Thursday, August 07, 2014

Notes on Software and Teaching Jobs in Higher Ed After Skimming the Pew report on AI, Robotics, and the Future of Jobs

Pew canvassed 1896 people by sending an " 'opt in' invitation to experts who have been identified by researching those who are widely quoted as technology builders and analysts and those who have made insightful predictions to our previous queries about the future of the Internet." 

52% think technology will lead to new and better jobs overall, as it historically has done, but 48% worry that technology is replacing jobs and not creating enough new ones.
Oremus framed it this way:
Replacing manual labor with machines on farms and in factories was one thing, the worriers say. Those machines were dumb and highly specialized, requiring humans to oversee them at every stage. But the 21st century is witnessing the rise of far smarter machines that can perform tasks previously thought to be immune to automation.

Today’s software can answer your calls, organize your calendar, sell you shoes, recommend your next movie, and target you with advertisements. Tomorrow’s software will diagnose your diseases, write your news stories, and even drive your car. When even high-skill “knowledge workers” are at risk of being replaced by machines, what human jobs will be left? Politics, perhaps—and, of course, entrepreneurship and management. The rich will get richer, in other words, and the rest of us will be left behind.

All of which has brought John Maynard Keynes’ concept of “technological unemployment” back into the academic discourse, some 80 years after he coined the phrase.
Both the Oremus piece, which includes links to other sources, and more importantly the Pew report itself, would do well in a writing course exploring technology and culture, or technology and economics. That nearly 1900 experts are divided, nearly evenly, on the future of jobs, makes this the kind of topic that will not have easy answers.
And remember, this use of technology is getting into education too, with the push for self-pace (lone learner) platforms that seemingly seek to be student proof or teacher proof  by using adaptive tools, automated assessment, interstitial quizzing and other means to funnel students over a prescribed learning path until they come out the other end having satisfied the learning gauntlet. Pearson's Propero is good example of this kind of learning technology; it's being offered to colleges as a way have students take a course and earn college credit, all with no instructor. It's software designed to replace a person with an MA or PhD who would otherwise teach the course.

So it's not just future jobs that are at stake, but current jobs too. There's a balance lost, especially in learning, when machines take over decision making. Software with fixed paths, even if adaptive and adjusting the content to what is not yet been mastered (as measured by what the software can measure), really doesn't teach learners to learn.

Deep learning requires that learners get advice from fellow learners and instructors, reflect on their prior learning choices and next options, and then make decisions about their own learning, good and bad ones, and learn from those decisions, whether good or bad, something about the subject of the course as well how they learn.

But I don't want to say all software is bad. Adaptive learning, automated writing assessment tools, data on engagement, and other information gleaned from learner action and learner choices, can do great things for students and teachers. When designed to support a learning community, defined broadly as students and teachers talking to and learning with and from one another, educational software can provide evidence, show patterns, recommend learning strategies that students can discuss with classmates, teachers, advisers, and academic coaches. Learning, especially learning to read, write, and think critically is hard and sometimes messy work. Software can supply an ordered view into that messy process. Learning comes from understanding what the views mean or tell, and making choices about what is working, what not, and making changes, experimenting, and reassessing over and again.
Good technology -- software, for example, like Eli from MSU, or MARCA from UGA's Calliope Initiative, or MyReviewer from USF, TTU's RaiderWriter, LightSide's project to use AWE as a basis for student/teacher discussion instead of teacher replacement -- can all give teachers and students useful evidence of learning, or of engagement, that becomes the basis for reflection, planning, conferencing, talking to peers, revising, learning from seeing what is working, what is not, and from all that helping the writer and learner to make a choice -- provides a matrix for transformative assessment -- about what to work on next and how to go about it.
And that's what's useful about the Pew report and its divide. It teases out possibilities as well as perils.
Also, as an aside, Pew URL is interesting for a cool social networking feature. As you read the sampling of comments some of the 1896 people made, you'll see embedded in the text Twitter's bird icon. That indicates excerpts from the findings that Pew has tweeted and a link to the tweet where it was used. Kind of cool and rhetorical move worth its own discussion.

