NYTimes columnist Zeynep Tufekci’s essay, “The Machines Are Coming” describes the impact of algorithms on jobs and offers examples of how computers are increasingly taking over more and more assignments that were formerly thought to require human interaction. Some examples are manning call centers, reviewing medical tests, serving as border guards, and “interviewing” applicants for jobs. One of the rationales for this is cost savings.
Machines aren’t used because they perform some tasks that much better than humans, but because, in many cases, they do a “good enough” job while also being cheaper, more predictable and easier to control than quirky, pesky humans. Technology in the workplace is as much about power and control as it is about productivity and efficiency.
Tufekci describes how machines are deemed to be superior to humans because they don’t “…get sick, ask for higher wages, have a bad day, aging parent, sick child or a cold.” After suggesting way technology could replace humans in the workforce, she does point out one way technology could enhance the quality of the workplace and achieve more than better productivity and efficiency:
In the 1980s, the Harvard social scientist Shoshana Zuboff examined how some workplaces used technology to “automate” — take power away from the employee — while others used technology differently, to “informate” —to empower people.
For academics, software developers and corporate and policy leaders who are lucky enough to live in this “informate” model, technology has been good. So far. To those for whom it’s been less of a blessing, we keep doling out the advice to upgrade skills. Unfortunately, for most workers, technology is used to “automate” the job and to take power away.
I believe technology could be a boon to schooling if it was used to “informate” teachers instead of being used to “automate” instruction… and if schooling was based on “informating” instead of “automating” it would be providing children with a completely different set of skills. Pandora does a relatively good job of divining my tastes in music (they introduced me to many new bands and performers)… and the NYTimes and Google do a relatively good job of divining articles that might interest me (the Times sent me this one!)… and Amazon does a decent job of figuring out movies that I might like. But while Pandora plays music I might like, but my music teacher introduces me to music I can master and play for enjoyment… and while Google and the Times send me articles like the ones I’ve read before, I find the articles posted by friends on Facebook whose insights I value give me newer perspectives… and my daughters who go to movies regularly and viewed videos with me for years have a much better sense of the movies I’d enjoy than Amazon’s algorithm. Computer algorithms can give me information that’s helpful but only humans can give me information I trust. Schooling that uses standardized tests to measure performance could replace teachers with robots Schooling that seeks to motivate students to want to learn independently and values character development can only have humans in charge.
To date, teachers are among those workers whose jobs are envisioned as ripe for automation… especially if one holds the belief that schools are factories whose “product” is a batch of students whose “quality” can be measured using a standardized test. As noted frequently in this blog, the factory school paradigm is the basis for the standardized testing paradigm and contrary to the humanistic approach progressive educators value. If one holds fast to the factory school paradigm, automation is the answer to improved productivity and efficiency. If one believes that each human being possesses unique skills “informating” is the only way to go. Here’s hoping the liberal arts majors prevail over those who write code and prevail over the MBAs and engineers who value efficiency and productivity over human interaction.
A recent Slate essay, “Welcome to Kindergarten. Take This Test… And This One”, describes the testing gauntlet imposed on NOLA students in the name of accountability. Alexandria Neason writes about the experiences of third year Kindergarten teacher Molly Mansel’s challenges in administering computerized tests to her entering kindergartners. The first challenge was teaching them to use a mouse when most of them were used to swiping screens on phones and pads. Then came the test itself:
Mansel’s students started taking tests just three weeks into the 2014–15 school year. They began with a state-required early childhood exam in August, which covered everything from basic math to letter identification. Mansel estimates that it took between four and five weeks for the teachers to test all 58 kindergarten students—and that was with the help of the prekindergarten team. The test requires an adult to sit individually with each student, reading questions and asking them to perform various tasks. The test is 11 pages long and “it’s very time-consuming,” according to Mansel, who is 24 and in her third year of teaching (her first in kindergarten).
The rest of the demanding testing schedule involves repeated administrations of two different school-mandated tests. The first, Measures of Academic Progress, or MAP, is used to measure how students are doing compared with their peers nationally—and to evaluate teachers’ performance. The students take the test in both reading and math three times a year. They have about an hour to complete the test, and slower test takers are pulled from class to finish.
The second test, called Strategic Teaching and Evaluation of Progress, or STEP, is a literacy assessment that measures and ranks children’s progress as they learn letters, words, sentences, and, eventually, how to read. Mansel gives the test individually to students four times throughout the year. It takes several days to administer as Mansel progresses through a series of tasks: asking the students to write their names, to point to uppercase and lowercase versions of letters, and to identify words that rhyme, for example.
