Cover Story

The power of information technology is putting another important issue on psychologists' plates besides free access: the ability to share large-scale electronic research databases.

The concept--actually mandated by federal grants--is in some ways just as controversial as free access (see main article). Some scientists welcome it with open arms; others see problems with it being imposed on them. But data sharing has only recently received wider attention as electronic options become more widely understood and used--and psychology in particular is still trying to catch up to its promised opportunities, says Merry Bullock, PhD, APA's associate executive director for science.

Three factors have kept psychologists from using it more widely, Bullock believes. For one, it has not been part of psychology's tradition to build or share large-scale data sets as has been the case in other scientific disciplines. Relatedly, "The typical psychological study is smaller and more likely to use idiosyncratic measures than studies in other scientific fields that regularly share data bases," she explains.

Another impediment is that much of psychology's data come from identified human subjects, Bullock says, so privacy and confidentiality issues must be addressed before data sharing becomes more commonplace.

A third obstacle is that wide-scale data sharing simply isn't part of the culture of the field, Bullock notes. "Collecting one's own data tends to be more highly valued in psychology, and it makes up part of our traditional reward structure," she explains. "There are some wonderful exceptions of psychologists conducting large-scale projects that build and make use of digital data sharing and databases, but by and large we are not very good yet at using these kinds of data." And Gary VandenBos, PhD, APA's publisher, notes that there are also additional costs to researchers for preparing the raw data for public release in a standardized format.

That said, the potential benefits of large-scale data sharing far outweigh the costs, Bullock believes. Pluses include increasing the power and generalizability of psychological findings by increasing effective sample sizes, getting more value from costly research investments and encouraging research across investigators and institutions.

To help psychologists get more comfortable with the idea, APA's Science Directorate has launched two initiatives, Bullock notes. One is an advanced training institute to help researchers learn secondary analysis of large databases. Funded by the National Institute of Child Health and Human Development (NICHD), the program's first session was last August in Chapel Hill, N.C., and had participants work directly with longitudinal data on 1,364 families from NICHD's ongoing Study of Early Child Care (SECC). This year's institute, May 17-24, will also be held in Chapel Hill and use the SECC database. 

The other initiative is an expanding Science Directorate Web page on data sharing and archiving, which features publicly accessible behavioral and social science databases, some of which are immediately accessible or free, and others of which require registration, fees or membership, says Bullock. (Go to www.apa.org/science and click on "data sharing" on the "Science A-Z" list.)

"We hope to inspire psychologists on the potential value of data sharing," Bullock says. "At its best, digital data sharing allows scientists to ask new questions and conduct smart and innovative research."

--T. DeANGELIS