Seth's Blog

This photo illustrates a method I have used for many years to get work done, usually writing. Every six minutes of work, I make a dot or line. One hour = 10 marks = a box (counting method from Exploratory Data Analysis). I use a stopwatch. I make a mark when I am more than halfway to the goal. If I glance at the clock and it says 4 minutes (more than halfway to 6 minutes), I make a mark. If I glance at the clock and it says 10 minutes (more than halfway to 12 minutes from 6 minutes), I make a mark. I only zero the clock when I take a break. I use one piece of paper per day.

I devised this. It is based on an effect discovered by Allen Neuringer and Shin-Ho Chung called quasi-reinforcement. Neuringer and Chung studied pigeons. They found that if you give a pigeon food every 500 times it pecks a key, it will peck the key slowly (say, 2 pecks/minute). If you give the pigeon a brief flash of light every 20 pecks — a marker that shows it is doing the right thing to get food — it will peck much faster (say, 4 pecks/minute). The flashes of light are quasi-reinforcement, said Neuringer and Chung — they have some but not all of the properties of ordinary reinforcement, such as food. By themselves, the flashes of light don’t interest the pigeon. It won’t peck a key to get them. The amazing thing about this effect is that it doubles how hard the pigeon works without raising its salary.

I noticed improvement — it was easier to write — within about 20 minutes the first time I tried this. I chose six minutes as the unit because shorter times were more distracting and longer times less effective.

I told Gary Wolf about the dots method two years ago and he’s been using it ever since. He says it is good for getting started on something he needs to write. After he gets going, he stops doing it. He uses it as an example of the value of self-tracking. I too find that after I get going on something, I need it less. If I stop, however, I drift backwards toward doing less productive stuff or nothing.

Gary asked me about this a month ago and I started doing it again (instead of percentile feedback). I noticed something I had never noticed before, which was that the system lifted my whole energy level and gave me a “can’t wait to get started” feeling in the morning. This too made it easier to get stuff done. It reminded me of some rat research I’d done. Put a rat in a Skinner box and it will explore for a while. If it doesn’t get any food,  after a while (10 minutes?) it will stop exploring and curl up in the middle of the box. However, if I give the rat a pellet of food at random times (at the rate of one pellet/minute), it will keep exploring the box indefinitely. Learning psychologists have emphasized that when you reward an action, you make it more likely. The rat experiment I just described suggests a second effect: when you give reward — at least, when reward is rare — you make all actions more likely. You increase exploration, not just the rewarded response. When I was a young professor I went to a two-week neuroscience program at Dartmouth. It was all lectures. The other attendees were graduate students. I had little in common with them. There was little to do in the town, besides eat Ben & Jerry’s. The next town was 8 miles away. I couldn’t find anything I enjoyed doing. After a week, I had trouble getting out of bed, like the rat curled up in the middle of the Skinner box. A psychiatrist might have said I had major depression. I flew home and was fine.

I have posted several times about using what I call percentile feedback to boost productivity. Percentile feedback means comparing your current performance to your previous performance using a percentile. If the current performance is in the middle of your previous performances, the percentile is 50, for example. Percentile feedback is easy to understand (scores above 50 are better than average) and is sensitive to small improvements — so even small improvements are rewarded. My implementation had three other helpful features: 1. It adjusted for the time I woke up to make different days more comparable. 2. It measured efficiency (time working/time available) to further improve comparability across days. 3. It was graphical. I made a graph of efficiency throughout the current day versus previous days. It greatly increased how much I worked every day.

I love it and wish I had it for everything I measure. Unlike so many feedback systems, it is realistic and encouraging. I found it worked extremely well — to my surprise, actually. It’s not so surprising I would think of it because it vaguely resembles an animal-learning procedure. (Animal learning is my area of expertise within psychology.)

Nick Winter, one of the developers of Skritter (which I use), recently started to use it. He gave a much-too-short QS talk about it in Pittsburgh a month ago. I asked him about his experience. He is as enthusiastic as I am. He wrote:

The percentile feedback has been a huge success–I’m getting way more done than I ever did, and I’m much better at prioritizing toward my main project. Seeing the graph going in real time has been much better at making me aware of what I need to do to hit high targets each day. I will do a full writeup on this, and on my self experiments, when I finish this iOS app and stop focusing so much on work. The short teaser goes something like this:

Phase 0: just tracking normal work at end of day in a Google Doc, average 2 hours a day on iOS development

Phase 1: tracking normal work and iOS dev separately in the Google Doc, average 4 hours a day on iOS development

Phase 2: using Beeminder to have better graphing and goal incentive for iOS dev, average 5 hours a day

Phase 3: first three weeks of using percentile feedback, average 6.4 hours a day

Phase 4: second three weeks of using percentile feedback, deciding to really push it based on the positive feedback from my metrics (more productivity, more happiness), average 9.4 hours a day

So now I’m getting close to averaging 70 hours of focused iOS dev a week and it feels great. In a normal work place, “time spent working” != “productivity”, but for me they’re very similar as long as my energy is good, which it almost always is now.

