Consequently, a more direct decision process involves:
1. Listing or ranking the feasible and most desirable data input dimensions, considering availability and economy, while recognizing that better data usually costs more in both money and time
2. Listing or ranking the most desirable output dimensions and characteristics
3. Determining the most desirable and feasible Input-to-Output matchings that survives a cost/benefit evaluation
4. Matching the feasible Assumptions-to-Process connections that are compatible with the Input-to-Output dimensions
5. Constructing other options and then selecting the most promising feasible profile.