Is the Internet Bad for Science? Not really!

In continuing with the controversial blog posts about science, Wired has a preview of an article coming out in the prestigious journal Science, which hypothesizes that the Internet allows for such pinpoint precision in finding an article that researchers miss out on interesting topics because they don’t have to leaf through paper journals. The article author, James Evans, has data to back it up. From the article:

“As more journal issues came online, the articles referenced tended to be more recent, fewer journals and articles were cited, and more of the citations were to fewer journals and articles,” writes Evans, who analyzed the citation patterns of 34 million journal articles that went online between 1998 and 2005.

It’s an interesting hypothesis, and it makes some sense. The method I use for keeping up with research is to get e-mail alerts from the journals I read updating me with their new issues and sending me the Table of Contents. It is automatically sent, then I can scan the titles to find something interesting.

This has the benefit of keeping me up to speed on the most recent research, but does not let me find old research on a topic that may interest me. This the main points of the article, and that may be a shame. Literature reviews may just not be as thorough as they used to. Also, though, some sciences (such as cognitive psychology) have really exploded recently, and only the more current articles are the relevant ones. I tend to cite a few tried-and-true articles, but most of my literature review focuses on the recent stuff. I’m curious to see how it breaks down by the age of the field.

Still, food for thought.

The End of Theory? Unlikely!

Wired Magazine posted an essay online that proposes that:

1. We have a ton of data and because of how much information we can store (what Wired calls “The Petabyte Age”), we can always have it on hand.
2. The incredible amount of computing power we have allows us to sift through data and run every possible statistical test on it until something comes up significant.

From the article:

There is now a better way. Petabytes allow us to say: “Correlation is enough.” We can stop looking for models. We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot.

The author of the article, Chris Anderson, is not a scientist, and it shows. Researchers already get enough false correlations without running thousands of statistical tests - five or ten is often enough to find some significant correlation, let alone thousands.

What Mr. Anderson is proposing is a reversal of the scientific method. First, collect a bunch of data. Then, run thousands of statistical tests to see what correlated. Then, invent a theory to explain that correlation. It’s so painfully simplistic and shows such an utter lack of understanding about the way science is done that it’s laughable. In psychology, how do we know what data to collect? If we go about research atheoretically, then we would need to collect every possible piece of data about every person so that we can throw it at the computer to see what pans out. Without theory, there is no place to begin.

Sadly, even the editor-in-chief of Wired does not understand science well enough to write an informed article.

Memory for faces is better than memory for other objects, but only under specific conditions

An article written by me, my advisor, and another professor at George Mason University just got accepted for publication in the journal Cognition:

Wong, J. H., Peterson, M. S & Thompson, J. C. (in press). Visual working memory for objects from different categories: A face-specific maintenance effect. Cognition.

Abstract: The capacity of visual working memory was examined when complex objects from different categories were remembered. Previous studies have not examined how visual similarity affects object memory, though it has long been known that similar-sounding phonological information interferes with rehearsal in auditory working memory. Here, experiments required memory for two or four objects. Memory capacity was compared between remembering four objects from a single object category to remembering four objects from two different categories. Two-category sets led to increased memory capacity only when upright faces were included. Capacity for face-only sets never exceeded their non-face counterparts, and the advantage for two-category sets when faces were one of the categories disappeared when inverted faces were used. These results suggest that two-category sets which include faces are advantaged in working memory but that faces alone do not lead to a memory capacity advantage.

---

So what does this mean? Other research has demonstrated that visual working memory seems to be able to hold more information about faces than other objects such as cars or inverted faces (which are processed in an entirely different manner from faces). This is a sensible conclusion, as faces have been shown to be processed uniquely in terms of perception. However, this is the first evidence that faces are unique in terms of working memory.