Friday, August 01, 2014

Getting Ready to Have no Office

Macmillan Science and Education, where I work, is consolidating offices. New York offices at 33 Irving Place and 41 Madison Ave, along with one or two other locations will move into one building in Manhattan. Our company leaders herald the new office layout -- an open design -- as a campus, where private offices and/or personal cubicles give way to rows of computer terminals, scattered couches and chairs, maybe some tables in nooks and crannies, or small, fish-bowled conference rooms for meetings.

I work in Boston and once leases expire, the plan, we hear, for Boston will be as it is in New York and was in London: a move to a campus-based office model. In anticipation of that move, though it may be years away, I removed from my current office anything personal because campus offices eliminate the personal. I also removed work related materials -- manuscript hard copy, notes from meetings, scholarly books I own that I consult, books we publish, print outs, whiteboards and other workplace desiderata.

So now my office looks like this:

My office as it now stands. Once the whiskey is finished, the glasses will be packed in that white box on the low left shelf and the box will go home with me. The laptop is on a spare chair as a way to create a standing desk. 

Of course, when MHE sets up their campus offices, they'll supply plants, color, furniture, and other amenities, that, like a new library or student union on a college campus, will look contemporary and vibrant, designed no doubt for cross-colleague communication and serendipitous discoveries of new ideas. That's a key pitch for the campus version of the open office. And too as an education publishing company, operating on a campus metaphor resonates with the mission to foster research and learning.

In the campus office, employees share space, much the way that on a college campus the library, learning center or student union each provides open seating for students to use on their visits to those places. Visually, based on glimpses from similar MSE offices in London pictured in a company video, the offices resemble a fairly well appointed college student union or library. Here, take a look:
:
A shared computer area in an open office, where I believe any open terminal can be used by any person who needs it. Much like a computer lab on a college campus.

A conference room, a tradition in traditional offices as well. So nothing new, but often these are now designed so that walls are glass.

Little nooks where colleagues can work in semi-private in small groups or alone. 

Tables placed in places where there's room, borrowing from library design.
If you've been to a college campus in the U.S., you know that the images above all are like the kinds of spaces colleges provide to students and faculty: open, no assigned or named seating, workstations may be present, but laptops abound, students cluster where they will, classrooms, conference room, study rooms, nooks for two or three, supplement the commons areas.

If you've been to college campuses, have walked through areas like these, you'll have seen energy: students in conversation, small study groups at work in a conference room, fingers flying on keyboards, and the immobile work of quiet readers in lone concentration. The campus office hunts for that vibe. Whether it can capture it, I'm not so sure.

For when students and faculty go to those public spaces, they carry in with them only what they can carry out. They do not bring plants, nor photos of family or friends, nor art from their walls, nor mementos, nor cork boards or white boards, nor more books than can fit in a backpack, nor anything else that they might leave out or hang or use or keep in their their faculty offices or dorm room or apartment off campus. They cannot carry those things because there's no where to leave them. Their use of the space is temporary, impermanent, and random.

The campus work spaces that look like the images seen above are not, like a publishing office, used 8, 9, 10, 11, 12 hours a day, five days a week. On a college campus, unlike an office borrowing the design, open work stations, open seating, shared conference rooms and small group nooks are all temporary and interstitial spaces, way stations used on the way from offices/dorms/apartments to classes. Or, if in the library, used for short bursts, relative to where most time is spent.  (Though when deep into research that requires months of long days, the library can feel like home.)

