Before calculating the various effects, one must first run a set of experiments. These experiment sets
may be “full factorial”, which include all combinations of factors and levels, or they may be “fractional
factorial”, which include a smaller number of experiments based on an orthogonal matrix of level
combinations. The feasibility study [6] considered only five factors, so a full factorial set of
experiments was possible. For the present study, the number of factors is increased to ten, so a
fractional factorial design is used to reduce the number of necessary experiments per façade. The
resolution of a DoE scheme refers to the quality of information obtained by the set of experiments,
specifically the amount of possible confounding with higher order interaction effects. For example, the
main effects of a Resolution III or Resolution IV design may be confounded with two-level or threelevel
interaction effects, respectively, and all higher-order interaction effects will also be confounded.
A Resolution V design results in non-confounded main effects and two-level interaction effects.