Our evidence supports that conclusion but also narrows it down. We found no specific advantage for faces if you have to remember a bunch of faces together or a bunch of other objects together (all houses, butterflies, or bodies in motion). This was seen for remembering two or four objects. However, if you have to remember two objects from two categories - two faces and two houses, for example - your memory was better than for four objects from the same category. Additionally, this only happened when faces were part of the set. Remembering two butterflies and two houses together did not lead to better memory.

Therefore, faces are unique in memory, but an advantage is only seen when faces are remembered with non-face objects. Why? It’s possible that there are distributed (but overlapping) representations of faces and all other objects, almost like a Venn diagram with two partially overlapping circles. When only faces or only other objects are remembered, only one of the two circles is activated in support of memory. When faces and other objects must be remembered, both circles can activate, increasing memory capacity. This theory, however, is for another experiment that we’re working on.

By the way, this is the same research that was presented at the 2008 Vision Sciences Conference held last May. Click the poster image below for the PDF:

Mental Doping: The future is here

From an article in The Economist (pdf):

For thousands of years, people have sought substances that they hoped would boost their mental powers and their stamina. Leaves, roots and fruit have been chewed, brewed and smoked in a quest to expand the mind. That search continues today, with the difference only that the shamans work in pharmaceutical laboratories rather than forests. If asked why, the shamans reply that they are looking for drugs to treat the effects of Alzheimer’s disease, attention-deficit disorder, strokes, and the dementias associated with Parkinson’s disease and schizophrenia—and that is the truth. But by creating compounds that benefit the sick, they are offering a mental boost to the healthy, too.

Such drugs are known as cognition enhancers. They work on the neural processes that underlie such mental activities as attention, perception, learning, memory, language, planning and decision-making, usually by altering the balance of the chemical neurotransmitters involved in these processes. This week a report from the Academy of Medical Sciences, a British learned society, says that a large number of such brain-affecting drugs are likely to emerge over the next few decades.

This was likely inevitable. Many diseases of cognition, such as Alzheimer’s, are treated by enhancing cognitive processing. In patients with a disease, this brings them closer to normal cognition. For those who are already functioning normally, the drugs seem to increase their cognitive abilities even further. Alzheimer’s drugs pinpoint memory centers of the brain, Ritalin focuses attention, and the list goes on.

Already, a poll of readers of the journal Nature showed that 20% of readers take some kind of brain-boosting drug. In the highly competitive “publish-or-perish” world of academic research, will taking these drugs be vilified like steroid use, users branded as cheaters? Will scientists need to go through drug testing? Or will this become commonplace so that every professor needs to take them in order to make tenure?

Vision Science: Reaching across disciplines

I am back from Vision Sciences, and it was a great conference. There were lots of good talks and posters, and I found some research that is relevant to my dissertation. Thankfully, nothing actually scooped my dissertation, so I don’t have to start over. Whew.

What I’ve noticed in the past few years is how broad the research has gotten at VSS. It used to be primarily behavioral research. Participants performed a task that involved some aspect of visual cognition, and accuracy, reaction time, and sometimes eye movements were recorded and analyzed. There is still a lot of this kind of research today, and it is still valuable.

In the past few years, though, there has been an explosion of neuroscience and computational modeling. A huge portion of psychological research is veering towards neuroscience, and this is the next logical step in understanding the mind and brain. Also, it gives psychological research an air of legitimacy, since we can tell our friends and family we’re neuroscientists, not psychologists. But studying the brain in conjunction with the mind makes sense.

Additionally, developing computer models to mimic and understand how visual processing takes place makes sense towards developing artificial intelligence and being able to test various theories about different phenomena. Psychologists are often not computer scientists as well, so it is promising to see links being forged between these two disciplines.

As a whole, the field has broadened beyond psychology, so the term “Vision Science” is apt. This is analogous to some schools offering a Cognitive Science program, which looks at cognition from not only a psychological standpoint, but also from philosophical, computational, and biological ones as well. Before we know it, graduate training will expand from learning experimental design to programming in MATLAB and analyzing fMRI data.