"I know no safe depository of the ultimate powers of the society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is not to take it from them, but to inform their discretion by education. This is the true corrective of abuses of constitutional power." - Thomas Jefferson 1820

"There is a growing technology of testing that permits us now to do in nanoseconds things that we shouldn't be doing at all." - Dr. Gerald Bracey author of Rotten Apples in Education

Search This Blog

Wednesday, November 2, 2011

Has no one in education read any Isaac Asimov?

The invention of the computer has supposedly heralded a new dawn of civilization.  It has triggered an information explosion that has transformed society. But anyone who has read about Clarke’s HAL (the Heuristically programmed ALgorithmic computer) in Space Odyssey,  Sky Net in Terminator, or Asimov's VIKI in I Robot knows the danger of turning over human control to machines.  Though thousands of times faster, with the ability to consider millions more data points than the human brain, computers still have some basic limitations. They only do as they are told (programmed) and their exercise of pure reason does not jibe with human reasoning.  This is why SciFi story lines based on computer reliance typically end so poorly for humans.  So why are we touting the benefits of turning over the evaluation of our children’s education to computers?

Various software programs (Project Essay Grade [PEG],  Intelligent Essay Assessor [IEA], E-rater,) that utilize algorithms to grade student writing samples, are being pushed as the solution to the growing demands on teachers’ time to grade the increasing number of assessments required by NCLB, ESEA, CCS and other local assessments.  This software could only be developed once “good writing” was analyzed and broken down into component parts.  This was done by Jane Schaffer who gave us the alphabet soup of the Schaffer Writing Method.  Those who have children in school now should be familiar with her “chunk paragraph” and its use of the TS, CD, CM, CM, and CS. This is essentially paint by numbers for writing.

The theory goes that if you have a student who has no idea how to write an expository essay, you give them this framework to fill in and they will end up with something literate.  They will have a Topic Sentence (TS),  Concrete Details (CD), Commentary (CM) and a Concluding Sentence (CS).  This much is true.  Will you have something that makes sense or is compelling? This is not guaranteed by the Schaffer method.  The finished work could be full of dangling participles, noun-verb disagreement, or erroneous conclusions.  Most humans reading the paragraph would recognize these errors. However, if you sprinkle in the right amount of key words, concepts and have the correct sentence order, you can fool a computer.

When the University of California at Davis tried out such technology a couple years back, lecturer Andy Jones decided to try to trick e-Rater.  Prompted to write on workplace injuries, Jones instead input a letter of recommendation, substituting "risk of personal injury" for the student's name.

"My thinking was, 'This is ridiculous, I'm sure it will get a zero,'" he said.  He got a five out of six.

A second time around, Jones scattered "chimpanzee" throughout the essay, guessing unusual words would yield him a higher score.  He got a six.

A University of MO sociology professor, Ed Brent, developed his own grading software, loading it with keywords and term relationships that the software would scan for.  Once this information was learned by the students, they used it to better their scores.

In Brent's class, sophomore Brady Didion submitted drafts of his papers numerous times to ensure his final version included everything the computer wanted.  "What you're learning, really, is how to cheat the program," he said.

Giving the teacher what they want is the current goal of all students, so this student’s approach is neither unreasonable, nor unethical.  He is learning something, just maybe not what he, or his parents, thought they were paying for.

As we rely more and more on computers, success in the future may be defined by one’s ability to fool the computer - a hacker’s nirvana.   The experience in Atlanta, where 150 teachers were found to be changing students answers on standardized bubble tests in order to improve their scores, shows that this is a lesson humans can learn quickly and well.  The Atlanta school district relied on computer scanners to determine how well the students were doing.  It took almost a decade, and a few humans, to notice that anything was amiss with the scores.  We are currently generating so much data, that computers can handle, but few humans are reading.   What is being missed in this data and how much more will be missed in the future as we increase the amount of data exponentially?

The Atlanta school district incident resulted in many people losing their jobs. But in all the stories about the cheating, no one ever mentioned what would be done for the victims, the students who didn’t actually perform that well on those tests yet were promoted anyway.  Their education was a casualty of standardized testing and computer analysis.  

We have yet to achieve Asimov’s first law of robotics; "a robot may not injure a human being or, through inaction, allow a human being to come to harm." Since a robot is really a computer with the power of locomotion, maybe Asimov should be required reading for all school bureaucrats so they can make sure their computers at least follow the first law.

No comments:

Post a Comment

Keep it clean and constructive. We reserve the right to delete comments that are profane, off topic, or spam.

Site Meter