X hits on this document

303 views

0 shares

0 downloads

0 comments

47 / 108

SECTION 3: DATA COLLECTION, FIELD OBSERVER TRAINING AND STATISTICAL METHODS (PHASE 3)

This Section of the Report Covers:

3.1

Determining a Reasonable Sample Size for Drawing Statistical Conclusions about the SQG Group

3.1.1

Sample Size Needed to Benchmark an Individual State’s Sector Performance

      • 3.1.2

        Sample Size Needed to Compare Performance Levels between States

    • 3.2

      Universe Identification and Random Sampling

    • 3.3

      Steps to Ensure Field Data Quality

    • 3.4

      Receipt and Analysis of Collected Field Data

      • 3.4.1

        The ERP Performance Analyzer

      • 3.4.2

        Quality Assurance Procedures for Data Analysis and Reporting

    • 3.5

      Observations and Lessons Learned from Phase 3

The third phase of the project involved applying statistical and data quality assurance procedures to sample selection and data collection. There were three general data collection objectives for Small Quantity Generators of Hazardous Waste (SQGs):

  • o

    Collect data on enough SQGs to be able to say that their performance reflects the performance of all of the SQGs in the state with enough precision and reliability for the results to be useful to state decision makers.

    • o

      Pick facilities randomly so as to not introduce bias into the results.

    • o

      Collect the data in the same way with the same interpretations so the results are comparable from state to state.

3.1 DETERMINING A REASONABLE SAMPLE SIZE FOR DRAWING STATISTICAL CONCLUSIONS ABOUT THE SQG GROUP

The States Common Measures Project had two primary analytical goals: 1) benchmarking SQG performance in each individual state and 2) comparing SQG performance across states to identify performance differences that do not appear to be due to chance. Since states could not inspect every SQG, statistics had to be used to draw conclusions from inspections of a sample of each state’s facilities. The sample had to be large enough to provide enough certainty that the observed results reflected actual conditions, precise enough to provide the states with useful information, but small enough to be done with available resources.

The States Common Measures Project Final Report

47

Document info
Document views303
Page views303
Page last viewedSun Dec 11 00:33:05 UTC 2016
Pages108
Paragraphs3821
Words29990

Comments