Just a quick follow-up to last week’s post on changes in higher education. The New York Times published an article on Friday, highlighting two new books on the future of the American academy and picking up some of the points I discussed last time:

The labor system, for one thing, is clearly unjust. Tenured and tenure-track professors earn most of the money and benefits, but they’re a minority at the top of a pyramid. Nearly two-thirds of all college teachers are non-tenure-track adjuncts like Matt Williams, who told Hacker and Dreifus he had taught a dozen courses at two colleges in the Akron area the previous year, earning the equivalent of about $8.50 an hour by his reckoning. It is foolish that graduate programs are pumping new Ph.D.’s into a world without decent jobs for them.

But the real meat of the article is an overview of some interesting, but slightly terrifying, proposed solutions:

As for the humanities, let professors do research after-hours, on top of much heavier teaching schedules. “In other occupations, when people feel there is something they want to write, they do it on their own time and at their own expense,” the authors declare.

The authors being “Andrew Hacker, a professor emeritus of political science at Queens College, and Claudia C. Dreifus, a journalist (and contributor to the science section of The New York Times)”. You can thank them below.

There’s nothing like an overtly contentious statement to bring in the traffic. And as they go, this is a pretty good one: “Why higher education is like a Ponzi scheme“.

The linked post is actually for a radio program, the content of which was based on this original article by a professor of psychology from the University of Kentucky. In it, she argues that there aren’t enough tenure-track jobs to support the PhD students coming through the system and that students are exploited to prop up the teaching and research of over-stretched professors:

“In short, I think academia shares many of the classic elements of a social trap: It is in most faculty members’ and departments’ best interests to recruit a lot of graduate students. Churning out PhDs is one of the major metrics of departmental ‘success’. Departments need graduate students to teach their classes, and faculty members need them to run their labs. Yet, as in any social trap, when everybody acts in their self-interest, a negative collective outcome ensues.”

Her solution? Not to accept any more PhDs:

“I’m no longer willing to pin my students’ prospects for their futures on an ephemeral job market that shines in the distance like a mirage … I don’t want to be part of the problem any more, and I think I will sleep better knowing that I am no longer contributing to an academic job market that bears an uncomfortable resemblance to a Ponzi scheme on the verge of falling apart.”

I take issue with some of the article being rather loosely written. For example, she states that during the admissions interviews, “I ask them [prospective students] what they want to do with their PhDs. They all reply that they want a tenure-track job at a research university.” (emphasis added). Now this may have something to do with psychology as a discipline, and she does comment that applied jobs in that field are in decline, but anecdotally I would say that a large portion of post-docs, let alone PhDs, are in those positions not to gain an academic post, but to acquire unique skills for an increasingly competitive private sector. To pull a number out of the air, I would say maybe 1 in 5 post-docs goes on to an academic post. And in the UK, this is all part of degree inflation: 1385 full-time doctorates where obtained in 1994/95 and 14165 in 2008/9, a 10.2 times increase! In the same period, the total number of full-time HE degrees obtained (only) rose by a factor of 6.7, from 15601 to 104260 (data from HESA). So in this case, we shouldn’t stress too much about providing everyone with tenured positions.

The corollary to the “all PhDs want academic jobs” thing is that, assuming consistent cohort sizes, it implies we should have the same number of PhDs as post-docs as lecturers as professors, lining up along a smooth career path. Clearly this isn’t the case: academia is inherently hierarchical, always has been and always will be.

But this, I think, is where things get interesting. Let’s make the heroic assumption that PhDs (and post-docs) are cheap labour. Is this such a bad thing, provided that these individuals enter into the arrangement with open eyes? This site is, after all, called academic productivity: if PhDs and post-docs can do teaching and research more cost-effectively than professors and tenured staff, wouldn’t that be better from a societal perspective?

For those of us in the UK, this may not be a hypothetical question. There is currently an “Independent Review of Higher Education Funding and Student Finance” underway, led by Lord Browne (ex-chairman of BP). It has a pretty broad remit to consider how higher education should be funded while ensuring access for talented individuals and one idea I’ve heard suggested is that professors should do less teaching and instead leave it to post-docs or dedicated teaching fellows. From an economic perspective, the appeal is obvious: post-docs earn a fraction of the salary of professors, who of course can better use their time securing grants etc. Indeed, this already happens to some extent with tutorials and labs, but why not lectures as well? Why should a world famous professor be teaching first year calculus?

Well there are a couple reasons. First, many professors enjoy it. Teaching enables them to keep in touch with students and the contact is not one way: students can often ask off-kilter questions that inspire new ideas for research. Second, for new lecturers, teaching is an essential part of professional development. It provides a valuable opportunity to practice presentational skills and, through the planning and preparation of course materials and activities, it can help lecturers to solidify their existing specialist knowledge while also learning new material.