The surprising insight is that changing the way that I measured my work performance–from spreadsheet, to better spreadsheet, to graph, to better graph–has had such a huge impact. I have been working on maximizing work productivity for four years, ever since starting the startup, but in the last six months I’ve become radically more effective. I love the percentile feedback graph design!

You can see his implementation on his homepage.

I recently posted about how doctors act like predators, in the sense of having what Jane Jacobs called “guardian values” (e.g., loyalty to other doctors is more important than honesty to patients). Here is an example of medical behavior that coming from an ordinary business would be shocking:

On February 21 [2012], I had my evaluation for a kidney transplant at a university-affiliated medical center about 100 miles from where I live. The way this institution operates, it takes about 8 months to get from initial referral to evaluation and there are all kinds of diagnostic tests in between (see previous blogs for more details). Once you are an approved transplant candidate and an organ becomes available, you go to the hospital and have surgery. The average stay for a kidney transplant is about 3 days and then you are discharged to a local hotel for 5-7 days. During that time, you return to the hospital every day for blood work, monitoring of the immunosuppressive medications and patient education. Also, you must have a full-time caregiver. That can be a friend, family member, stranger off the street corner, but they must be with you at all times to ensure that you are eating, taking meds, bathing, etc. Also, driving is prohibited until about six weeks post-transplant so the caregiver is also a chauffeur and attends the educational activities as a back-up in case the patient becomes incapacitated or symptoms of rejection appear.

In short, your caregiver must be able to put their own life on hold for about two weeks with as little as two hours notice. When you think about it, that’s a pretty tall order to fill. I have a caregiver, he happens to be a member of this forum. He is a dear, dear friend and always will be if only for the fact that he is willing to undertake this role with only the merest of acquaintance. He is more than willing to put himself and his home at my disposal if necessary. I won’t call him out by name, he obviously knows of whom I speak, but I truly feel as though Karma has smiled on me since our paths have crossed.

So the evaluation finally rolls around. Caregivers must be present during the evaluation. We check in at the medical center and are shown to an exam room. We are seen by a barrage of clinicians; dietician, nephrology resident, nephrology attending (the doctor in overall charge of my medical care while at the transplant unit), and the transplant surgeon. There are physical exams (kind of interesting since my caregiver knows me pretty well, but not THAT well), an EKG and a side trip to the lab. At the lab, the phlebotomist doesn’t pay any attention to my advice about using a butterfly catheter and proceeds to draw 20 (count ‘em, 20) vials of blood for type, cross match, antigen levels, etc, etc through a Vaccutainer. About halfway through, my vein collapses and she has to switch to the other arm, this time with a butterfly. After that, a chest x-ray. Back up to the 9th floor for our final meeting of the day; the social worker.

Up until this time, everything had been encouraging. I can’t say enough good things about the clinical staff, they were all wonderful, professional, warm, willing to answer questions, etc. My transplant surgeon looks like he should be on a TV medical drama, he can unzip me any time! The good vibes ended the minute we sat down with the social worker. She informed me that I would be required to have a second caregiver, a backup so to speak. WTH? People that can call a halt to their lives don’t grow on trees. Talk about hitting a brick wall. Here’s a sample of the conversation:

Social worker: What will you do if you are discharged to home and you can’t take care of yourself?
LadyDoc: Well, if I can’t take care of myself then I guess I shouldn’t be discharged, should I?
Social worker: Well, you could always go into a nursing home.
LadyDoc: Over my dead body.

And there you have it, the standoff. I have looked through every single printed word and email that I have ever gotten from this institution (and I keep very good records) and there is NOT A SINGLE WORD about having a second caregiver. The only family I have in the area is my daughter and she has two little boys under the age of five at home, so I can hardly ask her. My circle of friends is painfully small, many are disabled and not up to the challenge and the others have lives of their own.

