or subjective. But when you start trying to prove something predictive, or worse, start telling others to change their behaviour on the basis of your research, then you need to turn up to the fancy dress party with more than a fig leaf, saying you've come as Adam.

Surely there's more evidence?

Hang on, I hear you say, is that all the evidence there is? Well, yes, at least on this specific topic. Oh to be sure there are many studies that show that certain colours have certain connotations amongst certain people. And there are many other studies that show that certain colours can 'prime' us to respond in a variety of unconscious ways. But nothing- and I believe I'm throwing my perfumed gauntlet out here- NOTHING to suggest any clear link between the use of the red pen and...well, what is it the critics say is happening anyway? For one thing, the claims they make are maddeningly vague: red ink is 'negative'; it's 'discouraging'. Oh really? And how would you even go about measuring that kind of thing? How on Earth could you control for it? Do we ask 1000 children aged 14 to describe their feelings towards corrections in their books? And another 1000 to say how they would have felt if it had been in mauve rather than vermilion? *slaps forehead at the boneheaded nonsense of it all*
But hang on: our heroes have something to say about the claimed 'effect' of red ink on students:

'To our knowledge, this demoralization has not been empirically demonstrated, and would be an important complement to the current findings.'

That might just be the understatement of the red-inked century. Let me just repeat that:' no empirical evidence'. No studies. None. No proof whatsoever other than a gut feeling in educators that when kids look at a book covered in red ink corrections, they get a funny feeling in their tummies, and perhaps it's the red ink that's to blame? As many people have commented, it would seem perhaps more intuitively likely that people brought negative associations to the fact that they were corrections rather than the fact that it was red. After all, red can have millions of associations: danger, certainly- the red of a wound - but also thrills, passion, romance-  the blood racing in your veins because you're alive, for example.
Finally I would like to make one point: maybe I WANT kids to feel a little bit alert when they see red on their books? Maybe the connotations aren't exclusively negative, but rather represent a state of heightened alertness corresponding to increased attention paid to corrections. Maybe, just maybe. The point is that, as Ben Goldacre famously repeats, 'things are a bit more complex than that.'

And I sincerely hope that I haven't offended anyone by typing in black ink. Click here for a more soothing draft in magenta.

Finally, I think what particularly offends me about this subject  is that, were it relating to an aspirin, or a new technique for removing surgical stitches, it would be subjected to an enormous level of scrutiny before it could be released, as it were, into the wild. Not so in the field of educational science, where mutants, hybrids and sickly, runtish ideas are set free to breed and settle where they will. If this was an aspirin, it would have been laughed at; in education, it's adopted as a mantra. It would be laughable were it not so ubiquitous, or so representative of the slavish manner in which teachers are expected to assume a new position, no matter how servile, or follow any fashion, no matter how impractical, if their masters demand it.

I haven't been able to track down exactly when this idea escaped the laboratory from which it spawned, or how, or by whom. I can find reports of schools in the UK (primaries at first, then secondaries, Australia (Queensland) and America (loads of places) adopting the abolition of scarlet scribing dating back to 2003, in this report from the BBC. To quote:

'Penny Penn-Howard, head of school improvement for Sandwell Council, said: "The colour of the pen used for marking is not greatly significant except that the red pen has negative connotations and can be seen as a negative approach to improving pupils' work. Therefore, it is quite legitimate for a school to have a consistent policy that it uses a different colour."'

Which is another example of why I'll be glad to see the back of some people who claim to be employed to 'improve schools'.  If it isn't significant, why have a policy? How does red connote 'negatively' etc? Does that mean that tomato sauce and Christmas have negative associations? Blimey, someone better call the head of marketing at Coca Cola and tell them they've been getting it all wrong. Penny Penn-Howard has the answers.

My Stella Challenge to the red-ink flat-earthers

I'm not James Randi, and I don't have a million dollars, but I bet anyone a pint that they can't produce meaningful research that shows that the colour of the pen has any significant effect at all. I suspect my pint jar is safe from the toffee hammer for some time. In the meantime, I'll be choosing to mark in red as much as humanly possible, more in spite of the apparently still current dogma than for any other reason. Mind you, it stands out nicely against all the black ink of my student's work.  
Plus of course I tend to write in human blood. I like the connotations. It shows I care.


  1. I do purple. It's my favourite colour. No other reason. (Oh, except that the kids tend not to write in purple. So they can see what I've written. Tends to be helpful.)

  2. Unbeliever! Didn't you know that purple alienates the borderline D/C candidates, by alluding to an unachievable state of royalty?

  3. I use invisible ink - it does not infringe pupil rights by implying incorrectness where none may exist. Works well.

  4. I mark in pretty colours, pink, green, turquoise. It helps to keep ME interested. Have you ever marked a full set of books where the children make the same mistake across the board and found yourself gently correcting the first few books but ( as you get more and more RAAAAR )the last person in the class bears the brunt of your wrath and has the words DO NOT 'whatever the grammar crime' almost CARVED into the page in massive latters? That was red pen too.

    Really interesting reading. I will share this at the next staff meeting!

  5. Don't mark at all: instead, get the Student Council to peer assess the work. Perhaps they can take the lessons too.

  6. Personally I like the argument that red has connotations of wrong, because teachers scribbled all over books in red. How do we fix it? Scribble in purple! Or Green! Because then children will be fooled into thinking they're wonderful because there's no red. Sounds like the same logic that had one of my teachers calling her top groups 'Z' because "if I call them A, they think they're the top group and we don't have top groups, just one that does harder work" (her others were Y and X, shockingly).

  7. From the lofty heights of absolutely no research at all, when I ask my tuition students "What colour shall we use this week?" the answer is often ...... red. Some students have a strong, consistent preference for a special colour. Poor things, I insist on using a different colour each week for easy checking of amount of work done when I review each month or so. So purple or orange only gets a run on alternate weeks.

    But by and large, you'll find every second week marked in red for many students. It's bright, it's clear and it seems to be attractive to those with no particular preference among colours generally..

    Where's my grant? I must be able to get some money for this from someone.

  8. Having just read the paper you linked to, I'm shocked that it actually made it into a peer reviewed journal. It reads like it was written by a first year undergrad.
    This would be what Feynman refereed to as a cargo cult science.

  9. Great stuff. There's a fairly amazing poster by the graphic designer David McCandless which you can find here:
    - the poster shows the various connotations which different cultures associate with different colours. Oddly enough I see no mention of red as being bad for marking in any culture...

  10. So don't mark in red because it represents love and passion?
    Such a weak argument as tom pointed out.

  11. The only useful things I can find to say about the choosing the shade of one's marking ink are:
    1. Avoid the colour the kid has written in for fairly self evident reasons
    2. I like peer/self assessment to be done in a different colour so that students can distinguish my marking from theirs.

    I mark in red for preference, but when push comes to 'will you bloody well mark our books, sir" I'll use whatever pen's at the bottom of my bag.