In a step towards common sense, Australia has determined to prohibit computer scoring of national tests.
Now the next step towards common sense would be to get rid of the national tests.
One great piece of news for Australia is that the Finnish education expert Pasi Sahlberg has taken a position at the Gonski Institute of Education at UNSW in Sydney.
GERM and NAPLAN won’t survive Pasi’s onslaught of brilliance and direct talk for long.

To be clear, this is about computer scoring of essay exams.
LikeLike
I’m sure some techies in Australia and elsewhere are very upset that MIT computer scientists Les Perlman has pointed out the major problems with their” robo essay marking” system
We need many more academics like Perlman prepared to question and challenge the outlandish claims and outright lies told by polished techno-salesmen to pawn their software and hardware boondoggles off on an unsuspecting public.
If more people had questioned Bill Gates (a flunky used software salesman), we never would have been saddled with Common Core and all the software and hardware disasters that have occurred in districts throughout the country (in LA and elsewhere)
But at least members of the general public have an excuse for buying into the claims. They are ignorant of the technical aspects.
But computer academics can not claim ignorance and therefore have no excuse.
Academics who do not speak out against robo essay marking and other frauds are complicit.
And those academics who actually produce the stuff should be drummed right out of the academy.
LikeLike
Read Todd Farley’s book (MAKING THE GRADE) to realize that human scoring isn’t much – if any – better. Maybe we could accept that writing isn’t something that can be standardized, much tested by a standardized test? (Not that any other subject can be either, but I’d settle for a minor miracle at this point.)
LikeLike
I tend to agree.
LikeLike
There are undoubtedly many problems with human essay scoring, not least of all that it is inherently subjective (as is any computer algorithm written by humans), but, as Perlman has shown, computer scoring lacks the capacity to distinguish between sense and nonsense.
Perlman has demonstrated that the computer scoring can be fooled into thinking that well written gibberish ( that follows grammatical rules) deserves a high score.
LikeLiked by 1 person
Maybe they can train kangaroos to score the open-ended questions. I understand they are unbiased.
LikeLiked by 1 person
Fred,
Great one. Thanks for the TRAINED KANGAROOS comment. True at so many levels.
LikeLiked by 1 person
“Koalas for Dollars”
A court of kangaroos
Got nothing on koalas
Which testing firms will use
To score and boost their dollars
LikeLike
At least Australian teachers were respected enough to decry the absurd idea of computers scoring writing exams. There is no way a computer can process some of the subtle points of writing like voice, tone or irony. If fact, in America it was determined that computers gave high scores to writing samples containing polysyllabic nonsense as computers cannot “think.”
LikeLike
It’s bad enough when human graders are instructed to follow a rigid rubric that ignores the more subtle aspects of writing for the formulaic, but to replace human graders with a machine that does not even have the ability to distinguish these subtleties is ludicrous. Thank goodness Australia is resisting the tech juggernaut.
LikeLike
The basic issue is that people are far too accepting of output from computers.
Ever since computers came on the scene, the general public has largely treated them like oracles. One techy billionaire (Larry Ellison) even had the hubris to name his software company Oracle.
Cathy O’Neil writes about the many problems with computer algorithms (bias, inaccuracy, just plain wrongness) at the Mathbabe blog.
People would do well to apply a heavy dose of skepticism to anything that comes out of the tech sector because much of it is just garbage from a subject matter standpoint.
Just because you can program a computer does not mean you know anything at all about the subject of the program.
I encountered a great example of this recently when my nephew was doing math problems online. He had to do a series of multistep problems and if he got anything at all wrong, he had to start all over with a brand new series of multistep problems.
If that were not bad enough, there was no indication of what he did wrong or how to correct t his mistakes.
The software is total garbage, written by some brain dead “coder” who does not know squat about teaching.
I worked in software development for a long time and I can tell you that this is very common: people who can code but know next to nothing about the subject of their coding, in my nephew’s case
Teaching math.
LikeLike
And by the way, I’ve worked with a lot of computer programmers who came from many backgrounds (science. Engineering, math, even philosophy) and the best ones were not “computer science” majors.
In fact, from my experience, “computer science” grads are the absolute worst when it comes to “coding before understanding a problem”
CS departments don’t seem to teach the importance of really understanding a problem before coding a “solution”.
LikeLiked by 1 person
“CS departments don’t seem to teach the importance of really understanding a problem before coding a “solution”. SCARIEST STATEMENT. A most serious danger in that these exact “coders” are now taking over the writing of modern-day school (computerized) curricula/lessons.
LikeLike
The same type mysticism applies to statistics. VAM was accepted on face value and widely applied because it was heralded in by Gates with the false assumption that if it is mathematical, it must be unbiased and accurate. It was all false assumptions, smoke and mirrors.
LikeLiked by 1 person
This is a good move. Next to the link to the newspaper report is a discussion of the use of public funds for private education…and discontents with that.
LikeLike
Is it possible that Australia took notice of what happened in New Zealand and now Australians want to follow New Zealand and drive out the corporate reformers of public education?
LikeLike