Brian Spiering's Musings

Peer Reviewing Gamification

2012-02-10T06:47:00.000-08:00

Peer reviewing in scientific journals is a black box. There are known inputs and outputs but the process itself is mostly unknown. As a result, there are few incentives built into the current system to make peer reviewering both quick and high-quality.

Peer-reviewing could be improved by gamification, explicitly defining the rules and what winning looks like. Reviewers/players could be tracked on a public or private scoreboard. Reviewers would earn points for completing their reviews early. They move up to more prestigious journals/leagues by the quality of their reviews. Submitters or editors could then pick reviewers based on objective metrics of their work. It would also provide another data point for merit-based academic promotions.

What I Use: The Two Minute Rule

2012-02-08T06:53:00.000-08:00

The Two Minute Rule from Getting Things Done revolutionized how I ~~check~~ process email. I complete every action, that takes less than two minutes, associated with the current email before moving onto the next email. The most common actions are adding events in my calendar, tracking specific next actions, and replying with canned responses, via TextExpander.

Keyboard shortcuts are the engine that drives that machine. I avoid switching costs by not touching the mouse. As my muscle memory develops, I can complete more actions into that same block of time.

The small stuff is off my plate, and my mind, before I enjoy the entrée of the day.

The new face of scientific publishing

2012-02-06T04:50:00.000-08:00

I am part of the scientific journal publishing machine.

I have written published papers. Currently, I am peer-reviewing journal articles. I am on the cusp of submitting another paper for review. I truly enjoy that aspect of my science work. However, there is unnecessary friction in the current system, if the goal is to have the best science in the world as quickly as possible for people to read and use.

The paywall issue is a canary in the coal mine for this industry. The journal publishing industry is predicated on a false sense of scarcity. One example is the notion of finite space. This is true for printed journals. That limit no longer exists since journals are now primarily electronic.

I am not ready to completely jump off the sinking ship of conventional journal publishing, but I am willing to work for a system better suited to the digital age.

What counts as evidence in your world?

2012-02-03T04:44:00.000-08:00

Is it a story, an individual data point, a single data set, or a meta-analysis?

The hard work of science

2012-02-01T05:42:00.000-08:00

A theory can be proved by experiment;

but no path leads from experiment to the birth of a theory.

- Manfred Eigen

Proposing new theories is the hard work of science. It requires insights and guts. The theory you propose will be proved wrong, since all theories are wrong. It is far easier to run another experiment. Experiments are usually a series of technical problems (e.g., How do we find subjects? How do we analysis the data?). It feels safe to correctly solve technical problems. It is harder to create the time and space to develop a new theory.

The importance of trivia

2012-01-30T04:36:00.000-08:00

Given the almost constant access to the facts of the world, what is the function of knowing trivia?

One element of creativity is combining two or more different ideas (e.g., someone was the first person to put peanut butter and chocolate together for the eternal betterment of the world). The free recall of facts and figures is grist for that creative mill.

Trivia can be the start but is never the finish.

How much data is enough?

2012-01-27T08:03:00.001-08:00

How do you know when you have collected enough data for a project?

Do you collect data until ...

you pass a criteria set before the start?

you pass an arbitrary statistical significance test?

you find something interesting?

you find something meaningful?

you run out of resources (i.e., money, time, or subject pool)?

you satisfy every reviewer?

you get published?

What I Use: Wolfram|Alpha Computational Knowledge Engine

2012-01-25T05:26:00.000-08:00

Wolfram|Alpha Computational Knowledge Engine is a sleeping giant. It is a different service than the typical search engine by organizing the knowledge it provides. That basic synthesis helps make sense of the Internet's informational fire-hose tendency.

Wolfram|Alpha is an intergal part of my workflow as a trusted source for specific knowledge. That knowledge is more important to my work than just facts. One specific example to Matlab programming is finding the precise RGB values by simply typing in the common name. It also provides fact-based context, either to generate ideas or frame conversations. It is a great service that is getting better. I have seen the consistent and continuous improvements in the both breath and depth of the service during the time I have used it.

A good place to start is the examples page.

Tufte makes curry

2012-01-23T05:20:00.000-08:00

Yesterday at a big box supermarket, I stumbled on this:

These two coconut milks appear to be nearly identical products. Given they were representing themselves as commodities, I was solely making my choice based on price. However, the store didn't make my choice simple or easy.

The can on the left retails for $1.39. The can on the right retails for $3.00 for 5. Furthermore, the prices are listed in different font sizes. Why isn't the price listed for an individual can, even though the cans differ in size by 0.5 oz? I was drawn to the red box, which reads "$1.65 per pound" for the left one and "the price is "$3.95 per quart" for the right one! The store chooses different metrics for the same goods, thus making a meaningful comparison very difficult. Imagine one gas station selling gas by the gallon and another station selling it by the pound. Because those stickers are unnecessarily different on multiple levels, the don't serve their function as more objective sources of information.

There is both a science and art in creating meaningful information. A foundation in basic science is a good place to start.

Real Problems

2012-01-20T04:46:00.000-08:00

Problems worth solving are not always the same as:

• the easiest to solve problem

• the most apparent problem

• the first problem

• the latest problem

• the loudest problem

• the most popular problem

• other people's problems

The Neurobiology of Enchantment

2012-01-16T04:43:00.000-08:00

One key tenant of enchantment is applying the power of unexpected positive surprises. Everyone likes unexpected positive surprises. There is a biological reason for it, like all human properties.

