U.S. Department of Commerce

X-13ARIMA-SEATS Seasonal Adjustment Program

You are here: Census.govSubjects A to ZX-13ARIMA-SEATSSeasonal Adjustment PapersPapers by Year › Abstract of Findley, Wills, Aston, Feldpausch, and Hood (2003)
Skip top of page navigation

Diagnostics for ARIMA-Model-Based Seasonal Adjustment

David F. Findley (1), Kellie C. Wills (2), John A. D. Aston (3), Roxanne M. Feldpausch (4) and Catherine C. H. Hood (5)


We present two types of diagnostics for SEATS and similar programs. We start with modifications of the diagnostic used by SEATS to detect underestimation or overestimation, meaning inadequate or excessive supression of frequency components near the target frequencies, e.g. the seasonal frequencies in the case of seasonal adjustment. The modifications use time-varying variances of the finite-length filter output instead of the constant variance associated with the infinite-length filter. The SEATS diagnostic is shown to be substantially biased toward indicating underestimation even when the estimation is optimal, a situation where the modified diagnostics are unbiased. The second type of diagnostic considered is an adaptation of the widely used sliding spans diagnostic of X-12-ARIMA. The adaptation is a method for determining the span length appropriate for model-based-adjustment as a function of the ARIMA model's seasonal moving average parameter.


Residual seasonality, spectrum, SEATS, underestimation, overestimation, sliding spans

(1) David F. Findley was the Senior Mathematical Statistican for Time Series and is now a consultant, U. S. Census Bureau, 4600 Silver Hill Road, Washington, DC 20233. email : david.f.findley@census.gov

(2) Kellie C. Wills is currently a doctoral student.

(3) John A. D. Aston is currently at the University of Warwick. email : j.a.d.aston@warwick.ac.uk

(4) Roxanne M. Feldpausch is currently a consultant.

(5) Catherine C. H. Hood is currently a consultant. email : cath@catherinechhood.net

[PDF] or PDF denotes a file in Adobe’s Portable Document Format. To view the file, you will need the Adobe® Reader® Off Site available free from Adobe.

This symbol Off Site indicates a link to a non-government web site. Our linking to these sites does not constitute an endorsement of any products, services or the information found on them. Once you link to another site you are subject to the policies of the new site.

Source: U.S. Census Bureau | Center for Statistical Research and Methodology | (301) 763-1649 (or x12@census.gov) |  Last Revised: November 19, 2012