A Case Against Bayesian Neuroscience

Bayesian inference is a powerful methodology but can be over-extended. Above is an example of over-extending the Bayes framework to neuroscience. Given its power, Bayes inference can easily account for behavioral phenomena, but it does not provide a plausible neural mechanism. There are no Bayesian neurons or areas in the brain.

Starting with the underlying architecture of the brain (i.e., neurons, synapses, and neurotransmitters) puts  neuroscience on a path to develop more robust models. Models built from the underlying neural architecture of the brain would be more complex than the simple elegance of Bayesian models but provide the opportunity for deeper insights into real brain processes.

The Best Use of My Time

It is the common, easy path to let others govern minutes. Check-in with my boss for a quick task. Check email and shoot a quick response. Peruse a social media website. There will always be someone else available to dictate my next action.

Those minutes add up to the hours of the week (and months of the year).

At the end, I have to stand true for my time.

Like most aspects of life, it is better to the opposite of what is common and easy. I reverse engineer time management. What do I want true in one year? Then, how do I spend the next minute to get there?

Free Range Model Comparisons

Comparing models is fundamental to all science, for example, Newtonian vs. Einsteinian (Physics), Uniformitarianism vs. Catastrophism (Geology), and Specialized vs. Generalized brain areas (Neuroscience).

Model comparison implies your current patterns of thoughts (i.e., a paradigm) could be replaced with different patterns of thought that better fit the world-as-is (i.e., the data). Just opening the door to that possibility is powerful (and scary).

Imagine a world with that potential for openness in all aspects of life.

Design Patterns in Science

Design patterns are reusable solutions to commonly occurring problems within a given context. For example if I write a computer function to present a stimulus, I reuse that function in all future experiments.

The same heuristic can be applied to other recurring problems in science, ranging from lab maintenance (e.g., having a current consent form) to creating sustainable lab culture (e.g., training new researchers).

It worth spending time to create and maintain the infrastructure to solve a problem only once. As I create a deliverable, I also create a separate file of notes and steps that enabled me to solve the problem. My design element kit includes both the final product and production capacity elements.

The real art is applying the same design patterns across problems and projects.

My New Postdoctoral Fellowship

I am delighted to announce my new postdoctoral fellowship with Thomas Carlson at The University of Maryland.

He has assembled a dynamic team to conduct ground-breaking research in object recognition vision science. I look forward to contributing to it. The position is a great platform to continue my life mission - researching and sharing neuroscience. The first step is transitioning my previous category learning paradigms to a categorization paradigms, a distinction with a difference with regards to the questions worth asking. I will be expanding my neuroimaging toolbox to include magnetoencephalography (MEG).

If you are in the same neck of the woods (either physical location or research area), drop me a line.

When to hold fast and when to scuttle

There are two kinds of research mistakes: A mistake that scuttles a project and all the rest.

An inherent design confound could end a research project. In category learning research, participants respond to stimuli with motor movements. It is hard (but impossible) to get a response without a motor component. The category learning paradigm includes all the processes associated with motor responses. Any project that wants to look at category learning orthogonally to motor responses is dead in the water.

All other mistakes do not stop a project unless you let them. The worst case scenario is more pilot data than you intended to collect.

An example of a mistake that appears stops a project is discovering someone has published a similar study. Many people scuttle projects at this point. I do not. I now have the opportunity to be a artist and create something more interesting within already proven paradigm.

Pick the right fight the education revolution

While developing materials for my fall class, I am modeling other classes that have leveraged current technologies (e.g., ubiquitous and asynchronous access to information).

However, the world-as-it-is-now lacks many of the resources to easily and successfully bridge the gap between the predigital classroom and a classroom that maximizes current technologies.

It is the classic trade-off between production and production capacity. My primarily mission is producing a great class, regardless of digital integration. During the course of the process, I will increase my own technology production capacity and possibly do the same for future classes.

One example of raising the production capacity of university education is profiled here.

The Heart of Science

There is a cottage industry of popular press articles, blogs, and podcasts that rehash findings from scientific journal articles. While constructing those narratives, the ambiguity and nuisance of science is airbrushed. The wrinkles of low sample size are lessened. The birthmarks of confusing correlation and causation are removed.

The same happens within scientific journal articles. The introduction and discussion section construct a narrative around the facts and figures of the method and the results section. The rough edges of confounds and raw data are smoothed over.

The heart of science lies within the method and the results section.

When trying to understand the science of any topic, eat the raw heart first.

Improving at the Craft of Scientific Writing

On my continuing quest to improve my writing ability, I stumbled upon these nuggets.

Start with the second link first - "Techniques for Clear Scientific Writing and Editing." The six principles and pithy examples are my Elements of Style.

"Say it Simply: Tips for Clear Writing" shows complex original scientific sentences and the improved versions. Most vitally, it explains the rational for the improvements.

Book Suggestion: My Stoke of Insight

I just finished My Stroke of Insight by Jill Bolte Taylor. It is an informational and inspiring read for anyone interested in the brain and effects of stroke.

Jill gives the best basic introduction to brain anatomy and function I have read. She makes a complex topic as simple as possible, but never simplistic. She recognizes the separate features and functions of the two brain hemispheres to create our coherent conscious experience. She suggests methods to maximize each of them separately. That concept is also presented in this book - Pragmatic Thinking: Refactor Your Wetware.

The book is equally inspiring. She uses her traumatic event as a catalyst to change the meaning of her life. The books opens a door into alternative ways of the perceiving the world, without being dictating the precise path.

On a personal note, my mother had a stroke 10 years. The book provided me insights into what she went through during the event and on her recovery path.

I am so enthralled with the book I am assigning it as supplement reading for my fall course in Sensation and Perception.