When you look over the printed output, pay particular attention to the following:
1. Rejected stars: These are marked with an asterisk (*) in the right margin when the residuals are listed. An occasional reject is normal; but clumps of rejected observations are not. If every observation of a standard star is rejected, it may be misidentified. If the standard values of a standard star are all rejected, there may be a catalog (or copying) error; or the wrong star may have been observed.
Abnormally faint observations of program stars are often due to the dome being in the way. Be careful to check the dome slit while observing, if it is not controlled automatically. Another cause of abnormally faint star readings is confusion of star and sky measurements (see item 4 below).
2. Reading plots: Remember the conventions used in low-resolution plotting: the $ symbol marks overlapping points; ^ and v mark points beyond the upper and lower edges of the plot (think of them as arrowheads). Some plots have fitted lines indicated by a series of dashes. The residual plots identify different stars with different symbols; these are given in the tables of residuals.
3. Trends in residual plots: Similar trends in the run of residuals with time in different bands usually indicate either instrumental instability (zero-point drift) or varying extinction. Instrumental drifts are usually a function of temperature and/or relative humidity; but a bad high-voltage supply can produce irregular variations. Extinction variations are usually larger at shorter wavelengths, and may show short-lived dips as wisps of cirrus cross the sky.
4. Observations with negative intensities: These can be caused by star observations misidentified as sky, or vice versa. Check your data very carefully for errors. Ordinarily, sky data should be much fainter than the star measurements, for stars brighter than about magnitude 15. If you have fainter program stars than this, use a CCD to obtain the considerable benefits of simultaneous measurements of star and sky. The sky is considerably brighter in the infrared, or during auroral displays. Here, large fluctuations in sky brightness can occur, occasionally producing negative star intensities after sky subtraction. If your sky fluctuations are appreciable, compared to your faintest program star, be sure to chop back and forth quickly between star and sky to subtract the fluctuations accurately.