The social worker called me a few days later to see if I had changed my mind and it suddenly began to sound like a sales pitch. She was touting all the advantages of this particular institution but I just don’t see it. I am now turning my attention to medical centers where the inpatient stay is closer to 5-7 days and then the patient is discharge directly to home, none of this stay-in-a-hotel stuff. I can’t think of too many places where germs and nastiness run more rampant than a hotel. I am so frustrated, I feel as though the last 7 months of my life have been an utter waste of time. Furthermore, the evaluation day was wasted; if we had met with her first we could have simply gotten up and walked out and said “Thank you for playing, please try again”.

In case you needed any convincing that customers for health care differ from customers for other services. (The difference: they are more desperate.) Think of this example if you are sure that government-run health care must be worse than the current system. You can learn what happened next at the link.

In March I discovered that looking at a graph of my productivity (for the current day, with a percentile attached) was a big help. My “efficiency” — the time spent working that day divided by the time available to work — jumped as soon as the new feedback started (as this graph shows). The percentile score, which I can get at any moment during the day, indicates how my current efficiency score ranks according to scores from previous days within one hour of the same time. For example, a score of 50 at 1 p.m. means that half of the previous days’ scores from noon to 2 p.m. were better, half worse. The time available to work starts when I get up. For example, if I got up at 4 a.m., at 6 a.m. there were 2 hours available to work. The measurement period usually stops at dinner time or in the early evening.

This graph shows the results so far. It shows efficiency scores at the end of each day. (Now and then I take a day off.) One interesting fact is I’ve kept doing it. The data collection isn’t automated; I shift to R to collect it, typing “work.start” or “work.stop” or “work.switch” when I start, stop, or switch tasks. This is the third or fourth time I’ve tried some sort of work tracking system and the first time I have persisted this long. Another interesting fact is the slow improvement, shown by the positive slopes of the fitted lines. Apparently I am slowly developing better work habits.

The behavioral engineering is more complicated than you might think. My daily activities naturally divide into three categories: 1. things I want to do but have to push myself to do. This helps with that, obviously. 2. things I don’t want to do a lot of but have to push myself away from (e.g., web surfing). 3. things I want to do and have no trouble doing. But the recording system is binary. What do I do with activities in the third category? Eventually I decided to put the short-duration examples (e.g., standing on one foot, lasts 10 minutes) in the first category (counts as work), keeping the long-duration examples (e.g., walking, might last one hour) in the second category (doesn’t count as work).

Before I started this I thought of a dozen reasons why it wouldn’t work, but it has. In line with my belief that it is better to do than to think.

I fixed a few problems and eliminated the one Windows-specific function so it can be used with Macs.

The new version is here.

I have put a requested R workspace on my website so that you can download it. The percentile feedback workspace compares your productivity (time spent working/time available to work) today to previous days. When I started using it, I became more productive. Here is an introduction. Here are all posts about it.

This is not for everyone. You need R installed to use it (of course) and you’ll need to know at least a little R. You must edit a function called save.ws so that the workspace is saved in the right place. I have used it under Windows XP.

Warning: This post, written for the Quantified Self blog, has more repetition than usual of material in earlier posts.

In January, after talking with Matthew Cornell, I decided to measure my work habits. I typically work for a while (10-100 minutes), take a break (10-100 minutes), resume work, take another break, and so on. The breaks had many functions: lunch, dinner, walk, exercise, nap. I wanted to do experiments related to quasi-reinforcement.

I wrote R programs to record when I worked.  They provided simple feedback, including how much I had worked that day (e.g., “121 minutes worked so far”) and how long the current bout of work had lasted (e.g., “20 minutes of email” — meaning the current bout of work, which was answering email , had so far lasted 20 minutes).

I collected data for two months before I wrote programs to graph the data. The first display I made (example above) showed efficiency (time spent working/time available to work) as a function of time of day. Available time started when I woke up. If I woke up at 5 am, and by 10 am had worked 3 hours, the efficiency at 10 am would be 60%. The display showed the current day as a line and previous days as points. During the day the line got longer and longer.

The blue and red points are from before the display started; the green and black points are from after the display started. The red and black points are the final points of their days — they sum up the days. A week or so after I made the display I added the big number in the upper-right corner (in the example, 65). It gives the percentile of the current efficiency compared to all the efficiency measurements within one hour of the time of day (e.g., if it is 2 p.m., the current efficiency is compared to efficiency measurements between 1 p.m. and 3 p.m. on previous days).