Another rebuttal is that students will demand to be taught by professors. I can’t say the extent to which is true but clearly, higher education is increasingly being seen as a paid-for service with the accompanying “client” expectations. If I go to university X for course Y, it may be because of their Nobel prize winning professor and I may be very cross indeed if that person never sets foot in a classroom. To some extent, I think this is an issue of managing expectations. Universities should offer prospective students courses taught to the highest standards, supported by quality assurance mechanisms like training for teachers and regular feedback from students. But this does not mean that the professors should be doing the teaching per se. There’s no reason to think that a well-trained teaching fellow couldn’t develop and deliver a course as effectively (if not more) than a tenured professor who may excel at research but not teaching.

All of this is to say that calling academia a Ponzi scheme is (obviously) a bit sensational. However, academia is certainly a hierarchical institution and its functioning does depend on the cost-effective labour of PhDs and post-docs. Indeed under current budget constraints in the UK, this labour will become increasingly important and may substantially redefine the delivery of undergraduate higher education. It doesn’t mean that potential PhDs should be discouraged from entering the system but certainly there needs to be upfront clarity about their true role and career prospects.

Do you wonder why people without funding do research? Naw, probably not, because you do it too . Getting grant money involves a huge effort and most people do not have grants. However, everyone tries their best to get time to do research. In fact, universities encourage their faculty to focus on research at the expense of teaching time. This article covers a few theories on why this might happen. For example, Students gravitate toward research orientations, and Research quality has become a proxy for teaching quality. Interesting that two economists wrote it.

News: The Mystery of Faculty Priorities – Inside Higher Ed

How much of the advice we take is based on solid empirical evidence? Surprisingly worrying little! I’d love it if someone actually tries to put together an estimation (let me know if you know one!).

The Chronicle, in a surprising streak of opinion articles, finds that Strunk and White’s claims are mostly baseless:

Simple experiments (which students could perform for themselves using downloaded classic texts from sources like http://gutenberg.org) show that Strunk and White preferred to base their grammar claims on intuition and prejudice rather than established literary usage.

50 Years of Stupid Grammar Advice – ChronicleReview.com

If academics take advice without questioning the evidence, I wonder what will save the general public . Good to see people at The Chronicle debunking BS; I have fallen prey of recommending Strunk and White myself…

"Well, it's round, apple-y and …"

I want to introduce the idea of intended learning outcomes (ILOs) as a template for planning your productivity. Planning is a key part of the Getting Things Done (GTD) system but it’s perhaps an overlooked one. I think part of this problem is that it can be difficult to coordinate plans over the various recommended time horizons: career, 5 years, this year, this week, etc. ILOs help overcome this obstacle by clearly defining what you hope to learn and over what period.

An intended learning outcome is therefore simply a normative statement of your intended learning over a specific period. They are typically used in the design of courses, modules and learning sessions by helping students and teachers understand the scope of a particular lesson and how it fits with other content. For example, in a teaching situation, a set of ILOs might be:

By the end of this lecture, students should be able to:

• list three key features of apples
• compare apples and oranges
• assess whether an unknown fruit is an apple

This simple example shows the main features of a well-written ILO:

• they contain a clear timeframe.
• they should be SMART; that is, specific, measurable, achievable, relevant, and time-bound. Note for example that the first LO states exactly how many apple features students should be able to list.
• they should be phrased using an infinitive verb, chosen to represent an appropriate range of cognitive learning classes. Table 1 of this helpful ILO guide lists the various levels and some sample verbs. For example, to test surface-style knowledge-based learning you may wish to use verbs such as “list” or “recall”. However to test deeper understanding of a concept, higher level verbs like “evaluate” or “argue” can be used. (See here for more on surface versus deep learning styles).
• they should transparently support assessment. That is, as a teacher, you could easily design a test to assess how well you had taught the appley-lesson and similarly students would know exactly what was expected of them.
• Although not explicitly shown here, ILOs should also fit in with other objectives. They may for example be guided by overall course objectives.

