Executive Director's Column

Infrastructure for the Science of Psychology

Chief among twenty-first century challenges is to identify and build the kind of infrastructures that will allow our discipline to prosper and grow.

By Steven Breckler, PhD

Oceanographers have large ships to help with their work. Astronomers and astrophysicists share an amazing network of telescopes perched on mountaintops and that orbit distant planets. Biologists depend on electron microscopes, economists, sociologists and political scientists prize their large-scale surveys, and geographers are counting on geographic information systems to carry their science into the 21st century. These are all examples of infrastructure - shared and costly resources that provide widespread benefit for one or more scientific disciplines.

What is the infrastructure of psychological science? We don't have boats, telescopes, or satellites to do our work - nor do we need them. Most of the work of psychology gets done with off-the-shelf and general-purpose technology: microcomputers, tape recorders, video recorders, and lots of paper and pencils.

Psychology can claim credit for the technology of measurement and assessment, which offers a ubiquitous infrastructure for research on personality, individual differences, learning, and psychopathology. Perhaps not unique to psychology, research with animals depends importantly on species that have been carefully bred for specific characteristics. This is infrastructure.

Perhaps the best example of a potentially sharable and costly resource for psychology is in the area of cognitive neuroscience, which depends heavily on magnetic resonance imaging (MRI) facilities and expertise. Indeed, the innovations of functional MRI have produced a revolution in our understanding of human and animal cognition.

Thinking of fMRI as an infrastructure for psychological science helps to highlight several important issues. One is a trend for researchers to want their own - to have their own scanner right down the hall, and available for their own use 24/7. In part, this is an understandable response to a common situation in which researchers must currently borrow or rent MRI time (and not typically prime time) from a facility housed in the medical school or university hospital located across campus or even across town.

Of course, MRI facilities are expensive to build and to maintain. It makes more sense - pragmatically and economically - for scientists to share MRI resources rather than each acquiring their own. The problem is that this approach flies in the face of the culture of scientific psychology, which tends to favor individual control and ownership of data, facilities and equipment. The federal funding agencies have been willing to indulge researchers by providing support for very local imaging facilities, reinforcing the culture of individual proprietorship.

The idea of a shared resource is critical. fMRI has not yet achieved the status of true scientific infrastructure, because the technology and the facilities are not widely available. It still takes a lot of money and friends in the right places to gain access to this resource. As a discipline, psychological science could reap enormous benefit by investing its energies and funding resources in building imaging facilities that are widely available and shared by all of its citizens.

My sense is that the idea of sharing is not a popular one in scientific psychology. A case in point is the use of large-scale datasets. Psychologists like to collect their own data, analyze their own data, and publish their own data. This is reasonable and sensible when the scientific problem can be properly investigated and understood within the limits of a single laboratory in a single location relying on a locally available population of people or animals. Yet, many problems worthy of investigation are too big for this approach. What do we do when a nationally- or internationally-representative sample is needed? What do we do when the efforts of hundreds or even thousands of researchers are needed to collect the relevant data? What do we do when the cost of collecting those data exceed the size of an average single-investigator grant from NSF or NIH?

The answer is that we pool our resources - our money, our talent, our local facilities - and we create a shared data resource. One example relevant to psychology is the NICHD-funded study of early child care. Another example, currently being contemplated by a consortium of federal funding agencies, is the National Children's Study which would examine the effects of environmental influences on the health and development of more than 100,000 children across the United States, following the children from before birth and into early adulthood. These are both examples of true scientific infrastructure - shared and costly resources that provide widespread benefit for one or more scientific disciplines.

As we charge into the 21st century, the science of psychology faces many challenges. Chief among them is to identify and build the kind of infrastructures that will allow our discipline to prosper and grow. APA is well-positioned to take the lead in identifying the infrastructure needs of psychological science and advocating for support in federal funding agencies. But we can't do it alone - we need to share this one.