Reward Prediction Error (RPE) theory predicts that unexpected positive surprises are more rewarding than expected positive events. Think about wining the lottery vs. last week's paycheck, even if they are same monetary value. The underlying mechanism of this effect is the unique biology of dopamine. Dopamine is the stuff of addiction (e.g., casinos and the *caines). That addiction can be positive or negative.

Dopamine and enchantment are powerful stuff. Use them wisely.

By the way, Enchantment is Guy Kawakisi 's latest book. Check it out. I did from my local library.

What I Use: Keyboard Shortcuts

2012-01-11T16:41:00.000-08:00

It should not come as huge surprise I'm addicted to keyboard shortcuts, given my love of TextMate* and my fully-loaded 17'' MacBook Pro. Together we take on the world. Keyboard shortcuts let me fly through the busy work and allow me to focus on creating great stuff. You can find THE list of mac keyboard shortcuts here and a good textmate keyboard shortcut cheat sheet here.

Those are just the raw tools. Most people try to start by memorizing the first one that catches their eye. They try to remember, by sheer force of will, it and hope to recall it during their usual workflow. It could sit unused in their "psyche buffer" for 1 second or 1 day. That is the hard and inefficient path towards becoming a knowledge worker ninja. Additionally, it is easy to become overwhelmed with almost never ending list of keyboard shortcuts.

I have an experimental cognitive psychology twist on learning keyboard shortcuts.

It is the "glue game." Every time you sit down at your computer, your hands are "glued" to the keyboard. You lose the current round in the game when you touch the mouse (the trackpad counts, too). Your score is how much work you can do without touching the mouse. After you lose the current round by touching the mouse, you look up the specific keyboard shortcut for that action. You back up a couple of actions steps and use the shortcut. The next round of the game starts after you get up and sit back down. Each round should last a little longer and you should learn a single specific meaningful shortcut.

It is the easiest and quickest way to learn the actual keyboard shortcuts you use. It is called chaining in the business.

One of my favorite chains is:

⌘tab, repeat until textmate is selected

⌘N

⌘S

Five tabs

Up and down arrows into Dropbox

Return

That chain puts me in textmate, opens a new file, and saves to it to my dropbox folder, which lives in my sidebar. I can capture and, more importantly, never lose every idea, brilliant or not, I have. I started by learning ⌘tab link and then I added the ⌘N link. I added one link at a time until I created that complete chain.

Leave a comment with your single favorite keyboard shortcut chain.

*Do not comment or ask me about textmate 2!

Science by proxy

2012-01-09T05:49:00.000-08:00

I value primary data and theories over science by proxy. Science by proxy (e.g., popular press articles, podcasts, and blogs) are the warmed-over rehashings of primary science. Frequently, it reposting without analysis. Primary data and theories are difficult. They require domain knowledge and a deep reading to gain understanding. Science by proxy on the other hand is addictively easy to produce and consume.

Science by proxy points to actual science but is not actual science. In a similar way, google is links to the web (not the actual web), wikipedia is facts about the world (not the actual world), and facebook is information about your friends (not your actual friends)*.

* I read that keen observation somewhere but am unable to cite it.

n of 1 always beats n of 0

2012-01-06T14:35:00.000-08:00

An anecdote can be a first data point.

And data trumps speculation.

A Cheat Sheet for MATLAB

2012-01-05T06:25:00.000-08:00

I love MATLAB.* It is a seamless workflow environment from experimental design to manuscript-ready figures, including data collection, high-level analysis, and computational modeling. As a result, my work is faster, easier, and higher quality by remaining within a single computer application.

However, MATLAB has a steep learning curve. I have been using it for so long I forget how difficult it is to initially learn. MATLAB is literally a foreign language with its own vocabulary, syntax, and thought patterns. Over the years, I have developed a method to teach Matlab with minimal time and pain. The complete method is beyond the scope of a blog post, but I can share a quick tactic. After the initial ramp-up, I hand my mentee "MATLAB for the Faint Hearted". It is the one-stop shop cheat sheet for MATLAB, my mentee no longer has to consult the multi-volume(!) Matlab manual to answer a simple question.

I wish I could give credit to the originator of "MATLAB for the Faint Hearted," but the exact origins of the document are lost in the mists of history. I have updated and expanded it over the years. Now I have decided to share it:

"MATLAB for the Faint Hearted."

PDF version

Quick Tip - If you ever get really frustrated with Matlab, type "Why" in the Command Window!

* Do you have to shout it every time you say it?

Looking past the fringe

2012-01-03T08:47:00.000-08:00

The fringe of science gets the most attention. The new and weird of science is overrepresented in most media. The central tenants, or core, of science is not "news," by definition. However, a stronger understanding of them in a chosen field will provide a better long-term return on investment for your limited time and attention.

Gift giving in the information age

2011-12-30T05:53:00.000-08:00

Given the world as it is right now, how do you give a holiday gift?

I realized over the course of this season the best gifts are unexpected, should reinforce human connections, and can even call to a higher purpose.

Most people reading this post have enough money to buy all the average physical gifts they desire. There is very little physical scarcity in today's world. If it exists and I can connect to the internet, it can be at my home in a couple of days.

If I give a gift that is expected, be it a trending gift of the moment or a variation on a tradition, there is very little delight. It is better if I try give a gift from my heart, from one human to another human. To make a gift truly special, I try to call to a higher purpose. That higher purpose could be helping the people the gift receiver cares about or an improved version of the gift receiver themselves.

The art of this gift giving style is calling to a higher purpose without preaching.

Unquestionable Beliefs

2011-12-28T08:46:00.001-08:00

Everyone has unquestionable beliefs (Hint - They probably wouldn't call them beliefs.). If a belief is unquestionable, it is a waste of everyone's time to discuss it. Either it is two people reinforcing their common world view or two people talking past each other. The more interesting space is where beliefs are allowed to be questioned.