Adopting ILOs for personal productivity requires very little modification. From the points above:

• Time frame: You still need to be clear about the time frame of your intended activities. In a later post, I’ll talk a bit about assessing your progress against ILOs but the early tip is to try and align the time frame of your ILOs to existing performance evaluations, such as your annual performance review.
• SMART-ness: They should still be SMART goals.
• Verbs and cognitive learning: This one may require some modification. Chances are you won’t find it very useful to list three new facts in a year’s time. However you may wish to divide your goals into similar easily assessed things, like “Publish two papers” along with more abstract things. Remember though you want to be specific: try and avoid statements like “Understand post-modernist geography” (bit of a losing battle that…) and go for something like “Write a two-page summary of major themes in post-modern geography” instead.
• Assessment: Again we’ll tackle this in a later post but try to keep your annual performance review in mind while writing your ILOs.
• Dove-tailing with other objectives: This is a key one. GTD advocates planning your goals over a series of different time horizons. If you write ILOs or other productivity statements for, say a five-year horizon, you should be able to use these to guide the creation and assessment of your “next-year” learning objectives.

In the next post, I’ll look at how to make the most of intended learning outcomes.

Jose posted an article last week about one person’s PhD experience, highlighting many of the common difficulties encountered when doing what’s largely a self-directed research project. There are loads of books about how to finish a PhD that expand on these questions – of supervisors, organizing your time and so on – but I’ve found that their advice can be frustratingly abstract. When I started my PhD I couldn’t help but wonder “yes but what should I do RIGHT NOW?”.

One useful trick I discovered was to set myself regular assignments. If you’re coming to a PhD from an undergrad or Masters level degree, chances are you’re more used to having teachers give you tasks rather than setting off into uncharted waters on your own. What’s more, you’ve got a big mountain of work sitting in front of you labelled ‘lit review’ and it can be hard to know where to start.

I tried to overcome these problems by dividing up the task into about 8 two-week long assignments. First, I did some brainstorming and together with my supervisor identified the subject areas with which I should become familiar, before dividing these topics into specific research questions. In my case, these were things like “How do people and societies respond to new technologies?” and “Describe the policy issues associated with metering of microgeneration systems in the UK”. I then gave myself two weeks to write an essay on these topics. It gave a clear direction to my reading and by the end of it, I could present my supervisor with a tangible product that we could then discuss.

In theory these mini-reviews could be edited together into your lit review chapter, making one of the most difficult parts of writing up much easier. In practice though I found that, because I was working in a fast-moving field, much of the material I’d gathered in the first months of my PhD was out-of-date by the time I got around to writing up three years later. But the essays only needed some updating, it was excellent writing practice and it was an invaluable way of establishing a routine at the start of a daunting project.

As a footnote if you have some experience in teaching and learning theory, you may recognize this technique as an application of learning outcomes (see this for a brief introduction). My assignment questions can essentially be re-written as:

In two weeks time, I will be able to:

• Describe the policy issues associated with metering of microgeneration systems in the UK
• etc

You can evaluate your own success against these outcomes, recognize how far you’ve come, and be clear about what you’ve achieved for things like your transfer report. The above link has some good tips on how to write and use learning outcomes.

The efficient academic google group has a thread on a really interesting problem. Any hack addressing this has a high chance of saving several hours per week for those of you who teach.

Given lecture material has three components:

1. Slides for digital projection (preferable PDFs rather than PowerPoint or Keynote)
2. Lecture notes to support what I need to say and remember
3. Lecture handout

I regularly update all three, but I am finding keeping all three in sync to be a bit tedious.

I’m not sure what the solution is, but what I am visualising is some sort of single document, where you  write the lecture handout. I could then update this with new information between presenting the lecture.

If you have a solution, drop by and post it there (or here!).

The Chronicle (newspaper) has a good website with a very active forum. I have been monitoring it for a while, and I can certainly say that there are very informative threads in there. It seems to be a very good place to get privileged hard-to-find information about subtle topics such as what is a good job offer, whether a particular department should be red-flagged because of internal fights, or how to negotiate a start-up package. This forum may well be old news for many, but it was an interesting discovery for me so I’ll just post about it just in case it’s useful for anyone.

They have a section on “balancing work and life”. I wonder how many similar forums centered around a profession have one. Scary.

Dan posted in his blog that he had managed to get seven papers out in the open literature in January. I had to interview him.

AP.com: How do you manage your daily workload?

Dan Navarro: A lot more pragmatically than I used to. I put an hour or so aside each morning to cover the miniature administrative rubbish – it’s not really enough time to do it properly, but I’ve started to realise that most of it doesn’t matter very much, so I can cut-and-paste a lot of things (Incidentally: never throw away a good piece of bureaucracy-speak, like a research profile or a course description. You can reuse it about 10 times before anyone starts to care). I tend to do intellectually heavy things throughout the morning and into the early afternoon. I tend to take a bit of a siesta in the late afternoon – I don’t sleep, but I do switch off a bit (sometimes I do paperwork). I find this makes it easier to do something useful in the evening.

AP.com: How do you prioritise?

