Studying dream meaning is difficult due to methodological pitfalls, including vague and inconsistent definitions of dream content variables used as independent measures. When patients were allowed to repeat their dream and report associative content, the number of initial references to death and separation variables nearly doubled compared to a standard single report of manifest content. There was minimal or no correlation of these variables between manifest and associative content, showing associative content's unique contribution. A new interview technique, the Staged Interview Technique, is described to control and measure these potential distortions.
Using dream content as an independent variable to study dream function requires consistent definition of that content. A dream interview technique was described that controls for distortion from associative content and repeated reporting. Dream reports obtained with this technique were rated for references to death, mutilation, separation, illness, guilt, shame, and falling. The presence or absence of all seven variables varied considerably depending on how the dream report was defined. To use dream content as a predictor in dream function research, associative content and repetition of the dream report must be systematically controlled.