Data, Method, or Narrative

2011-12-19T08:20:00.000-08:00

What do you spend the most time and attention on?

What do you read first?

What do you read most often?

What has the most value in science?

A map of the brain: Allan Jones on TED.com

2011-12-07T17:21:00.000-08:00

Allan Jones shows a powerful direction for the future of neuroscience. He combines the best of science (i.e, technical skills and cutting-edge technology) with the latent power of crowd sourcing. You can explore those human brains here.

Slip Slidin' Away in Science

2011-12-02T12:00:00.000-08:00

“Life is like skiing.
Just like skiing, the goal is not to get to the bottom of the hill. It’s to have a bunch of good runs before the sun sets.”
Seth Godin

Here is how my work in science is similar to skiing:

Some runs are easy; some runs are hard. A classic day has the proper mixture of both. Every paper won't get published in Science or Psychology Review. Some data sets will end up in the file drawer.

The best equipment isn't necessary but can help bring out the joy. The synthetic clothes and properly-tuned skis of science are access to data and fast computers. Those elements are more available and cheaper everyday.

The solitude of quiet runs and writing are intrinsic elements. If they are embraced, they can be delightful. However, ski buddies and science colleagues are more likely to push me in interesting directions, often towards personal blind spots.

Sometimes the logistics of skiing (e.g., traffic and snowy mountain roads) keeps me from even getting to the mountain. That is a shame. Sometimes logistics get in the way of my science goals (e.g., publish quality data, propose new models, and mentor as many people as possible). I need to be more mindful of minimizing everything that does not help me to achieve my goals.

I wonder how many good science runs can I get in before my sun sets.

Free Data

2011-12-01T05:45:00.000-08:00

In the past, there has been a scarcity of data. Data was expensive to collect and analysis. It is cheaper every day. What happens when it is free? Will that decimate how you do science? Or will you use it to ask and answer more interesting questions?

A friendly reminder about the fringe of science

2011-11-28T12:48:00.000-08:00

The fringe of science still has to play by the rules (e.g., falsification) to get the honor of being called science.

Leveraging the long tail of posters

2011-11-25T05:03:00.000-08:00

Chris Anderson points out in the The Long Tail an aggregator is necessary to manage increased options, for example the Sears catalog aggregated consumer goods and Google aggregates web sites. Currently, there are no scientific poster aggregators to leverage the long tail of posters. That is an opportunity for a risk-taking conference to build a community.

Bonus - Requiring posters to be a vector image would facilitate digital searching.

The center, frontier, and fringe of science

2011-11-23T07:49:00.000-08:00

What follows are my personal, working definitions of the center, frontier, and fringe of science*.

The center of science are the concepts and principles taken for granted in a particular field. This is the material taught in the classroom. An neuroscience example is neurons exist and have certain properties. The frontier of science are novel conjunctions and hypotheses within established paradigms. This is the day-to-day work of science. Reward prediction error is an neuroscience example. The fringe of science is outside of current scientific paradigms. This is can be found at post-conference happy hours. The idea that neurons are able to predict 500ms in the future what will be required of them is fringe neuroscience.

Each of those elements has an inherent medium. Textbooks are most often contain the center of science. They take massive amounts of resources and time to produce. It makes sense to have them reflect the central tenets, which are unlikely to change. The bulk of journals, posters, and talks are at the frontiers of science. Those media take less resources and time to produce and can afford to take more chances. The best medium for the fringe of science is blogs. The cost of blogging is effectively free. It provides the opportunity to try ideas and receive immediate feedback. Blogs can serve a vital function in the world of science.

* Excuse me if I'm "reinventing the wheel." I could not find a digital or hard copy of Notes on the Nature of Fringe Science. I would greatly appreciate if someone could provide or point me to a copy.

Plenty of work to do

2011-11-21T05:18:00.000-08:00

One of my missions at Neuroscience 2011 was finding interesting single-unit recording studies, there was no shortage of high-quality studies. One common, but not universal, phenomenon was a dearth of coherent theories for the work being done. It appeared to me history and technology were the greatest driving forces behind recording sites and task selection choices. While each individual study is well-executed, there could be much greater value with coordinated effort between theorists and empiricists. Theorists can elucidate the most important questions, in contrast to the empiricist's intuitive focus on the most convenient questions.

There is still plenty of work to do for a theoretical neuroscientist who likes to get his/her "hands dirty."

The long tail of posters

2011-11-18T07:43:00.000-08:00

An example of a power law graph showing popularity ranking.
To the right is the long tail; to the left are the few that dominate.
Notice that the areas of both regions match.
Picture by Hay Kranen

Not so long ago, posters were one time only events. All the time and energy that went into the construction of a poster would only have a brief flicker of life during a conference. Shortly after the conference, the poster would either be hung in a research hallway or tossed into a poster graveyard found in the corner of every laboratory.

Posters now can live a second, longer digital life. Scientists have the tools to easily host and post their posters. You can find mine here.

Most posters will have their biggest impact during the physical conference, the head of "the long tail", but the digital afterlife provides posters with a continuing opportunity to be social objects.

Keep Moving and Get Out of the Way

2011-11-16T04:53:00.000-08:00

"Keep moving and get out of the way" is my mantra for collaboration.

I always try to move a project forward, sometimes inch-by-inch.

At the same time, I get out out of the way. I try not to be the bottleneck in collaborations.

It is Ape Law that works in science.

Poster Success!

