First-person reports seem to refer to conscious contents and pose a problem for behaviorism. According to logical behaviorism, the meaning of a first-person report consists only of the publicly observable conditions for verifying it, even though the speaker may not have to observe those conditions in order to make the report. According to operationism, first-person reports are merely verbal responses, and "sensation" is operationally defined as that which is measured by a set of publicly observable procedures.
Many behaviorists interpret first-person reports as verbal responses to discriminative stimuli. These stimuli may consist of overt behavior, environmental variables responsible for behavior, or covert events. Even under this interpretation, first-person reports are not scientific observations but are rather verbal behavior from which theoretical concepts such as intervening variables and hypothetical constructs can be inferred. One reason for not including first-person reports as scientific observations is that they are unreliably linked to their stimuli.
Behaviorists also interpret the first-person reports of "psychological measurement" which purports to measure "psychological," or "subjective" magnitudes. Under an intervening-variable interpretation, psychophysical experiments are like defining experiments in which the experimenter, not the subject, assigns a value to an intervening variable. The behavior of the subject can be explained as the transfer of verbal behavior from situations in which intersubjective rules are defined to a situation in which they are not. Under a hypothetical-construct interpretation, psychophysical judgments can be construed as responses to covert events. Nevertheless, it is an error to view the subject, rather than the experimenter, as measuring these covert events.
These interpretations of psychophysical scaling can be extended to cover a large body of research on "subjective experience." In general, these experiments can be understood as studying how subjects follow their "inclinations" in transferring intersubjec