Hypothesized Pathway and Bias

Abstract

Science proceeds by a progressive process of posing scientific hypotheses and conducting experiments to test them. When this process is directed toward a treatment to improve human health, it is called a clinical trial. The great philosophers of science argued that a good scientific hypothesis “sticks its neck out” with a quantified prediction which can only be tested in an experiment that is free of bias. They further argued that failure should be welcomed as an opportunity to refine and progress, and replication is the best route available to obtain conclusive evidence that a treatment works. In behavioral clinical trials, the scientific hypothesis is translated as the hypothesized pathway. It is the quantified path by which a behavioral treatment is hypothesized to improve a behavioral, biomedical, or chronic disease endpoint. It guides the design of the trial, the interpretation of results, and the next step in a progressive, translational program of research. A trial is inconclusive when it falls prey to any of a number of known biases. Because a behavioral trial synthesizes a behavioral treatment with a health endpoint, biases identified in behavioral sciences, medicine, epidemiology, and biostatistics can all pose potential problems. Mastery of this complete range of potential biases will promote the rigor and conclusiveness of any single behavioral trial.

“A proper theory sticks its neck out”

Lewens, 2016 [1].