2011-11-14T07:32:00.001-08:00

A big thank you to everyone who took time to stop by my poster.
I enjoyed engaging with so many people.

Bonus - PDF version of my poster

Posters are social objects

2011-11-13T07:00:00.000-08:00

Posters are one of the primary social objects of science.

It is easy to start people talking with posters. Posters are both the pretext and context for a conversation. They are a conversation script with their predefined beginning, middle, and end. The poster medium is inherently one-on-one or small group. People are comfortable communicating in that context, in contrast to the contrivance of "giving a speech."

All that and a good acronym

2011-11-12T11:55:00.000-08:00

Walking around Neuroscience 2011, I wonder how science ideas spread. An idea needs to first catch my attention before I start spreading it. Here is what I look for:

Data that is unexpected and unexplainable
Data that helps choose between two competing theories
A theory that explains previously unaccounted for data
A theory that synthesizes multiple, separate data sets
A method that generates those types of data and theories

A good acronym doesn't hurt.

My Neuroscience 2011 Poster Tomorrow (11/13)

2011-11-12T05:41:00.000-08:00

My poster "Expectancy violation and functional connectivity in musical syntax processing" is tomorrow.

Date and Time - Sunday, November 13, 2011, 8:00am - 12:00pm

Board - JJ7

I will be there the entire time.

Neuroscience 2011 Zeitgeist

2011-11-11T06:27:00.000-08:00

Every year the Society for Neuroscience conference has an unofficial theme. Looking at the upcoming events, neuroeconomics is my guess for Neuroscience 2011's theme.

Classical economics makes assumptions that are frequently untested. This creates an environment for unbridled speculation. Adding a neuroscience perspective to economics provides constraints that can create the foundation for meaningful conversation and science.

Stop me if you heard this one before ...

2011-11-09T04:58:00.000-08:00

A theoretical physicist and a mathematical psychologist walk into a bar. Sorry, I don't have an actual joke. Just an observation:

Math is elegant. That elegance has a dark side and seduces people people away from the real world of data. Math can unequivocally prove their results, whereas real world data never proves anything. At its best, data can falsify a theory. More often data is consistent with several theories. However, messy real world data can be far more valuable than the fragile perfection of math.

Society For Neuroscience Conference Meet-up?

2011-11-07T07:36:00.000-08:00

Society For Neuroscience Annual Conference starts this Saturday, November 12. I'm looking forward to attending every day. If anyone wants to meet-up for coffee and interesting conversation, contact me.

(I recently relocated to the Washington, D.C. metro area so I know a couple places that are off the beaten path.)

Holmium

2011-11-04T07:01:00.000-07:00

I'm currently enjoying The Elements, an amazing book that profiles the entire periodic table. I was surprised to read about Holmium which is used to concentrate fMRI magnetic fields. I have done research with fMRI for years and never heard of it.

It made think me about the amount of time fMRI researchers focus on the technical minutia of fMRI (e.g., Larmor precession). Those technical aspects are only tangentially related to creating meaningful research. Understanding Holmium and Larmor precession doesn't directly contribute to better research. Just as I leave Holmium to the magnet makers, researchers should leave pulse sequences to the physicists and technicians.

fMRI researchers should spend their time asking better questions, collecting and analyzing the best data, and effectively communicating the results.

My Limitations

2011-11-02T13:54:00.000-07:00

I don't have access to _________ technology.

I don't have a grant for _________.

I don't have a degree in _________.

I don't have access to _________ patient population.

Everyone has limitations. What am I going to do within mine?

What I Use: OmniGraffle

2011-10-31T06:33:00.000-07:00

I'm polishing my poster for the upcoming Society for Neuroscience Annual Conference. It is just about the time when I relearn how much humans inherently underestimate the time to complete large projects. I had the the additional thought - The foundation of science is data, but humans are still the audience. It behooves everyone to make posters look "pretty." Software should support that goal with minimal cost (e.g., financial or time spent learning the program).

I have used a variety of software programs in the past to create posters. My first experience was using Microsoft's PowerPoint, the de facto option. It is hard to make posters look pretty in PowerPoint. Creating a slideshow with 1 slide that is 3X5 feet exposes nonlinearities in the software. I spend time wrestling with those nonlinearities rather creating content. I had a brief infatuation with Adobe, mostly Illustrator. There is inherent friction in that system that takes ~5 minutes to open. In my experience, it is too powerful for posters. Given my level of design skill acquisition, I need less choices and more automation. However, Illustrator can make things very pretty if you spend the time to learn it. Recently, I experimented with Scribus. Being an open-source program, it is free but has bugs. As with Illustrator, you have to spend time learning the program before making anything of value.

Currently, I use OmniGraffle for posters. It has the right number of features while being fast and easy to learn. You can create great looking design elements almost immediately. It is straight forward and quick to export current drafts to share with collaborators or the final draft for printing. Check it out if you want your poster to look pretty with minimal fuss.

The Lanuage of Music?

2011-10-28T06:10:00.000-07:00

The "language of music" is a common metaphor both inside and outside of research. It is a powerful, but ultimately limited, idea. Could there be better ones? Here are couple:

Music as nonverbal cognition.
Music as an implicit process.
Music as a series of computations.
Music as music.

Science Edition of The Daily Show

2011-10-27T05:03:00.000-07:00

Yesterday's Daily Show covers science in the Daily Show's informative, but still fun, style. Watch it here.

The Daily Show makes references to the belief in global warming or evolution. All beliefs are wrong. Some are more wrong than others. The better question is what are you going to do with your beliefs. I can do useful things with my (current-but-probably-wrong-in-the-long-term) belief in global warming and evolution.

