Making Sense of Biostatistics: Types of Probability Sampling

Sampling methods have important implications for the quality of research results. Ideally, a clinical research study could draw on a database of all potentially eligible patients. However, clinical studies are usually limited to interested patients drawn from site populations and advertising programs, a small fraction of the total population that may, further, be biased in a variety of ways. Nevertheless, this column will discuss three variations on the ideal case as a baseline for the next column, which will deal with the real world of clinical research.