The purpose of this review is to discuss the use of the scientific method of problem-solving in the development of an experimental design. It reviews the basic concepts that allow us to evaluate data statistically. Common problems that can arise from inadequate experimental design, investigator bias, or inappropriate statistics are also mentioned.
The actual statistical methods are not discussed in detail-such a description can be found in any applied statistics text. Moreover, statistical software packages are available for computers of all sizes and, if used appropriately, can be very helpful to the medical investigator. Several of the most commonly used mainframe software packages are referenced at the conclusion of this review. Most computer vendors can provide information concerning statistical software packages for microcomputer application.
USE OF THE SCIENTIFIC METHOD IN MEDICAL RESEARCH
Employment of the scientific method in medical research can help maximize the beneficial effects of our research efforts. Although the steps of this procedure may seem natural to the experienced investigator, a review of these basic principles can help prevent inappropriate studies.