Technology Always Comes First

2011-10-26T06:14:00.000-07:00

You can either create technology or creatively use it. There is an inherent lag between technology creation and use. That choice extends to science. I can spend my limited time and attention creating novel, improved methodologies. Alternatively, I can use the current, suboptimal methods and do my work today.

Self-Paced Courses

2011-10-24T06:25:00.001-07:00

One of the classes that shaped my views on higher education was my undergraduate introduction to psychology class. I knew that I wanted to study psychology as science by the time I arrived at college. Introduction to psychology was a necessary evil. I signed up for an experimental version at San Francisco State University during my freshman fall quarter. There were two distinct aspects, out of class and in class. Out of class, all the reading was online and self-paced. (It was 1997, before widespread video on the internet). In class, you asked questions and took quizzes. It was perfect for me. I completed the course before Halloween with an A. Interestingly, that class had one of the largest failure rates in the entire university.

Self-pacing is a two edged sword. I was an outlier. I went as humanly possible. I had a clear goal and moved effectively and efficiently towards that goal. Most of the other students didn't. It wasn't their self-pace wasn't too slow, it wasn't a pace. They stood still and were wanting for a push that never came in that class.

Initially, I had thought the format of class was a bad idea. Why create a system that allows most people to fail? It is possible to create a system in which people succeeding is the default. Now I realize the early pruning is a benefit to the students themselves. Those students didn't have the ability to self-pace, an important skill for knowledge work. They had a lifetime of being primarily pushed forward (i.e., it is the 4th grade and now you study fractions). I would do something that wasn't an option - offer a class to teach self-pacing skills.

Temporarily Unbroken

2011-10-21T10:20:00.001-07:00

The world is a mess. Or more precisely, the world is messy. Humans naturally put a structure on it. In cultural terms, that structure is called a world-view. In scientific terms, that structure is called a paradigm.

All world-views and paradigms are, to varying degrees, wrong. There is always data that doesn't fit into the structure. That doesn't mean abandon all structures. All world-views and paradigms are temporarily unbroken but can be useful in getting things done.

What I Use: Mendeley

2011-10-19T10:14:00.000-07:00

During the course of doing science, I have to scale a small mountain of scientific papers. I need a trusted system to handle all that "stuff." My trusted system is Mendeley.

It does everything I want a reference manger to do. Mendeley is a single seamless system for every stage of workflow (i.e., collect, process, organize, review, and do). The keyboard shortcuts allow to me to focus on what I'm doing and move beyond thinking about how I'm doing it. The best part - it is free. I recommend Mendeley to any students, instructors, or researchers.

"That Is An Emprical Question."

2011-10-17T04:28:00.000-07:00

That is one of the favorite and frequent sayings of my Ph.D. adviser. It was said most often in lab meeting after a graduate student made a crazy conjecture. It brought us back to the real world. The conjecture was the start. Before we assumed too much and became too attached to the idea, there is work to be done. Experiments to be designed, data to be collected, statistics to be analyzed, and papers to be written. Then we would in a better position to understand.

What would happen if that was the default question in the world? Applied to everything from what should you do today to how we should tackle social policies.

Scaling Unstructured Category Learning

2011-10-13T07:55:00.000-07:00

Scientists, like all humans, are attracted to structure. Scientists studying category learning have spent most of their time and attention on structured categories. The structure underling the categories could be either rule-based or feature relationship-based. However, there are many categories that are unstructured. One example of an unstructured category is "my passwords." Hopefully, there is more than one exemplar in that category, and none of them are "password." "My passwords" are most likely alphanumeric strings, but there is no underlying rule or feature relationship uniting members of the group.

Very little is empirically known about unstructured category learning. One aspect I have recently investigated is the finite human capacity to learn unstructured categories. I was manipulating the number of exemplars per category to validate a particular set of stimuli. For this set of stimuli as I increased the the number of exemplars per category, performance could be characterized as "very easy," "easy," "kinda of hard," and "impossible." I'm not unclear what underlies this particular anecdotal result. It would better serve the field to better understand this empirical phenomenon and other unstructured category learning properties before assuming neurobiological mechanisms.

Meaningful (And Enjoyable) Science

2011-10-10T10:05:00.000-07:00

On my lifelong quest to better understand how to "solve the right problems at the right level for the right reasons," I stumbled upon this article. It suggests weighing the gain in knowledge against difficulty as criterion for selecting the right problem to solve. If Csikszentmihalyi's challenge vs skill concept is added to the mix, science can be meaningful (and enjoyable) for everyone (including myself).

One person's tool ...

2011-10-07T07:54:00.000-07:00

Last weekend, I was walking through a flea market and saw typesetting equipment for sale as home kitsch. That was tool was someone's livelihood, now it is a wall hanging. A single individual was the last person to own it as a tool. What happened to that person when the business changed? Did he or she consider themselves a typesetter (wedded to a technology) or person who brought books into the world (wedded to a process)?

This applies to the science of psychology. The tachistoscope is well on its way to kitsch. How long for the fMRI scanner?

Do you define your job as a technology or a process?

Anhedonia & Category Learning

2011-10-05T09:28:00.000-07:00

A recent study found that unmediated individuals with major depressive disorder have reduced caudate response to rewards. Given the caudate's role in category learning, I would expect anhedonia to adversely effect performance in information-integration category learning tasks. No one, to the best of my knowledge, has investigated the correlation between the two. Any takers?

Ben Goldacre on Science

2011-10-03T12:38:00.000-07:00

This video reminds me:

Doing science is fun.

Talking science is fun.

More importantly, teaching science should be fun.

