Skinner (1953) defined attention behaviorally as "...a controlling relation -- the relation between a response and a discriminative stimulus (p.123)."
For some authors, like Skinner, attention is simply synonymous with stimulus control. For others, attention refers to selective stimulus control. Here, only a portion of the available discriminative stimuli in the environment (so-called nominal stimuli) actually control operant behavior (so-called functional stimuli). Experimental analysis can disclose precisely which of the available stimuli in the environment actually control operant behavior.
In operant conditioning, attention can be experimentally examined by varying the contingencies of reinforcement for making an operant response in the presence of different discriminative stimuli. Three key attentional effects in operant conditioning are: overshadowing, blocking, and relative validity. Each effect reveals that only a subset of the available stimuli may effectively control operant behavior.
A further important insight is that, without making an appropriate observing response, discriminative stimuli cannot exert control over operant behavior. A motorist who runs a red light might be said to have been inattentive; perhaps a better statement might be that he failed to look at the traffic light, thus precluding his coming under the control of its color.