Observe people’s actions. Sometimes these obser- vations are very simple, as when a researcher notes which of two items a person selects. At other times, however, the observations are more elaborate, as when judging whether someone is acting warmly or coldly toward another person, and require that interrater reliability be established. Interrater reliability refers to the level of agreement among multiple observers of the same behavior. Only when different observers agree can the data be trusted.
The advantage of observational methods is that they avoid our sometimes faulty recollections and distorted interpretations of our own behavior. Actions can speak louder than words. Of course, if individuals know they are being observed, their behaviors, like their self-reports, may be biased by the desire to present themselves in a favorable light. Therefore, researchers sometimes make observa- tions much more subtly. For example, in experiments concerning interracial inter- actions, researchers may record participants’ eye contact and seating distance to demonstrate biases that would not be revealed using more overt measures.