The Real Cocktail Party Effect

2011-09-30T08:01:00.000-07:00

"So what do you do?"

The social devil on my right shoulder wants to say, "I solve the right problems at the right level for the right reasons." The social angel on my left shoulder more often wins and I say, "I do research." That opens the door to the logical follow-up, "What do you research?" The angel responds with "Learning", but the devil wants to say "Neuroscience."

When I say the n-word, the conversation frequently stumbles or stops. It intimidates people which is not the best way to continue a conversation. I like the term "learning" because everyone has experience with learning. Later I sneak in the neurobiology. Usually I try to leave them with a "take home message." (It is hard to stop the teacher in me.) One of my favorites is the importance of feedback timing in procedural learning (starting on p. 125), not your usual cocktail chitchat but still interesting.

Book Review: We Are All Weird by Seth Godin

2011-09-22T08:05:00.000-07:00

I'm a big Seth Godin fan. I tore through his latest offering We Are All Weird and want to share it with the world. It is a manifesto, a concise document meant to change something in the world. The basic idea of We Are All Weird is that normal is not so normal anymore. The world is weird and getting weirder. It would behoove everyone, from marketers to academics, to recognize it and move forward.

It is a fun and challenging read, a rare and welcome combination. There are amazing examples of the world getting weirder that range through time and location. Since I read it on the Kindle, I was able to quickly dig deeper into the wide reaching references.

It was eye opening to hear Seth's take on one of my personal passions, the education system. He points out the current limitations of encouraging everyone to be normal. You lose both ends, the highest and lowest achieving students. We now have the tools and techniques to restructure the education system to reach every student more effectively. We can address the lower achieving students, who are frequently studying the wrong material for the wrong reason. We can also encourage the higher achieving student to the greater levels by becoming leaders and teachers.

I do take issue with his, often improper, use of the normal or bell curve. One of the primary prerequisites for assuming a normal distribution is a single variable that is at least interval scaled. For example, weight can be be understood by using a normal curve. However, Seth plots cultural behaviors from various decades on a normal distribution. Cultural behaviors are not a single variable and are nominally scaled. Additionally, normal distributions are only theoretical. Actual data (if appropriate) should be displayed in a histogram. This might appear to be a minor point, but when he misuses statistics he loses scientific face. He can make his argument without misusing statistics.

I recommend We Are All Weird, but get Linchpin first.

Publication: A critical review of habit learning and the basal ganglia

2011-09-01T08:07:00.000-07:00

I'm proud to present my latest publication - "A critical review of habit learning and the basal ganglia."

Abstract: The current paper briefly outlines the historical development of the concept of habit learning and discusses its relationship to the basal ganglia. Habit learning has been studied in many different fields of neuroscience using different species, tasks, and methodologies, and as a result it has taken on a wide range of definitions from these various perspectives. We identify five common but not universal, definitional features of habit learning: that it is inflexible, slow or incremental, unconscious, automatic, and insensitive to reinforcer devaluation. We critically evaluate for each of these how it has been defined, its utility for research in both humans and non-human animals, and the evidence that it serves as an accurate description of basal ganglia function. In conclusion, we propose a multi-faceted approach to habit learning and its relationship to the basal ganglia, emphasizing the need for formal definitions that will provide directions for future research.

PDF

Objective Criteria

2011-08-15T13:24:00.000-07:00

When is a paper ready for review?

When is a poster ready for the printer?

How should you decide knowledge work is done?

Frequently, it is subjective or extrinsic. It looks done or it is due (more often overdue). I strive for the higher standard of objective criteria. My work on a daily basis has objective criteria (e.g., I will not get another cup of coffee until I put 500 more words into Scrivener). It is harder, but important, to have measurable, therefore manageable, criteria at the project level.

One of my favorite is "Running The Stack." For every project, I have a stack of index cards. The stack includes defined outcomes, all currently-known steps for the outcomes, and a someday/maybe list for the project. "Running The Stack" is an out-loud reading of a project's index cards. It is a mental sluicing, looking for the golden ideas in the dross. If I can "run the stack" without generating any novel, actionable ideas, I know that when all the current index cards are crumpled in the trashcan I'm done with the project.

Footnote: It is allowed to move an index card to the stack of another project to clear the stack of the current project.

A Different Appoarch to Learning

2011-08-02T07:28:00.000-07:00

I believe that an over-learned foundation in the basics of a subject is a prerequisite for tackling more interesting problems. A system like the Khan Academy is the fastest method I have found to create (or reinforce) that foundation.

I'm off to brush-up on Calculus.

Bluish White Collar

2011-07-29T11:36:00.000-07:00

I make stuff. Everyday. I get up, grab my lunch pail, make coffee, and boot up my computer. I like to call the stuff I make widgets so I don’t get too full of myself. They are most frequently thoughts and ideas. Sometimes there is a physical manifestation (e.g., poster, presentation, or paper).

A major difference between me and the widget makers in the Rust Belt (RIP) is that I’m my own foreman. Yes, someone else writes the checks but most of the time it is up to me to decide which widgets to make today.

Good Push Back

2011-07-21T09:37:00.000-07:00

Right now the world changes faster than our mental models and systems. Asking "Why is ...?" is the best method to cut through the clutter of outdated thinking. The answer to that question often implicitly includes some combination of these statements:
"I hadn't thought about it."
"That is way it has always been done."
"Someone told me so."
"There is not a better way."

The mere act of making them explicit and named is the start of a better solution.

