Chapter 74. Principles of Statistical Methods for Research in Regional Anesthesia
A. The field of applied mathematics that analyzes data to prove a specific hypothesis
B. The study of variability
C. The field of applied mathematics that analyzes data to know what the minimum sample size is to enroll in research studies
D. The field of applied mathematics that should only be used to analyze data collected in large surveys and research studies
B is correct. It is seldom feasible to study an entire population, so samples taken from the population are studied. But each sample will be slightly, moderately, or vastly different from the next sample.
A is incorrect. We may find a result that is anticipated (given current knowledge in the field)—or—we may not (a surprise for everyone, including ourselves)! We do not set out on a research study with the intention of “proving” that we are correct.
C is incorrect. Estimates of a feasible and reasonable sample size are needed for inferential studies that test hypotheses. Nonetheless, there are descriptive studies in which sample size calculations are not needed.
D is incorrect. While data from large surveys (eg, national political opinion polls) are frequently collected and analyzed—particularly, in an election year—small surveys and research studies can also be important to evaluate.
Which statement is true regarding basic study designs?
A. The randomized clinical trial (RCT) is the gold standard of research designs, and thus should be used whenever possible.
B. When deciding on which study design to use, choose the one that involves the most treatment arms as this helps the statistician to stratify the outcome data to control for confounding.
C. When deciding on which study design to use, choose the one that collects the most variables as this enables the statistician to analyze the data with the more powerful, next-generation statistical approaches.
D is correct. Every study design has its pros and cons. Study designs that are overly complex are difficult to conduct and their data may be too entangled for clear interpretation.
A is incorrect. The hallmark of the RCT is that the investigator manipulates the exposure or treatment. Clearly, this is not always ethical—even if subjects are randomly allocated to exposure or treatment “arms.” Women who are pregnant cannot be forced to smoke or, conversely, not to smoke, after they have been enrolled in a study of ...