kitchen table math, the sequel: handwriting and grades and the wisdom of the crowd

Tuesday, October 9, 2012

handwriting and grades and the wisdom of the crowd

Even legible handwriting that's messy can have its own ramifications, says Steve Graham, professor of education at Vanderbilt University. He cites several studies indicating that good handwriting can take a generic classroom test score from the 50th percentile to the 84th percentile, while bad penmanship could tank it to the 16th. "There is a reader effect that is insidious," Dr. Graham says. "People judge the quality of your ideas based on your handwriting."
How Handwriting Boosts the Brain by Gwendolyn Bounds | WSJ | October 5, 2010
I absolutely believe that. And I doubt teachers can turn this bias off. Cognitive biases can't be turned off at will.

Grades and grading are a mess. Probably for writing especially.

Speaking of writing and grades, I had a thought the other day. It's not possible, under the current system, for teachers to grade papers according to an objective standard. An A paper for one teacher is a B paper for another teacher is a C paper for a third.

In theory, a testing company can achieve 'rater reliability' by dint of extensive training sessions, although Todd Farley's account of his experience in the industry makes me wonder.

But is training-up individual graders to apply the same standards as their colleagues (even if it's possible) the best approach?

Maybe not.

Any teacher can (or should be able to) correct a paper's grammar, punctuation, and spelling. I assume teachers are going to agree on grammar, punctuation, and spelling far more often more than they disagree.

Beyond that, however, I'm not sure you actually want a uniform response across teachers. "Writing" as a profession or a business obligation means writing for an audience of more than one reader, and the individuals who make up that audience don't necessarily agree amongst themselves that you've said what you've said or that you've said it well. Writers learn from these disagreements.

Maybe students would also benefit from a 'diversity' of reader reaction?

If I had my druthers, I would scrap the letter-grading of writing altogether, apart from scoring punctuation, grammar, and spelling, simply on grounds that the letter-grading of student writing is simply too inconsistent to be credible.

I would experiment with some kind of Intrade or Wisdom of the Crowd approach. Farm papers out to a bunch of readers who read quickly and check off a thumbs-up or thumbs-down option. Something simple. Then give everyone the results for everyone.

Students would receive a kind of polling or survey result instead of a grade: a rough sense of how well their papers worked for an audience compared to papers written by their peers.

Of course, students would need to be able to read the work of their peers to see what kind of paper produced what kind of global response.

Or -- here's a thought -- perhaps schools could create an extensive set of exemplar student papers that have been 'voted on' by a large number of instructors. As a teacher of freshman writing I would kill to have such a resource myself.

I don't know whether a system along these lines would offer useful or 'actionable' information to students.

But I think it might.

10 comments:

Anonymous said...

If you left the names on the papers the system would certainly be actionable in the proper sense of the word—you could get sued for doing it.

MagisterGreen said...

"I assume teachers are going to agree on grammar, punctuation, and spelling far more often more than they disagree."

Even that may be a step too far...

Personally I think it would be a massive boon for students (and teachers) to have "crowd-sourcing" of the grading process for papers. Have papers be read by a group of people who give a first-impression grade and average the student grade from that. It'll never happen for a whole host of reasons, but it'd be a great thing if it did.

Sadly, though, even a system like that wouldn't be able to get past individual biases for things like grammar, spelling, and so on. But with a wider array of inputs, you might be able to get a more reasonable sense of how a student is doing, never mind the wealth of information that would be shared across teachers about individual students.

As for sharing out papers...you'd need to get "exemplar" papers, almost certainly from graduated students who sign over the right to use their work (most likely with names redacted). Never happen, but should.

Anonymous said...

I always found that grading the first few papers in a batch took forever...and that I then would begin to accelerate. Having read a few set some parameters.

Finally, I started quickly skimming papers first and separating them into piles. Three piles is pretty easy --

good to great, understandable, made a point

I get what you're saying, but had to work at it, your point is a little obscure or not quite supported by your evidence, or the manner in which you convey your point is unclear, wanders, switches at times

The "what?!" papers -- it just doesn't hold together. There are two different things going on, you contradict yourself, you never get around to having a point.

Once you've got a preliminary sort, it's much easier to see differences within each group too (and yes, sometimes a paper looks different when read more closely and might move to a different pile). First pile is As and Bs, second pile Bs and Cs and third pile Cs, Ds, and Fs.

Auntie Ann said...

Perhaps some level of computer grading would help...at least to get in the ball park, and to examine grammar and spelling.

But I love this article on computer essay grading:

NYTimes

>> The e-Rater’s biggest problem, he says, is that it can’t identify truth. He tells students not to waste time worrying about whether their facts are accurate, since pretty much any fact will do as long as it is incorporated into a well-structured sentence. “E-Rater doesn’t care if you say the War of 1812 started in 1945,” he said. <<

>> “Once you understand e-Rater’s biases,” he said, “it’s not hard to raise your test score.”

E-Rater, he said, does not like short sentences.

Or short paragraphs.

Or sentences that begin with “or.” And sentences that start with “and.” Nor sentence fragments. <<

Catherine Johnson said...

Even that may be a step too far...

You think so?

Catherine Johnson said...

I always found that grading the first few papers in a batch took forever...and that I then would begin to accelerate.

I've had that experience ....

Catherine Johnson said...

Personally I think it would be a massive boon for students (and teachers) to have "crowd-sourcing" of the grading process for papers.

I'm so happy to hear you say that!

I'm really just thinking out loud, but the problem of teacher subjectivity - or maybe just teacher 'individuality'? - is real, and I don't see a way around it.

Catherine Johnson said...

As for sharing out papers...you'd need to get "exemplar" papers, almost certainly from graduated students who sign over the right to use their work (most likely with names redacted)

Absolutely - that would work.

I am in chronic need of exemplars; I have, at this point, exactly one to my name.

Catherine Johnson said...

I can try writing exemplars myself, but the one I wrote so far was way too sophisticated & doesn't work as an exemplar of good FRESHMAN work.

It's neither easy nor simple to imitate a good 5-paragraph essay written by an 18-year old.

Catherine Johnson said...

Auntie Ann - I like long sentences & short paragraphs!