DN: Mostly by looking at next week’s deadlines, and by thinking about the probable consequences of missing them. Invariably, there’s more deadlines than I can meet, so I start thinking about who I can afford to ignore for a while. What I notice, though, is that I end up clustering everything. For instance, at any given point in time, there’s usually several biggish admin things to do, none of which is actually urgent (note that this isn’t the opinion of the university administrators, but they tend to think that admin is the core business of a university anyway, and are hence untrustworthy). So I tend to ignore all of them, focusing on the core domains of research and teaching, until there’s a big enough admin pile for it to be worth spending a half day on. The clustering approach tends to be useful for me, because I don’t switch gears easily.

AP.com: How do you schedule your time and use calendars?

DN: I once spent a lot of time trying to come up with a complex management system using various different software packages, but found that none of them helped, and a lot of them actually got in my way. So I’ve gone back to a low-tech pencil and paper diary, and I work on the assumption that everything that I need to care about is in the diary. Any meetings not written in the diary don’t exist, as far as I’m concerned. Again, what I try to do is cluster the meetings as tightly as possible to avoid getting trapped (my favourite trick is scheduling a meeting with a student immediately after the nominal end of a committee meeting – so I’m forced to leave when the paperwork says so), and to stay in the same “headspace”. Whenever possible, I try to keep one or (if I’m lucky, two) days a week meeting-free. However, what I’ve learned is that I can’t tell anyone which days I’ve cleared up. Otherwise someone will inevitably try to insert themselves into that slot, since no-one else values my time as much as I do!

AP.com: How do you plan for the future and manage ongoing projects?

DN: In the big picture, very poorly. I seem to be unable to strike the right balance between long-range ideas and short-range projects. Perhaps as a consequence of the hyperbolic function for intertemporal discounting, I keep neglecting long-term projects in favour of short-term payoffs. At a local level, I think I do somewhat better. For me, what seems to be optimal is letting the environment do the work. The directory structure on my computer reflects the various collaborations that I’ve got going on – when one part of the project is completed (usually after a publication) I move it to a chronologically-organised archive. Any follow-up work starts a whole new set of files. Paper-writing files are usually kept separate from the research project files, since writing tends to involve a somewhat different way of thinking. So every time I open up my computer, it shows me the structure of what I’m working on. In terms of tracking the tasks associated with the projects, I let my email account do the work. Spam aside, my email is triaged. There’s the “immediate respond (or ignore) and archive” category, the “it sits in the inbox because I should probably look at it” category and the “this is actually important, but requires effort” category (which sits in the poorly named “current” folder). The net result is that the “current” folder tends to consist only of those emails that correspond to various research tasks and the occasional teaching or admin thing that requires me to think. Since the emails have good metadata – names, titles and most importantly, dates – it tells me everything that I’m supposed to be working on. Better yet, it’s robust. Because when something slips through the cracks, I invariably get a follow-up email. For the most part this works, but in the long run, the thing that I need to get better at is translating a single long-range project into a collection of short-range ones, to avoid neglecting the big picture. For my most recent grants (which tend to describe the long-range stuff), I’ve started doing this, and they’re working out much better than some of the older ones.

AP.com: How is your work influenced by goals?

DN: I really don’t know. I’m not sure if I have any goals that aren’t boring, pragmatic ones (make sure I finish the grant obligations, etc) or vague, useless ones (new stuff interests me). It’s kind of pathetic, but true, and after asking a range of other people about this, they seem to agree that I’m remarkably free of any explicit goals. That said, it’s interesting to note that again the environment can be made to do most of it for you – my “big goals” don’t actually exist anywhere in my head, but they do get bundled into grant applications, so there are some implicit guidelines that I’ve laid out for myself that way.

AP.com: How do you store and remember new information?

DN: Dunno. I don’t use any particular strategy on this one, so it’s all down to the mysteries of semantic memory.

AP.com: How do you write papers?

DN: Depends on who’s in charge. When I’m not leading the project, I’ll tend to follow orders, doing bits and pieces here and there to try to help out. When it’s my baby, almost every paper starts from a collection of figures that I think are interesting (in a particularly egregious example, one of my cogsci submissions this year manages to cram 11 figures and a table into a 6 page paper). Sometimes I’ve got chunks of relevant text I can cut and paste into the paper too, but other than that it tends to come together by writing bits and pieces to go with each image. The one thing I don’t do is try to figure out the big picture immediately. It rarely comes together like that until the very end.

AP.com: How do you deal with procrastination and manage deadlines?

