FAQ
FREQUENTLY ASKED QUESTIONS
Why analyze not summarize?
aLocal’s statistical model uses secondary US Census data focused on Zip Code Tabulation Areas (ZCTA’s) for variables related to NAICS code summaries regressed to demographic and spatial datasets. These variables are grouped using Principal Component Analysis (PCA) and then re-regressed with the dependent variables and PCA regression variables with eigenvalues greater than 1. Summary analyses are then tested for multicollinearity, and heteroskedasticity, whose results are then treated using additional control variables.
The final model is ultimately applied as a predictive formula to existing ZCTA variables, providing the user with a snapshot projection of which geographic regions should be avoided, given poor prospects, as well as those which show net positive demand and promise for potential market success.