Intermittent work in open, shared and public spaces, differs from working in office with a door one can shut for privacy and concentration, where there's a desk one leave work out on for more efficient resumption each morning,  and where small personal accoutrements -- art, plants, favorite pens, coffee cups or teapots -- make coming in a little more pleasant.  And so it's no surprise that some of the research (summarized here by Julie Beck in The Atlantic Monthly and here by Maria Konnikova in the New Yorker) on open offices shows that productivity falls, workers become more stressed, and the office becomes a place employees are happy to not go to when they can arrange to work elsewhere.

I'm among those who will not do well in an open office, not if I need to spend all my time there. As amenable and contemporary as their design may be, with color, furniture, use of space, they are by design more impersonal, and thus despite their colors and comforts, colder places to work. By design a worker has no permanent place, no private place. I've worked in factories and other jobs where I've had no place, but for maybe a locker to hang and store a coat. And in all those places, workers were easily replaced, some less so than others, but replaceable still. Now publishing isn't factory work. The offices MHE proposes differ mightily from the noise, grime, and grind of machinery I experienced in my factory jobs. 

These new office models -- where a place to sit is make due, a place to talk quietly or privately is make shift, and that make no room for the personal installation of art, photos, awards, degrees, books, papers, plants, toys, and yes, a bottle of whiskey maybe -- reduce the sense of who people are. And by that reduction, reduce the sense that people belong. Right now, when I go through our current offices, even when I colleague's not at her desk, I can glimpse who they are, what their children and grandchildren are into from where photos are taken, the kind of work they do from what stays on their desk, the cultural references they know from their art and knicknacks, how they think and plan from the way they plot out their whiteboards, and so much more. I can know them better as people with lives outside the office and colleagues with skills and talents inside the office.

But in the new office design, if people are not there that day, not in, there will be no trace of them. No sense of them. And in that lack of trace, these new office designs make it feel to me like people who work there are more transient and replaceable than before. Because when folks are out of the office they are as gone from view as passengers in an airport lounge whose flights have departed.

In a review of Cubed: A Secret History of the Workplace by Nikil Saval, Jenny Diski describes vividly an open office failure:
Jay Chiat decided that office politics were a bar to inspirational thinking. He hired Frank Gehry to design his ‘deterritorialised’ agency offices in Venice, California in 1986. ‘Everyone would be given a cellular phone and a laptop computer when they came in. And they would work wherever they wanted.’ Personal items, pictures or plants had to be put in lockers. There were no other private spaces. There were ‘Tilt-A-Whirl domed cars … taken from a defunct amusement park ride, for two people to have private conferences. They became the only place where people could take private phone calls.’ One employee pulled a toy wagon around to keep her stuff together. It rapidly turned into a disaster. People got to work and had no idea where they were to go. There were too many people and not enough chairs. People just stopped going to work. In more formal work situations too, the idea of the individual workstation, an office or a personal desk, began to disappear and designers created fluid spaces where people wandered to settle here and there in specialised spaces. For some reason homelessness was deemed to be the answer to a smooth operation.
So I see these open office plans coming but the offices, as resplendent as it they come to be, and as nice as they might be to visit or attend a meeting there, as soulless and empty. Which is why I've decorated my current office accordingly and only go into it when I have a face to face meeting scheduled with a colleague. Otherwise, I work from home, and when I travel to attend conferences or visit colleges as part of my work, from airports, from hotel rooms, and yes, the campus libraries and student unions my offices back home will soon come to imitate.




Friday, July 25, 2014

What Can Metacogntive Tutoring Research Teach Textbook Publishers?

Leonard Geddes, Associate Dean of Co-Curricular Programs and Director of the Lohr Learning Commons at Lenoir-Rhyne University, began a began a series of LearnWell Project posts 7/23/2014 with a first titled, "A Metacognitive Peer Tutoring Model: Linking Thinking, Learning and Performance in a Peer Tutoring Program."

With a research grant, Geddes asked tutors who he had trained in metacognitive tutoring to record the issues they believed were causing students to struggle in their learning. 40 tutors working with about 80 students in tutorials that lasted an hour logged 522 reports, using this template to record the learning problem:

Problem

Please identify the problem(s) which led the student to seek tutoring. (You can choose more than one option.)

