Often, the phenomena we are trying to measure involve abstract constructs (e.g., cognitive dissonance) that are captured with written responses or observed behaviors. Developing unobtrusive or indirect measures of already abstract constructs can be difficult, but necessary. In a laboratory setting, how could you use unobtrusive measures to assess the extent to which a person is currently experiencing the discomfort of cognitive dissonance? Would this laboratory-based, unobtrusive measure of dissonance be equally valid in field settings—why or why not? In representative field settings (i.e., contexts with high ecological validity), is there a way you might be able to indirectly or unobtrusively determine whether people were in a dissonant state?