Collection, Analysis, and Management of Data
Proper management of research conduct is essential to achieving reliable results and maintaining the quality, objectivity, and integrity of research data. The different steps of research should be monitored carefully, and research design should include built-in safeguards to ensure the quality and integrity of research data. This chapter addresses ethical conduct in different steps of the research process: hypothesis formation, research design, literature review, data collection, data analysis, data interpretation, publication, and data storage. This chapter also discusses methods that can help assure the quality, objectivity, and integrity of research data, such as good research practices (GRPs), standard operating procedures (SOPs), peer review, and data audit.
Scientific research is the systematic attempt to describe, explain, and understand the world. While all three main branches of science—physical science, biological science, and social science—study different aspects of the natural world, they share some common methods and procedures. These methods and procedures are designed to achieve the goals of science by helping researchers to acquire accurate knowledge and information. Researchers' compliance with the scientific methods and procedures will minimize falsehoods and biases and maximize truth and objectivity (Cheny 1993). One pillar of the scientific method is the idea that researchers should subject their theories and hypotheses to rigorous tests (Popper 1959). A test is an attempt to gather empirical evidence (or data) that tends to either confirm or disconfirm a theory or hypothesis. Ideas that cannot be tested, such as metaphysical theories or ideological claims, are not scientific hypotheses or theories. Some (but not all) tests involve experiments. In an experiment, a researcher attempts to control the conditions of a test in order to understand statistical or causal relationships between variables or parameters. For an experiment to be rigorous, a researcher must describe it in enough detail that other researchers can obtain the same results by replicating the experimental conditions (Kirk 1995).
Repeatability is important in experimentation because it confirms that others can carry out the methods and procedures used and attain the same data. Repeatability, or lack thereof, provides substance for public debate and inquiry. Private intuitions, hunches, faith, introspection, or insight can play an important role in generating new ideas to test, but they do not constitute rigorous proof. Therefore, all test results in science, whether from controlled experiments, field observations, surveys, epidemiological studies, computer models, or meta-analyses, should be open to public scrutiny and debate. Peer