It is common practice for most commercial systems to collect each quadripole several times in a row immediately one after another and average the results. This procedure is referred to as stacking. Although collection of stacked measurements nominally increases the duration of the survey, stacking improves the signal-to-noise ratio because noise, if random, is averaged out. In addition, the standard deviation of the stacked measurements (i.e., the stacking error) provides a means to quantify error and define data weights for inversion (although reciprocal measurements described next are preferred). Stacking errors are useful in QA/QC and can be used to weight data in inversion or form a basis for editing datasets prior to inversion. The number of stacked measurements should be recorded on field data collection forms if not recorded by the software. On some instruments, the number of stacked measurements is determined by calculating the running stacking error, with more stacks collected for measurements with larger error.