I started looking at the progress display often. To my great surprise, it helped a lot. It made me more efficient. You can see this in the example above because most of the green points (after the display started) are above most of the blue points (before the display). You can also see the improvement in the graph below, which shows the final efficiency of each day.

My efficiency jumped up when the display started.

Why did the display help? I call it percentile feedback because that name sums up a big reason I think it helped. The number in the corner makes the percentile explicit but simply seeing where the end of the line falls relative to the points gives an indication of the percentile. I think the graphical display helped for four reasons:

1. All improvement rewarded, no matter how small or from what level. Whenever I worked, the line went up and the percentile score improved. Many feedback schemes reward only a small range of changes of behavior. For example, suppose the feedback scheme is A+, A, A-, etc. If you go from low B- to high B-, your grade won’t change. A score of 100 was nearly impossible, so there was almost always room for improvement.

2. Overall performance judged. I could compare my percentile score to my score earlier in the day (e.g., 1 pm versus 10 am) but the score itself was a comparison to all previous days, in the sense that a score above 50 meant I was doing better than average. Thus there were two sources of reward: (a) doing better than a few hours ago and (b) doing better than previous days.

3. Attractive. I liked looking at the graphs, partly due to graphic design.

4.  Likeable. You pay more attention to someone you like than someone you don’t like. The displays were curiously likable. They usually praised me, in the sense that the percentile score was usually well above 50. Except early in morning, they were calm, in the sense that they did not change quickly. If the score was 80 and I took a 2-hour break, the score might go down to 70 — still good. And, as I said earlier, every improvement was noticed and rewarded — and every non-improvement was also gently noted. It was as if the display cared.

Now that I’ve seen how helpful and pleasant feedback can be, I miss similar feedback in other areas of life. When I’m walking/running on my treadmill, I want percentile feedback comparing this workout to previous ones. When I’m studying Chinese, I want some sort of gentle comparison to the past.

Here is another example of the efficiency graphs I’ve blogged about (here, here and here). The line is the current day; it shows how well I’m doing compared to previous days. It goes up when I work, down during breaks. The number in the right corner (“77″) is the percentile of my current efficiency (at the time the graph is made) compared to measurements within one hour (e.g., a measurement at 2 pm is compared to previous measurements between 1 pm and 3 pm).

The blue points come from before I started the feedback; the green points, afterwards. The red and black points are the final points of a day (that is, at quitting time). That the green points are above the blue points suggests that the graphical feedback helped. Here is a better way of seeing the effect of the feedback.

I didn’t expect this, as I’ve said. It is not “the effect of feedback”; before the graphical feedback, I’d gotten non-graphical feedback. It is a comparison of two kinds of feedback.

Why was the new feedback better? Here’s my best guess. It helped a little that it was pretty (compared to text). It helped a lot that it was in percentile form (today’s score compared to previous scores). This meant the score was almost never bad (from the beginning the percentile was was usually more than 50) and yet could always be detectably improved (e.g., from 68 to 70) with a little effort. I wish I could get such continuous percentile feedback in other areas of life – e.g., wwhile treadmill running. I think feedback works poorly when it is discouraging or unpleasant and when it is too hard to improve. When I taught a freshman seminar at Berkeley, I got feedback (designed by a psychology professor) that was so unpleasant I stopped teaching freshman seminars. Because it came only at the end of the term, it was hard to improve — you’d have to teach the class again to get a better score. Moreover, it compared your score to everyone else’s.  I think I was in the lower 50%, which I found really unpleasant. There was no easy way to give feedback about the feedback; maybe it is still in use.

In contrast, I love the feedback shown in the upper graph. Not only does it really help, as the lower graph shows, it leaves me at the end of the day with a feeling of accomplishment.

A few days ago I posted about the effect of efficiency graphs — graphs of time spent working/available time vs time of day  (see below for an example). I used these graphs as feedback. They made it easy to see how my current efficiency compared to past days. As soon as I started looking at them (many times/day), my efficiency increased from about 25% to about 40%. I was surprised, you could even say shocked.  Sure, I wanted to be more efficient but I had collected the data to test a quite different idea. In this post I will speculate about why the efficiency graphs helped. (more…)

After talking to Matthew Cornell a few months ago, I decided to try to measure how much time I worked.  Measuring it might help me control it. I’d done this before but hadn’t gotten anywhere. Maybe this time . . . (more…)