Doesn't grist the material in class
Experiencing difficulty seeing the relationship between what is covered in class and what is reflected on tests
Doesn't know to use the textbook or doesn't use the textbook
Doesn't know how to take notes
Attempts to memorize material only
Student is overwhelmed by the volume of information they are required to learn
Doesn't grasp what the professor is talking about in class.

Which lead to these results overall:

Results of Lenoir-Rhyne Tutors Observations of Student Learning Problems
Results from Leonard Geddes research on learning problems his tutors recorded encountering. Click on image to see more fully.




Now, there's a lot I don't know about this research --how tutors determine which problem is at play, for example. So I hope later posts by Geddes get at what the distinction is from "Doesn't grasp the material in the class" to "Doesn't grasp what the professor is talking about during class."  The first post reports that the categories were chosen from years of tutoring reports and documents, but hearing a bit more about the process for deriving the categories, especially where there seems some overlap -- is poor note-taking a cause of not grasping material? or does not being able to grasp make it hard to take good notes?

Note added 8/1/2014: Geddes second post is in fact getting at some of the questions above. I especially appreciate the views into actual tutoring sessions.

Geddes defines what metacognitive tutoring, writing:
Whereas traditional tutoring focuses on a particularly challenging subject area, and supplemental instruction addresses specific challenging courses, metacognitive tutoring focuses on students’ interaction with content, in general, across domains and academic tasks. We like to call it listening with a “third ear.” Metacognitive tutors address the immediate cognitive problems their students are experiencing while also remaining open to underlying metacognitive conditions that may be contributing to students’ academic problems.

I hope too that there's insight into a tutoring session, perhaps with some record of the discussion to illustrate more fully what metacognitive tutoring is in practice.

But all those and other questions aside, I'm engaged by the results above and like the idea of attempting to map learning problems, to excavate them and address them with students. So tutoring isn't about studying content alone, but studying with each student their own learning process and skills.

My role with Macmillan Higher Education Publishing involves the study of teaching and learning, and I wonder, looking at this, why textbooks score relative low as an issue, yet grasping material in the class is higher. Isn't a textbook a means of delivering course material? If tutors report that students know how to use a textbook, but that they still aren't grasping course materials, is there something textbooks can do more effectively?

When I started at Macmillan, it was with a company called Bedford/St. Martin's*, whose co-founder, Chuck Christensen, said to me in my job interview, that we were not in the textbook business, but rather the pedagogical tools business. And so as textbooks evolve with digital technologies to become more obviously pedagogical tools, where learning analytics, engagement analytics, adaptive learning, personalized learning, and other possibilities emerge, will there be a way to make our course materials such that students can grasp them more fully?

Is it possible to learn from the kind of metacognitive research Geddes and others are doing to build into pedagogical tools metacognitive aides for students?

I think so. Formative assessment that measures not just learning, but also that correlates learning with engagement, with prompts and questions to help students see if they're studying wisely, using their time well, taking good notes, and so on. Creating tools that invite written reflection -- that prompt note-taking while reading, that prompt active study planning (not just delivering links to recommending content after an assessment), that offer a learning journal, or the ability to form study teams. That is, I think it would be a mistake to simply make things that push and pull students, that force them to a path. Instead, we can make things that give students a formative look at where they are, where they need to go to meet course goals, and then choices to follow, suggestions, that students have to choose among.

Without that action -- student agency and choice -- metacognition means so much less.

And, given how important coaching can be in learning, making it so that students can let tutors see into the system, so that tutors can advise them on choices.

__
* In December, Macmillan reorganized its educational companies. Bedford/St. Martin's went from being an independent business, with its own president, marketing department, production, promotions, and other publishing infrastructure to an imprint under Macmillan Higher Education.

