Causal Analysis/Diagnosis Decision Information System (CADDIS)

• Data Analysis Methods
• Fundamentals of Data Analysis
• Get Data Analysis Tools

DA.2. Assuring Data Quality

The quality of each type of data collected and used during the causal analysis process will need to be evaluated. The main types of data used during a causal analysis include data from the case, data from elsewhere, and data produced from models. Ultimately, you must decide whether there is a sufficient quantity of each type of data meeting an acceptable level of quality to demonstrate the likely cause of the biological impairment.

At the start of a causal analysis, we recommend that you incorporate Elements of Systematic Planning to identify data needs and acceptance criteria. When considering data needs, it is helpful to classify the data you have in terms of their role in fulfilling conceptual model pathways. Although methods of data quality analysis and the standards for acceptable data may vary among data types and causal analyses, the most important part of any data quality review is the written record of the activities and analyses you conducted and the decisions you made.

Observational data — To determine whether observational data are adequate for the needs of the case, it is helpful to develop checklists and standard operating procedures to guide the data review. Steps in the evaluation of the appropriateness of any data include:

• Determining if there are non-quality constraints on the use of data in the case,
• Conducting environmental data verification and validation to ensure that data are appropriate and consistent with the intended use, and
• Conducting a data quality assessment examining variability, uncertainty, sample size, data gaps, etc.

Data produced from models — The quality of data produced by a model can be analyzed by following a process that documents the model outputs, the results of assessing these outputs, and the need to place caveats on the use of those outputs. In addition, project-specific documentation and reporting requirements must be met (e.g., peer-review of theory and equations; documentation of changes to the model, assumptions, theory, and parameterization). A detailed process for developing models or selecting pre-existing models to generate quality data can be found at Guidance for Quality Assurance Project Plans for Modeling (PDF) (121 pp, 615K, About PDF).

Additional Information

A wide range of comprehensive data quality resources that can be useful to any causal analysis, including information specific to non-EPA organizations, can be found at U.S. EPA's Quality System for Environmental Data and Technology. What follows is a list of some of those resources:

• Elements of Systematic Planning
• EPA Requirements for Quality Assurance Project Plans (PDF) (40 pp, 121K)
• Data Quality Assessment Tools
• Guidance for Quality Assurance Project Plans (PDF) (111 pp, 402K)
• Guidance for Geospatial Data Quality Assurance Project Plans (PDF) (106 pp, 1.4MB)
• Guidance on Quality Assurance Project Plans for Secondary Research Data (PDF) (2 pp, 7K)
• Guidance on Systematic Planning Using the Data Quality Objectives Process (PDF) (100 pp, 333K)
• Guidance for Data Quality Assessment: Practical Methods for Data Analysis (PDF) (219 pp, 1.7MB)
• Using Data From Other Sources - A Checklist for Quality Concerns (PDF) (6 pp, 142KB)
• Guidance on Choosing a Sampling Design for Environmental Data Collection (PDF) (178 pp, 1MB)
• Guidance on Environmental Data Verification and Data Validation (PDF) (96 pages, 387K)
• Resources for Planning New Data Collections

You will need the free Adobe Reader to view some of the files on this page. See EPA's PDF page to learn more.