Side 77 · Research substrate
Scientific publications
Experimental design determines what comparison is created before statistics are computed. This layer emphasizes allocation, preregistration, power, measurement and the limits of generalization.
—publications
—direct
—adjacent
—methodological
How to read this layer.
Analysis cannot fully rescue a weak design. Randomization protects some causal comparisons, but power, measurement quality, protocol transparency and population scope still constrain what an experiment can support.
DirectAdjacentMethodological
Mass rollout · Wave 2. Mappings are selective and designed to expose the strongest research substrate for this Side.
0 of 0 publications shown
Newest first · links open source records