ABSTRACT
In the preceding chapters I discussed the ways in which categories are defined, constructed, sampled, recorded, and measured. Measures, as I noted, are derived in part from the theories guiding the research. Among the most important attributes of measures are their technical properties, or the extent to which they are reliable and valid. Reliability and validity are necessary in the scientific enterprise as they help to insure that measures are both consistent, and measure what it is they purport to measure. While reliability and validity can be applied to any number of measures, such as mental test scores, behavioral observations, physiological measures, and self report inventories, they are most often discussed in the context of paper-and-pencil tests. Indeed, much of the theoretical and empirical work in these areas has been in the context of mental testing. In this chapter I will draw from this work to inform the discussion of reliability as it is applied to behavioral observations.
