In DOE, factors, blocks, and design resolution are determined during experiment planning, before data are collected. Factors are the independent variables selected for study, blocks are chosen to account for nuisance variation, and resolution is part of the experimental design selection. The CSSBB supplemental material explains that factors and blocks are design elements established in advance, while resolution reflects the level of confounding and is also chosen as part of the design setup. By contrast, the estimate of variance depends on the observed experimental results. The CSSBB materials describe the need for “an estimate of the inherent error variation” and place this in the context of replication, power, and post-data understanding of noise in the system. That variance estimate comes from the collected response data and is not fully known until the experiment has been run. This is fundamental in Analyze Phase thinking, because the team uses the experimental results to quantify error, assess significance, and separate signal from noise. Therefore, among the listed choices, the only item that can be determined only after data collection is the estimate of variance.

===========