Tuesday, July 22, 2014

Goodnight Dull Assignments and Hello Goodnight Moon

Would I assign Margaret Wise Brown's Goodnight Moon in a college writing course?
I think I would, were I teaching at the moment, after reading "What Writers Can Learn From 'Goodnight Moon'" by Aimee Bender in the New York Times (http://opinionator.blogs.nytimes.com/2014/07/19/what-writers-can-learn-from-good-night-moon/).
Bender describes getting several copies of the book birthing twins and settling in to read it for the first time. She then says,
The babies listened in their sleepy baby way, and as the pages turned, I felt a growing excitement — a literary excitement. Not what I expected from this moment. But I was struck and stunned, as I have been before, by a classic sneaking up on me and, in an instant, earning yet again another fan.
It also seemed to me to be an immediately useful writing tool.

“Goodnight Moon” does two things right away: It sets up a world and then it subverts its own rules even as it follows them. It works like a sonata of sorts, but, like a good version of the form, it does not follow a wholly predictable structure. Many children’s books do, particularly for this age, as kids love repetition and the books supply it. They often end as we expect, with a circling back to the start, and a fun twist. This is satisfying but it can be forgettable. Kids — people — also love depth and surprise, and “Goodnight Moon” offers both. Here’s what I think it does that is so radical and illuminating for writers of all kinds, poets and fiction writers and more.
Her piece goes to a reading of Goodnight Moon that explores and celebrates her observation.

I'm attracted to this kind of thing in a writing course because Bender's essay is assignable, and for many students who've read or have had the book read to them, their be the warmth of recollection, and for those who haven't read the book, it's readable.

I can see using the Bender essay and Wise Brown book to discuss form, to invite even the most novice and unsure writers to experiment with voice, to play with style, to think about the idea of simply finding a way to surprise their readers a bit.

Also, the idea makes sense as more and more states require their public colleges and community colleges to eliminate or reduce reliance on developmental reading, writing, and math courses. For example, in Florida, students who graduate high school can go right into a first year writing course even if a writing placement test they take indicates they would be better in a developmental writing course. A few years ago, in fact, they would have had no choice but to take that developmental course first, possible two or even three developmental courses. That's ended and so many professors are seeing in their first year courses writers with more varied ability, some stronger, some weaker. Or in other places, their are accelerated learning programs (see http://alp-deved.org/), progressive approaches to helping developmental writers stay on track and do well in a first year writing course along side students who did not require a developmental course.


To make more varied ability writing courses work, college writing teachers need to learn about differentiated teaching practices of the kind that elementary educators are trained to apply. A lot of college writing teachers already get to differentiated practices, but many will struggle to become comfortable with the approach.

Assigning first Goodnight Moon, discussing it, and then assigning Bender's essay carries with it a native differentiated element. It starts with a simply elegant book to read, a picture book, but one that is being read by writers, following Bender, to learn about writing. So while the book is a children's book, it's also a classic, great literature, and it teaches. The reading can kick off a range of possible discussions -- remembering other books read, or for those who did not grow up as readers or being read to, other stories heard or watched as children.  This would borrow from literacy narrative assignments.

The turn to Bender's essay would, for some weaker readers, be less troubling because they'll have read and discussed a bit the book Bender's writing about. That is, it's a different experience reading a review or analysis of a work of art one's already seen and thought about.  So placing Bender's essay in a digital setting where writers can share notes and comments as collaborating readers, including inviting them to read the comments that will have been archived at the New York Times site, makes reading communal.

And then because of Bender's points about how writing can surprise, float above its forms, subvert its own rules, the first year course can invite writers of all abilities to play with the rules and conventions of standard edited English. Even the seemingly most inept writer, the student who cannot get a subject or verb to agree, a tense to hold, a paragraph to cohere, can play with words and learn to find some joy in the experiments that come with trying to give their readers both "depth and surprise."