Is the data sample size an issue for the project?  I've done mine based on some trend work I've done for my company by state but it's essentially 16 data points, with each one using 5 years of data to form a trend.  It may not be enough data to evaluate predictability but I'd think it's enough to evaluate my process.

[NEAS: Data size is rarely an issue. We check if you understand the methods, not how credible the data are.]