Science-biased Conversations

2011-07-19T11:12:00.000-07:00

Most of the people in my extended social circle have earned master and doctoral degrees. It is a delightful mix that ranges from social scientists to engineers. In addition to proving their ability to sit still and delay gratification, a post-bachelor degree usually requires rigorous scientific steeping. This creates certain implicit guidelines for conversation. Scientists realize the difference between narrative and facts. Disagreements are frequently about definitions. They realize attributing causation is a tricky. Most importantly, scientific training reminds you that you could be wrong. This does not guarantee improved conversational content, but these implicit guidelines are more conductive to the successful exchange of ideas. I wonder what our national discourse would sound like it if was more scientific-biased.

My 3 Types of Questions

2011-07-10T18:33:00.000-07:00

My foundational research questions start with a "Why".

Why choose the basic science end of the research spectrum over the applied science end? Why focus on understanding "normal" brain function, in contrast to "abnormal" brain function? Why seek to understand the adult (18-40 years of age) brain, as opposed to the younger or older brain? These questions provide perspective and grounded motivation for what I do.

My primary research questions start with a "What".

What is the end goal of this study? What are the subgoals of this project? What does successful completion of this project look like? These questions set up the trajectory of any given project.

My secondary research questions start with a "How".

How best answer do I answer my primary research questions with my limited resources, whether they are money, time, or attention? How do I construct a project work flow? How do I best collaborate with other members of my team? How do I collect and analyze my data? These questions outline the way I walk the path of the project. They put into focus the vitally important details but are always addressed last.

Accepted to SFN 2011 Annual Conference

2011-07-01T11:23:00.000-07:00

My abstract for "Expectancy violation and functional connectivity in musical syntax processing" was accepted to Society for Neuroscience 2011 Annual Conference in Washington, D.C.

Session Type: Poster Session Number: 171
Session Title: Auditory Processing: Human Studies
Date and Time: Sunday Nov 13, 2011 8:00 AM - 12:00 PM
Location: Walter E. Washington Convention Center: Hall A-C

See you there!

Leveraging The Long Tail

2011-06-27T14:47:00.000-07:00

Psychology tends to get to data from a medium number of trials, about 100, from a medium number of subjects, about 30. Occasionally, psychology, particularly perception and sensation research, will get a large number of trials, say 10,000, from a few subjects, say 4 (including the first author).

What happens if you go to the edge. How about a large number of trials (a million trials would be a good start) from one person? How about 1 trial from a large number of people (again, I would start with a million)? Those edges are easier than ever to access via the internet. Using the "long tail" of the internet has been prototyped in science by the Galaxy Zoo project. It is time for psychology to join the game.

Required Reading

2011-06-22T09:04:00.000-07:00

The reading list for seminars and conferences sometimes border on the ridiculous. Since everything is electronic, it is additively easy to find and share information. There is always another article that is interesting and possibly relevant. I'm a fast reader but I shepherd my superpower in this instance. In my experience, there are more important aspects to a successful seminar and conference attendance than doing all the required reading.

I show up, my secret weapon. First, that means I am physical at the conference or seminar. Second, I am mentally there. I have "nothing on my mind", a Getting Things Done reference. I have my 8 hours of sleep. While I am physically in the room, I not checking my email, twittering, or looking up what is happening elsewhere. I know this is where I need to be in the universe. I feel like I'm cheating when I fully present at conferences or seminars, even though sometimes I don't all the mandatory reading.

Even though I just made an argument for not reading, sometimes it is better if you show up prepared. Here are my tips & tricks:

1. Read all the abstracts
2. Rank order the articles
3. Read the 1st paragraph of the general discussion
4. Summarize the article in your own words
5. Read the methods
6. Create a list of limitations based on the methods
7. Interpret the figures
8. Read the results
9. Read the discussion
10. Create a list of limitations based on the discussion
11. Read the introduction
12. Create a list of research they should have cited (mostly your own!)

Cognitive Chew Toy

2011-06-01T11:02:00.000-07:00

Every researcher I know has a “Cognitive Chew Toy.” A cognitive chew toy is an aspect of research that you love to think, talk, and obsess about. I have two cognitive chew toys, designing studies and creating workflows. I love both. They are a type of thought experiment that will be manifested in the real world. I think they are critical foundations to successful research. However, there is a dark side. Those “fun” aspects need to be balanced with the other facets of research that are less delicious but equally important.

Eating more corn doesn't make you a better farmer,

2011-05-10T08:53:00.000-07:00

just like more reading articles doesn't make you a better researcher.

This post was inspired by Merlin Man.

Digital Paper

2011-05-07T08:15:00.000-07:00

Every scientific journal article is now a pdf. However, most journals treat pdfs as digital paper. That is similar to thinking about cars like faster horses. Both are limited viewpoints. Neuroimaging is inherently multidimensional. The richness of the data can not be conveyed in a table and a screen capture. Supplemental materials, when they are available, are more tables and 2-d images. The next logical step in my own research is a video stepping through all brain slices. Why stop there? How about an 3-d image viewer for fMRI results? That would allow the reader to view any part of the brain from any viewpoint. How about a screencast of experimental procedures? That would be a better way to convey attentional blink paradigms.

Robustness over Prediction

2011-05-02T07:39:00.000-07:00

If I can't predict tomorrow's job market, how should I advise first-year students, undergraduate or graduate, who are preparing a job market in 5-7 years. Instead of providing a false of security via predictions, I advocate being robust.

Here is my short list of academic robustness:
Diverse research and teaching skills, within and between disciplines
No debt, both in loans and obligations
Low overhead, both professional and personally
Plenty of free capital, in time, attention, and money

This post was inspired by Nassim Taleb's book Black Swan.

The Next Generation Book

2011-04-30T09:11:00.000-07:00

Imagine applying that technology to neuroanatomy textbooks
(or neuroimaging software manuals!)

