Skip to main content
Water Quality Monitoring Across the United States
STARMAP (Space-Time Aquatic Resources Modeling and Analysis Program) - Department of Statistics at Colorado State University
Main menu Section menu Feedback Next

		How to Summarize

Verifying and Validating the Data. page 1 of 10

    A statistical summary is needed to provide information to meet the survey objectives. The design of the survey determines the way you analyze the data. They are not independent of each other. Communicating results is the last important step in this process. Reports, graphics, and tables can be used to summarize the results of a survey, but care must be used in the presentation. Improper use of graphics can alter the perception of differences between subpopulations, or regions. A poorly written report can have a detrimental effect on a carefully planned survey.

    However, the data needs to be verified and validated before any serious data summary should start.  Although the verification and validation process may seem to possess similar attributes, the difference between them can be seen when understanding the individual processes. Data verification is the first of the two evaluations to occur, being performed by all personnel.

    The ultimate goal of data verification is to ensure and document that the reported results reflect what was actually done.  Although there may be a need for an external data verifier, that does not relinquish the responsibilities of the sample collectors and laboratory analysts to verify their own data.  Data verification should be occurring at each level, by multiple personnel.  The data verification process tends to be predominantly an internal practice for the field and laboratory staff; however, depending on the project, it is not uncommon to have an external verifier.

    Whether or not the current verifier is an external or internal member, it is crucial that they are familiar with the programs requirements. Understanding the purpose of the study and all protocols will ensure that the data verification procedure will be effective.In addition to understanding the material, it is important to have all protocol documentation easily accessible.

    Developing environmental data of any kind involves the production of documentation and records, from daily field logs to electronic records; data verification is no different. However, first, the verifier should identify the source and location of all records produced, including all calculations performed on the input data.  In addition, the verifiers should determine the criteria or specifications against which the records will be compared.  The records obtained may be produced by multiple personnel with different formats, although they should be following the records-keeping procedures, the verifier should review the formats and get comfortable with the individual styles.

    The five steps to be verified are the: sample collection, sample receipt, sample preparation, sample analysis, and records review, and are discussed on subsequentt pages.

Main menu Section menu Feedback Section Menu Next