190 likes | 271 Views
Delve into the 2004 review of sampling methods in extrapolated new base-year studies, highlighting comparisons between extrapolated and nonextrapolated approaches, addressing concerns, and proposing improvements. Analysis of data accuracy, possible errors, and suggestions for efficient future studies are discussed.
E N D
Overview of the Final Report and Findings from the Review of Sampling Methods in Extrapolated New Base-Year Generation Studies May 11-12, 2004
Types of New Base Year Generation Studies • Nonextrapolated • “Actual” diversion data collected from diversion programs implemented within the jurisdiction • Extrapolated • Includes diversion data collected from a sample of the non-residential businesses and extrapolated to the entire non-residential business sector
Number of New Base-Year Generation Studies • Since April 1998, Board staff have reviewed 184 submitted new base year studies for Board consideration. • 143 Nonextrapolated studies • 41 Extrapolated studies
Background • Board expressed concerns regarding validity of extrapolated studies • Working group convened • Board staff performed analyses on extrapolation vs. nonextrapolation study methodologies
Extrapolated vs. Nonextrapolated Study Data Comparisons • Generation pounds per person per day was 79% higher for extrapolated methodologies • Diversion rates for extrapolated studies were significantly higher (53%) than nonextrapolated studies (41%). • Average source reduction tonnage in extrapolated studies exceeded nonextrapolated studies by an average of 300%.
Response to Concerns regarding New Base-Year Study Data • Base-Year Modification Certification Request form • Board staff began on-site verification of new base-year generation study data • Board pursued independent, third-party review of extrapolated study methodologies
Scope of Review • Review 20 extrapolated studies. • Only 2 of the 20 studies reviewed submitted sufficient data. • Provide observations and recommendations for improving extrapolated methodologies used. • Impact of poor quality or lack of study data.
Findings • The findings in the final report address the impact of errors relating to: • Study design; • Sampling errors; and • Technical data components of nonresidential diversion surveys.
Study Design • Jurisdictions should list assumptions made in the study design and use sample data to assess their validity and appropriateness. • Studies did not include listing of such assumptions. • Jurisdictions did not use the study data to test the validity of assumptions.
Sampling Errors • Accurate data on the selection of samples is needed for Board staff to fully evaluate the sampling methodology. • Lack of documentation regarding sample selection. • Studies employed nonrandom sampling methods.
Technical Data • Lack of documentation regarding nonresponse. • Most jurisdictions provided outlier analysis, but did not address all study design variables. • Lack of access to study data.
Additional Analyses • Case Study of One Extrapolated New Base-Year Generation Study • Review of 35 extrapolated new base-year generation study data • Diversion of Top 10 Businesses Compared to All Surveyed Businesses
Commercial Business diversion Represents Approximately 38% of Total Diversion 25% Residential NonResidential 37% Business Audits Res/ADC/Landfill Salvage 38%
New Base Year Case Study: Impact of the Largest Businesses (333) Audited Relative to the Total Business Audits (approx. 7,600) 90 80 70 60 50 % of Diversion tonnage from Businesses Audits 40 30 20 10 0 top 10 top 20 top 30 Business Generators
Diversion of Top 10 Businesses Compared to All Surveyed Businesses • By surveying 10 top businesses: • 94% of the jurisdictions captured 50% or more of the total diversion; • 60% of the jurisdictions captured 80% or more of the total diversion; and • 37% of the jurisdictions captured 90% or more of the total diversion.
Maximizing Efficiencies in New Base-Year Generation Studies • Many of the jurisdictions would have been better off targeting their largest diverters and submitting a nonextrapolated study. • These jurisdictions could have captured most of the industrial/commercial diversion without having to address problems associated with sample selection, nonresponse, outliers, or choosing an estimator.
Future Efforts • Board staff will: • Update the Board’s Base-Year Modification Request Certification form for extrapolated studies. • Update the Board’s Diversion Study Guide to explicitly underscore the importance of documenting all critical aspects of study design, implementation and diversion rate calculation. • Develop a Web-based reporting format.
Future Efforts (continued) • Board staff will continue to • Assist jurisdictions regarding their future new base-year generation studies; and. • Stress the benefits of targeting the largest businesses first to measure the non-residential sector.
Summary • Board staff were unable to perform a complete statistical review for 18 of the 20 studies due to insufficient study data. • Findings support the value of targeting the top 10 businesses.