Academic Picking

2011-04-28T12:37:00.001-07:00

Students are picked by universities.
Faculty members are picked by departments.
Postdoctoral fellows are picked by principal investigators.
Principal investigators are picked by grant committees.
Papers are picked by journals.

All that picking presupposes scarcity.

What happens when:
Universities aren't the only way get a education.
The grant system isn't the only way to conduct groundbreaking research.
Killing trees isn't the only to communicate science.

This post was inspired by Seth Godin's post.

3 Questions (and no answers)

2011-04-21T07:00:00.000-07:00

It is amazing (almost overwhelming) the amount of information that comes across my eyeballs over the course of a day conducting research. There is almost always one more article to read or one more set of analysis to do.

I ask the following questions every time something demands my finite research resources:

What is the theoretical underpinnings?

What is the empirical track record?

What is the practical implementation?

It is still important to ask these questions, even if there aren't always clear answers.

These questions are fundamental to the doing of quality research (in contrast to talking about the doing of quality research). However, they are not limited to just that endeavor. I'm starting to ask them of anything that demands my finite time and attention.

Book Review: Do The Work by Steven Pressfield

2011-04-20T08:43:00.000-07:00

Steven Pressfield's newest book escaped today (I think that books, like scientific manuscripts, aren't published, they escape). Do The Work is an in-the-trenches manual for guiding anyone from start to finish on any project that is important in his or her life. The book is targeted at anyone that does work. If you take David Allen's definition of work, anything you want to get done that's not done yet, then the book is applicable to everyone. In addition to being a pragmatic guidebook, Do The Work is a easy book to read. The text leaps off the page. It is less than 100 pages, as long as it needs to be but no longer. For someone that reads a lot, both aspects are appreciated.

The section most appropriate for scientific researchers is "A research diet" (page 19). Steven writes, "You're allowed to read three books on your subject. No more." My translation for researchers - " You're allowed to read three articles on your subject. No more."

It initially sounds counterintuitive. For someone who loves research, reading less seems fundamentally wrong. The questions that every researcher must ask is "Am I someone who reads about research or someone that actually does research?"

I'm guilty of procrastination via attempting to know everything. You will never find (let alone read) every journal article, book, or blog post that is relevant to your current project. Find the critical few, then go Do The Work. You can deal with "must read" stuff later.

At some point the Resistance will ask, "Are you missing something?" You must answer, "I could be BUT I can catch it later. If not, my collaborators will catch it. If not them, the reviewers will catch it. If no one does, it is not that important." Like in Fight Club, "No fear. No distractions. The ability to let that which does not matter truly slide."

I'm not suggesting that you do a low quality literature review, but I've witnessed every researcher I know procrastinate by overfixating on the literature review. Do no more than the minimum. Your work is more than reading other people's work.

Simply:
Go do something. Anything important to you.
If the Resistance attacks you (it will),then go read "Do The Work."

Changing Education Paradigms

2011-03-01T09:59:00.000-08:00

I wrestle with the ideas presented above when I mentor students. I interact with them at end of the "factory line". They pay attention and work hard (or else they wouldn't be upperclass college students). However, they struggle with real-life problem solving, generating novel ideas, and working in teams. Those are the traits I find most valuable in research assistants.

I retrain them to be Kindergarteners.

The Family Tree of Neuroscientists

2011-02-01T10:04:00.000-08:00

Neuroscientists are a geeky, fun loving group but tend to be inbred. In case you were wondering how inbred, this website shows the relationships between most neuroscientists.

A Couple of New Ideas (No Charge)

2011-01-01T10:06:00.000-08:00

My summer reading list includes Suzanne N. Haber's recent article "Functional Anatomy and Physiology of the Basal Ganglia: Non-motor Functions." (Don't Judge!) It begs many questions, in a good way. She outlines distinct neurological properties of the basal ganglia. Those properties might influence distinct behavioral properties. On page 37, she states medium spiny neurons are "bistable, shifting between two membrane states: an upstate and a downstate." Although this has been investigated, I think the full ramifications have not been found yet. Later, she discusses a set of interneurons that receive "powerful" input from cortex and respond faster and earlier than medium spiny neurons. The nature of this cortical input and the effect of being faster and earlier could have profound implications for basal ganglia functions.

I'm too busy with other projects to explore these ideas. I would love to see someone else tackle at them.

Applying Neuroscience

2010-12-01T10:08:00.000-08:00

An interesting take on neuroscience

Nassim Taleb on Neuroscience

2010-11-01T10:12:00.000-07:00

"As a skeptical empiricist I favor the experiments of empirical psychology to the MRI scans of the neurobiologists, even if they appear to be less ‘scientific” to the public"

Source (Thought #15)

Lately, I have been challenging myself by reading the work of Nassim Taleb. I suggest Fooled By Randomness and The Black Swan. His take on neurobiology does not sound favorable, but I don't think it applies to me. First, my science, although primarily MRI, is rooted in empirical psychology and mathematical modeling. I'm not fishing. I have falsifiable theories that I set out to disprove. If I fail, I tentatively accept them until a better theory comes along.

Second, I constrain my mathematical models of the human brain via neuroanatomy. Many times everything connects to everything in pure neuroanatomy, not offering the constraining parameters necessary for a powerful model.

One hope is functional anatomy which is coming into its own through new methods (e.g., DTI). It should provide increased constraints relating form to function. It is an exciting time, especially if you can question your assumptions about brain structures and their functions.

I leave you with science as art

The reconstructed white matter fiber tracts of the mid-sagittal plane of the human brain.