ACCOUNTABILITY IN RESEARCH

Accountability is a central issue in ethics and politics, one closely related to other concepts such as responsibility, integrity, and authenticity. In ethics, individuals are held accountable for their actions. In a democracy, citizens of the state ultimately hold politicians accountable. In both instances, however, there are questions about how such accountability is to be practiced, and in reference to what standards. Similar questions arise with regard to accountability in scientific and engineering research. Accountability in research or research accountability as general terms may thus refer to a range of concerns and practices related to the philosophies, policies, systems, procedures, and standards for analyzing and promoting ethical conduct in research.

In the worlds of business, finance, and government, accountability also implies a more specific reference to accounting in the sense of bookkeeping methods that involve maintaining the financial records of monetary transactions and the regular preparation of statements concerning the assets, liabilities, and operating results of some activity. To assure the accuracy of such financial accounts, one well-developed dimension of the accounting profession is auditing. Audits review and examine accounts to determine whether they are reasonably accurate and presented in an understandable manner. The attempt to adapt such methods from the fields of business and finance to those of scientific research is called data auditing (DA), and constitutes a special effort to assure accountability in research.

Historical Background

In the early history of modern natural science the methodological requirement that experimental results be reported in such a way that they could be reproduced by others, and the practice of accepting into the body of scientific knowledge only those results that had been reproduced, effectively made auditing a standard part of research practice. Even so, William Broad and Nicholas Wade (1982) argue that what is now called creative accounting was sometimes practiced in scientific research. For example, there is evidence that physicist Isaac Newton (1642–1727) made experimental data fit his theories, and that chemist John Dalton (1766–1844) cleaned up his data to obtain whole numbers for ratios on chemical reactions. Biologist Louis Pasteur (1822–1895) is alleged to have announced an anthrax vaccine before completing his experiments, and Cyril Burt (1883–1971) may have fabricated intelligent quotient (IQ) test results. Even Nobel Prize winner Robert Millikan (1868–1953) may have fudged his data. Other examples include the fabrication of animal test data by Industrial Biotech Corporation for the Food and Drug Administration (FDA), and the conduct of unethical high-risk experiments on psychiatric patients. In some cases serious adverse events, including deaths, were not reported (Shamoo and Resnik 2003).

A number of surveys indicate that students and researchers suspect that questionable conduct in research is widespread, accounting for 0 percent to 50 percent of all research. The actual percentage of questionable research practices is probably much lower—in the single digits (Shamoo and Resnik 2003, LaFollette 2000). The scientific community was initially slow to call for reforms in dealing with scientific misconduct. In response to media coverage of some serious lapses, commissions were formed and congressional hearings held to discuss accountability in research. Then-senator Albert Gore chaired hearings to examine concerns and urge reforms (LaFollette 1994).

The modern explosion in attainment of knowledge has resulted in profound changes in the social character of science. In 2002, nearly 3 million individuals worked as researchers in the United States alone, with about 1 million holding post-graduate degrees and controlling a budget of more than $250 billion. Science in has become mass science in the pattern of mass production and mass culture. Traditional means of apprenticeship and social pressure are not effective ways to uphold high standards for scientific knowledge in the early-twenty-first century. More explicit approaches must be developed.

It was in this context that the term data audit first began to be used (for complete references on this topic see Loeb and Shamoo 1989). Following a 1988 conference on the subject, the inaugural issue of the journal Accountability in Research announced its intention to "serve as a catalyst for the development of specific procedures and standards for acquiring, analyzing, and auditing" (Shamoo 1989, p. i).

DA Theory and Practice

The concept of auditing has a long history, including efforts in early Egyptian, Greek, and Roman civilizations by governments to develop ways to expose cheating by accountants. Modern accounting and auditing procedures have their immediate origins in response to the enormous expansion in business enterprises since the nineteenth century.

There are several kinds of auditors. External auditors are independent auditors who work in public accounting firms for identified clients. However because third parties use the information in the financial statements generated by these auditors, external auditors can be said to work also in the interests of society. Internal auditors work as employees within organizations; public and private corporations and government agencies have internal auditors. Government auditors are employed by government agencies to audit outside entities and individuals. An example of a government auditor is the Internal Revenue Service (IRS).

In addition, there are various types of auditing. Financial auditing examines the accuracy of an entity's financial statements. The resultant report can be used inside or outside the entity. Operational or performance auditing examines performance, management, or value-added operations, including cost-economy, efficiency, and effectiveness. Compliance auditing examines whether an organization is in compliance with specific rules and regulations, whether issued internally or imposed on the entity by a third party. Attestation engagements are given to public accounting firms for the purpose of examining the representations of an entity other than those that are traditionally included in financial statements, for example, those regarding systems of internal accounting control or investment performance statistics (Loeb and Shamoo 1989).

Auditing is an independent activity that reviews accounting, but is separate and apart from it. Its methods rely on logic, not accounting principles, to evaluate concrete issues. DA, as proposed by Adil Shamoo in the late 1980s, is modeled after financial auditing. The purpose of DA is to check the accuracy of derived research data by comparing it to the original raw data. This method can be used either randomly for a small number of data determined by a statistical method or when the data are suspect. Several publications have outlined the method since its initial introduction (Shamoo and Annau 1987).

The Future of DA

Accountability in research requires reviewing institutional policies (for example, those of universities) and examining the attitudes and behavior of researchers. Institutional policies are key because they dictate the tone and culture of tolerance in research conduct and are major influences on how and why researchers work on particular issues (Shamoo and Dunigan 2000).

Society demands accountability from researchers. This is especially true when the results of particular research affect individuals and communities. In the early-twenty-first century, accountability in research is an important and expanding area of interest to both professionals and the general public.

ADIL E. SHAMOO

SEE ALSO Misconduct in Science;Peer Review;Research Ethics.

BIBLIOGRAPHY

Broad, William, and Nicholas Wade. (1993 [1982]). Betrayers of the Truth: Fraud and Deceit in the Halls of Science. New York: Simon and Schuster.

LaFollette, Marcel C. (1994). "The Pathology of Research Fraud: The History and Politics of the U.S. Experience." Journal of Internal Medicine 235: 129–135.

LaFollette, Marcel C. (2000). "The Evolution of the 'Scientific Misconduct' Issue: An Historical Overview." Experimental Biology and Medicine 224: 211–215.

Loeb, Stephen, and Adil E. Shamoo. (1989). "Data Audit: Its Place in Auditing." Accountability in Research 1: 23–32.

Shamoo, Adil E, ed. (1989). Accountability in Research 1(1). Inaugural issue of the journal; contains complete references.

Shamoo, Adil E., and Zoltan Annau. (1987). Correspondence (Letter to the Editor), June 18, 1987. Nature 327(6123): 550.

Shamoo, Adil E., and Cheryl D. Dunigan. (2000). "Ethics in Research." Experimental Biology and Medicine 224: 205–210.

Shamoo, Adil E., and David B. Resnik. (2003). Responsible Conduct of Research. Oxford and New York: Oxford University Press.