The Multi-Evidence Paradigm as an Evidential System

Evidential systems describe how decisions / determinations are made and by whom. In the Statistical Causal Inference Paradigm, the evidential standard is a credible estimate of a bias-free causal effect. The evidential process involves statistical conditioning. Both the standard and process are determined by statisticians. Mechanistic scientists play no role in creating the system or determining its capabilities.

In the Multi-Evidence Paradigm, the evidential standard, processes, and deciding entities will depend on the evidence being used and the research situation (e.g., risk-reward considerations) - there is no single standard or process for all cases.

If mechanistic knowledge is considered, scientific experts must be involved in the process. Evidential standards will vary depending on the situation but will include, (1) preponderance of evidence and/or (2) progress in building causal knowledge. Many sources of information can contribute, including mechanism-observation consistency, theoretical analyses.

It will be very helpful if investigators will declare the evidential system they are using, acknowledge the possible choices, and justify the choice made. Again, starting a study with a consideration of existing causal knowledge will maximize our chances for building causal knowledge, permit integrative causal analysis, and help to clarify the literature for readers. Details of what works best will likely develop as more empirical cases are considered.