Precision concerns both the reproducibility and the carefulness of the measurement. An alcohol thermometer with a wide bore and markings that are very close together can not be read as precisely as an alcohol thermometer that has a narrow bore and markings that are further apart. If temperature is controlled carefully, the temperature may be more accurately maintained than the precision of the measurements with the wide bore thermometer. Precision depends on both the scale (precision) of the measuring device and the reproducibility of the measurement.
Accuracy is much more difficult to assess than precision or reproducibility. Accuracy is how close the result is to the actual value. For instance, the concentration of a solution is measured by measuring the color. Three measurements with the same spectrometer may be very precise, but the value may not be accurate if the standard was not prepared properly. Any number of errors in sample preparation or standard preparation may result in inaccurate but precise data. Instrumental calibration is critical in making accurate measurements.
In the image below, the target at the left implies accurate, precise data. The middle target may have an accurate average, but the points imply imprecise data. The target at the right implies precision without accuracy.
Laboratories prepare standards for assessment of the accuracy of methods in use. To provide a method for assessing the accuracy of methods between laboratories, NIST (National Institute of Standards and Technology, formerly the National Bureau of Standards) sells standards which cover a large range of samples. Even in companies where outside agreement is not important, these NIST products supplement in-house preparation of standards for additional checks of accuracy. NIST maintains a web site for analytical chemistry databases.
