The U.S. Geological Survey (USGS) collects water-quality information daily to monitor the Nation's water resources. Water-quality analysts collect water-quality samples for laboratory analysis and make field measurements of alkalinity, pH, and specific conductance. All USGS personnel who make these field measurements are strongly encouraged to participate in the National Field Quality Assurance (NFQA) Project. Contract and cooperator personnel who collect these field measurements to be used in USGS reports or stored in the USGS National Water Information System (NWIS) are also encouraged to participate in the NFQA Project.

In March 1979, U.S. Geological Survey began the NFQA Project. The project is designed to monitor the proficiency of alkalinity, pH and specific conductance measurements performed by USGS water-quality analysts. Initially, the project assessed only the performance of pH and specific conductance measurements; alklinity was added to the project in 1984.

The two specific objectives of the NFQA Project are to provide precision data for the field measurements and to identify water-quality analysts who need additional training. Annual proficiency samples are distributed to individuals who determine alkalinity, pH and specific conductance in the field. The data are summarized and the most probable value (MPV) is determined from all the data. No distinction is made between USGS measurements or data provided by the contract or cooperator personnel when determining the MPV. After the samples are analyzed and the results returned, a proficiency report is prepared by the NFQA manager and submitted to the appropriate USGS offices for review. An annual proficiency report rates values reported by the water-quality analysts as satisfactory, marginal or unsatisfactory.

The criterion used to describe the concentration of the data about the median value is the fourth-spread. The fourth-spread measures the data range of the middle 50 percent of the values (Hoaglin et al., 1983). The median value is the central value in the data set. The bottom 25 percent and the top 25 percent of the values are considered as outliers. The fourth-spread values are presented to NFQA Project customers.