These are pre-tests… and over the course of the year Mansel’s students will spend 95 hours taking these tests… and if Ms. Mansel’s performance rating is based on “growth” you can be certain they will spend many more hours in front of screens instead of playing with blocks or engaging in social play with classmates. All of this is being done in the name of maintaing international competitiveness with other countries. But when do other developed countries introduce reading and what does research tell us about this issue? David Elkind’s EducationNext article in 2012 addressed this question:
Evidence attesting to the importance of developmentally appropriate education in the early years comes from cross-cultural studies. Jerome Bruner reports that in French-speaking parts of Switzerland, where reading instruction is begun at the preschool level, a large percentage of children have reading problems. In German-speaking parts of Switzerland, where reading is not taught until age six or seven, there are few reading problems. In Denmark, where reading is taught late, there is almost no illiteracy. Likewise in Russia, where the literacy rate is quite high, reading is not taught until the age of six or seven.
So if research shows that premature instruction in reading increases the probability of reading difficulties, why are we introducing “academics” early? The short answer is that scientific evidence is immaterial in the politicized environment of American schooling today. The consequences on children are adverse whether or not they learn how to read earlier, for the 95+ hours they spend in front of screens are 95+ hours that could have been spent engaged in activities that would help them develop interpersonal skills and self-regulation.
Audrey Watters is always thought provoking, and her brief essay with the dating title “The History of ‘Personalization” and Teaching Machines” that I just came across teased out the question that is the title of this post. In the essay Watters suggests that personalized learning may not be a liberating force that enables students to learn what they want, when they want to. Instead it is a means of feeding students what they need at a pace that enables them to master the skills as defined by those “in privilege and power”.
The distinction is an important one. If “personalized learning” is defined as allowing a student to progress through the Common Core curriculum at a rate of speed that matches their capability to learn, then personalized learning is an efficiency to the factory model, akin to Skinner’s theories. If it is defined as allowing a student to learn what they want, when they want to it is more akin to Ivan Illich’s experiential De-schooling model.
From my perspective schooling should be designed to provide students with the tools— the foundational skills— needed to learn-how-to-learn and to connect them with those who can help them expand their knowledge once they have those foundational skills. The foundational skills can be delivered using a behaviorist approach: math and reading fundamentals, fundamental writing skills, and basic analytic skills are based on hierarchical frameworks that lend themselves to various forms of asynchronous on-line instruction. A teacher in these classrooms would intervene when a student is struggling with a particular concept but would not engage in “broadcast” instruction to a group of students. Otherwise, the teacher would serve as a facilitator, helping the student gain self-awareness and self-understanding through dialogue with the teacher, with peers, and with other mentors.
We have the technological capabilities to make the mastery of foundational skills more efficient and effective than it is today and the human resources available to provide each student with a mentor to help them find answers to the questions that they want to find answers for. “Personalized learning plans” should be able to achieve both ends.
I am slowly but surely shedding boxes of papers from the past and in doing so have reviewed journals I wrote in college, papers I wrote in graduate school, newspaper articles I wrote as superintendent of schools… and lesson plans from my two years of teaching middle school mathematics at Shaw Junior High School from 1970-72. As described in earlier posts, Shaw Junior High was a rough-and-tumble urban school with 3000 students on a split shift the first year I taught there and a 1600+/- school on a single shift the second year. During the first year, I found that the grade-level materials the district provided were inappropriate for my eight grade students, most of whom had not mastered the basic skills. Like most of my first-year colleagues, I encountered many discipline problems— most of which were brought on as a result of the difficulties I faced getting students engaged with the materials.
I was taking a graduate course on “Curriculum” and to complete an assignment for that course AND help me with my classroom management, I decided to write my own material for one of the sections I taught. I used some of the funds allocated to me to mimeograph a 30+ page set of materials that student could go through at their own pace. My wife, who was an artist, illustrated some of the pages with cartoon caricatures of me exhorting the class to “Do Your Math!”. With over 30 kids in the class, implementing this individualized learning was a challenge, especially since the notion of proceeding at their own pace was alien to the students. After a couple of weeks the students got the knack of it and settled into work on the material. The brightest kids in the class completed the packet quickly, but I found I could assign those same students supplementary problems and they worked on them without disrupting the class. Unsurprisingly, the most disruptive students in the class struggled the most with the work, but they were getting my personal attention to help them. I was observed in the class and while the assistant principal noted I was “not following the prescribed curriculum” he acknowledged that the class was orderly and on task… and my classroom management skills had improved.
This experience flashed before me when I read Tina Rosenburg’s Fixes column, “Reaching Students One By One” in yesterday’s NYTimes. The “Fix” Rosenburg describes is “Teach to One” a computer-based individualized program that can deliver exactly what I was attempting to deliver 44 years ago… and with Khan Academy, a wealth of web resources, and all kinds of tracking software teachers in PS 29 in Brooklyn are capable of accomplishing the goal of matching lessons to students far more effectively than I could. Rosenburg concludes her essay with this paragraph:
Critics ask a good question: Why should a school try an expensive, disruptive high-tech platform that’s still unproven? The answer is: in order to prove it. School of One takes comprehensive advantage of technology in ways that let teachers concentrate on teaching. That’s worth getting right. There may be ways to make it cheaper and more effective, but only through further experimentation. As for being disruptive, does anyone defend the current system? “We’re not aspiring to create the least disruptive program,” said Rose. “Our goal is a model that works.”
