Six years ago Congress passed the little-known Data Quality Objectives Act (DQOA). According to the Newsletter I just received from the Maryland Environmental Law Program and an article in the Maryland Bar Bulletin, "[t]he DQOA is a two-sentence mandate snuck into the fiscal year 2001 federal appropriations bill as a rider. It simply requires the Office of Management and Budget (OMB) to adopt mandatory guidance requiring each federal agency to adopt formal guidelines to ensure that information 'disseminated' by that agency meets certain data-quality criteria for 'quality, objectivity, utility, and integrity'." But it seems data quality is in the eye of the beholder. Industries, at least in the environmental context, are challenging state data and even potentially inaccurate spreadsheets under federal law.
The article linked to above asks, "Should environmental regulations be based upon a solid scientific foundation and reliable data before they become final and have the 'force and effect' of law?" This, however, begs all sorts of questions: What data is good enough? Should agencies be allowed to regulate in the face of uncertainty where the risk is unknown but the potential harms may be severe? What processes create good data in administrative agencies? OMB and EPA guidelines for ensuring quality data can be found here and here.
I'd be interested if anyone knows if the Data Quality Act has been litigated in other contexts besides environmental law.