DN: Pragmatics again. Since I always underestimate the amount of time something requires (a packing and unpacking effect, I guess…) there’s always a bit of a late night scramble to finish things by the deadline. However, since that tends to drain me pretty badly, I usually try to reward myself with some guilt-free procrastination for the next few days. Also, unless there’s some clear reward involved (e.g., refereed conference proceedings, grant dollars), I tend not to bother too much about actually meeting the “deadline”. I think only about half of my paper reviews come in on time, and that’s never seemed to stop the action editor (a) complaining but (b) sending more papers for me to review. Most of them I tend to let it slide for a week if need be. Finally, I tend not to worry that much about procrastination. I spend half my weekends reading, or playing around with data, or just plain thinking about research ideas. So what if I decide to play computer games on Monday? Research productivity comes in clumps – some days you’re shit hot, and other days you might as well be at the beach. The big thing for me is making sure that no-one gets in the way of those productive moments, and allowing myself to go have a beer when my brain is MIA.

AP.com: What motivates you to get out of bed in the mornings?

DN: Coffee. My interest in the higher things in life doesn’t kick in until about 9:15…

AP.com: What is the hardest challenge in being productive in the academic world?

DN: In general, I couldn’t say. I suspect there’s some big individual differences here. I know some people who are unproductive because they never bother to write up the work they’ve already finished, others who get paralysed by the inability to narrow a topic down, and others who can’t manage to stay on top of the endless barrage of teaching commitments and pointless administrata. But for my money, it’s striking the right balance between exploration and exploitation. If you don’t spend enough time exploring new territory, it’s easy to end up in a corner, and you don’t find new collaborators. Without new ideas and new people to work with, your research program dies (or worse, becomes boring).  However, if you don’t exploit your strengths, your rate of publication drops, which affects your research income. Without money and status, you can’t grab the best students or attract industry partners etc, and your research program dies. As an example – my advice to grad students looking for postdocs is to try to find one in your “second area”. If you go to your strongest area, you don’t learn enough new material and can’t bring many novel things to the lab. If you go further down than your second strength, you’ll take too long getting up to speed in your new environment to be truly productive. In general though, it’s hard to stay in the sweet spot. I think I tend to make both mistakes at times.

Here is my attempt at a general strategy for managing time. I define productivity operationally here by measuring it in terms of publications (of course, this definition may have critics).

The central point is that your time at work can be divided into productive and unproductive time (see graph), and that both are important; however we should try to maximize the productive time.

The graph may be biased towards the kind of work I do (modeling and experimental cognitive science); other disciplines may not have some of the activities, and the partitioning of your time may well be very different, so feel free to make your own graph with relevant tasks.

Of course there are many more tasks that could be included in this graph in all the branches, so feel free to make your own graph. A third major branch would be tasks that are unrelated to academic productivity, but we have to do anyway (to have a copy of life 1.0).

The delegable tasks are very few, this is bad news. Plus, my experience is that most academics have trouble delegating work and are unhappy with the results. This is why I posted about project management and its importance in the academia.

Note that unproductive time is very needed. You cannot just erase it from your life. In fact, if you do not read papers, do your paperwork, and do some kind of service to the profession, your career will be affected.

So what is a good solution for this conundrum? I propose to do some time boxing: set a time quota for productive time a day (or week) and do the same for unproductive time. The trick to end your day with a feeling of having achieved something is to be sure you have fulfilled your time in the productive area. I’d also recommend to do a large chunk of the productive tasks before entering the gray area of unproductive time (because that way you will not feel guilt). For example you could allocate 2 hrs without interruptions first thing in the morning to your most ‘profitable’ activity (e.g., writing the revision of that paper that needs to be resubmitted!), and then indulge in some unproductive, but pleasant aimless browsing.

One important argument is task dependence: sometimes you just need to have one of the tasks living in the unproductive time area done before you can get a major advance in another productive task (e.g., you need to fill some paperwork to get some funding, or to do to some committee reviewing before your PhD student gets some approval). These tasks usually have hard deadlines and always land you in a difficult situation where you find that their urgency eclipses all other priorities. This is a hard decision to make; I’d try to enforce some timeboxing still, while maybe relocating time in a way that temporarily favors the urgent task that has dependencies; but it should be a temporal thing, and you should go back to your scheduled time boxing that allocates most of your resources to your productive areas.

It would be good to distinguish between things that can decrease your productivity before they increase it, and things that just decrease your productivity.

The first group has things like learning math or programming languages, and testing productivity software. The second group has things like paperwork and service to the profession.

Although making this distinction is purely common sense, it is very useful because (for me at least) it is very easy to spend several hours working hard on something that lives in the unproductive time branch, and then be surprised with how little we have to show for it in the end.