Taken to its ultimate conclusion programs like “Teach to One” could compel schools to engage in the ultimate disruption: the replacement of age-based grade level cohorts with individualized tracking. Here’s hoping that the standardized testing protocols, with their implicit assumption that all children learn at the same rate, don’t marginalize programs like “Teach to One” that help each and every student experience success.
Why Governor Cuomo’s Education Proposals Are As Bad for Students As They Are for Teachers | The Nation
Third grade teacher Leah Brunski describes how Cuomo’s ideas will play out in the classroom… and it is not a pretty picture. One point Ms. Brunski failed to make in her analysis of various evaluation systems: there are two problems with the evaluation systems used in other countries from the “reformers” perspective: they are designed to HELP teachers and not PUNISH them; and they cost too much money to implement and spending money on education is not the outcome we are looking for— it drives down profits!
The New Yorker provides its readers with a “Daily Comment” feed that provides a column on a timely issue and, it being mid-March, the timely issue is the annual ritual of standardized testing in New York Schools. While I was heartened to see that the New Yorker was covering the emerging grassroots protests in opposition to the testing, I felt that Rebecca Mead’s article, “When a Teacher’s Job Depends on a Child’s Test”, neglected to emphasize the invalidity of the tests that are being used and failed to mention how the Common Core and standardized testing that accompanies the Common Core are being used to undercut the public’s faith in public education and thereby opening the door to privatization. A few phrases from the article will illustrate some of my concerns. To frame the issue of testing and teacher evaluation, Ms. Mead writes:
That teachers should be evaluated is an assertion with which no reasonable person involved with education—from a policy-maker to a parent—is likely to disagree. But how teachers might best be evaluated remains a contested science.
If a reader takes the time to click on the link, they will find that linking teacher evaluations to “growth” as measured by successive standardized tests is NOT “contested” any more than climate change is “contested” for the American Statistical Association has issued a statement indicating that the formulas used by states to evaluate teachers “can’t actually do this with sufficient reliability and validity” and, therefore, should not be used for any personnel decisions. There is no mention of this anywhere in the article, which is an injustice to those of us who are opposed to the emphasis on standardized testing in public schools.
Further along in the article, Mead writes: “Cuomo’s faith in the results of state tests as the best measure of the abilities of both students and teachers is not universally shared.” At this juncture she emphasizes the political debates on testing at the expense of the educational and statistical ones. While she provides a thorough recounting the thoughtful testimony offered in Albany by Mayor de Blasio and Chancellor Farina, she fails to mention the opposition of Superintendents groups, Principals associations, and school boards across the state, all of whom have expressed reservations about Cuomo’s plans based on the lack of educational value of the tests.
Ms. Mead is also remiss in allowing Arne Duncan to have it both ways on the testing issue, writing:
Even Arne Duncan, the Secretary of Education, whose department ties school funding to test results, has warned that “too much testing can rob school buildings of joy, and cause unnecessary stress,” noting that testing should only be one measure of progress. “In too many places, testing itself has become a distraction from the work it is meant to support,” Duncan wrote last fall.
She either doesn’t know that Mr. Duncan is the one who introduced the linkage of student tests to teacher evaluations with his misbegotten Race to the Top initiative or is giving him a free pass when he makes this disingenuous “warning”.
Ms. Mead provides a good overview of the nascent opt-out movement and notes that New York City is moving away from its reliance on tests as a means of determining placement in magnet schools. And, in the end, Ms. Mead DOES view Governor Cuomo faith in testing as a questionable political move:
In the light of such widespread skepticism about over-reliance on test results—and such widespread consensus about the detrimental effects engendered by teaching to the test—the governor’s doubling down on state test results to assess teachers’ effectiveness seems a questionable calculation.
From my perspective, though, in framing the opposition to testing as “widespread skepticism” Ms. Mead overlooks the settled science on value-added measures in the same way legislators in oil-rich states overlook climate change and, in so doing, perpetuates the public’s belief that standardized tests can be used to measure teacher performance.
Finally, and most importantly, Ms. Mead doesn’t challenge the notion that poor performance on tests is the result of poor teaching and bad schools. The tacit acceptance of Cuomo’s assertion that low test scores can be raised by ridding the schools of ineffective teacher effectively reinforces this meme and consequently reinforces the notion that schools can be “fixed” by firing bad teachers and replacing bad government monopoly schools with good free-market schools. More than anything, THAT toxic assumption needs to be challenged, for THAT assumption is diminishing the ability to attract people into public school teaching.