Most studies that examined the precuing of motor responses have been interpreted as indicating that response specification is a variable-order process. An apparent exception to this conclusion was obtained by Miller (1982) for the preparation of discrete finger responses. Precuing was beneficial only when the precued responses were on the same hand, suggesting that response specification occurs in a fixed order, with hand specified before other aspects of the response. Three experiments examined this discrepant finding for discrete finger responses. Experiment 1 demonstrated that with sufficient time (3 s), all combinations of responses can be equally well prepared. Experiments 2 and 3 showed that the precuing advantage for same-hand responses at shorter precuing intervals is due to strategic and decision factors, not to an ability to prepare these responses more efficiently. Preparation of finger responses, thus, also appears to be variable. This conclusion poses problems for Miller's extension of the precuing procedure to the evaluation of discrete versus continuous models of information processing.

Reeve and Proctor (1984) demonstrated that a precuing advantage obtained for certain pairs of finger responses in a four-choice task is a type of spatial-compatibility effect. This compatibility effect was attributed by Reeve and Proctor to translation processes that relate stimuli to responses. An advantage similar to that obtained with spatial-location stimuli also has been obtained with two-dimensional symbolic stimuli, which have no spatial-location attribute. Miller (1982a) presented evidence that the advantage obtained with symbolic stimuli is not a compatibility effect, and he argued that a translation account cannot explain this advantage (Miller, 1985). The present study used various response sets to demonstrate that the symbolic stimulus sets do show compatibility effects similar to those shown by the spatiallocation stimuli. The results were interpreted as supporting a salient-features coding principle in which both stimuli and responses are coded in terms of the salient features of each, with the translation processes based on the relations between the stimulus and response codes. We would like to thank James McAlarney III for programming the experiments and

The stimulus-response translation stage of human information processing plays a mediating role of relating stimuli to assigned responses. The translation stage has been implicated as the locus of a pattern of differential precuing benefits obtained in spatial-choice tasks (Proctor & Reeve, 1986; Reeve & Proctor, 1985): When pairs of finger responses from the middle and index fingers of each hand are precued, the two leftmost and two rightmost responses show the greatest benefit. This pattern of differential benefits, which occurs regardless of whether the hand placement is adjacent or overlapped, has been attributed to spatially coded representations of the stimulus and response sets in the translation stage. Experiment 1 evaluated whether the mediating role of the translation stage changes with practice. All precued pairs of responses showed equivalent benefits in the last of three sessions. This result indicates that the spatial representations used initially to translate between stimuli and responses have been altered to be more efficient or have been replaced by productions that directly specify fingers. Experiment 2 used a fourth session in which subjects were transferred from the overlapped hand placement to the adjacent placement, or vice versa. For subjects in the former condition, the pattern of differential precuing benefits reappeared in the transfer session. This lack of transfer is consistent with the hypothesis that task-specific productions develop with practice that directly relate stimuli to fingers. For subjects who practiced with the adjacent placement and switched to the overlapped placement, only a nonsignificant tendency existed for the pattern of differential precuing benefits to reappear. This failure of the pattern to reappear could indicate that spatial representations continue to be used to translate between stimuli and responses. Alternatively, as occurs with the overlapped placement, task-specific productions could be acquired that relate stimuli to fingers. If so, the failure of the pattern of differential precuing benefits to reappear would reflect a modification in the representations that are used for translation in the transfer session. Specifically, if subjects were coding the stimulus and response sets on the basis of the distinction between the two hands, as well as the spatial distinction, the differential benefits would be minimized because hand coding should benefit different responses from those benefitted by spatial coding. These alternative explanations were evaluated in Experiment 3 by having subjects who practiced with the adjacent placement switch to a placement in which the hands were crossed completely.(ABSTRACT TRUNCATED AT 400 WORDS)

Proctor and Reeve (1985) found that when two-dimensional symbolic stimuli are assigned to finger responses, the translation between stimuli and responses operates according to a salient-features coding principle. The present study addresses the issue of whether this principle also applies to situations in which the stimuli are spatial locations assigned compatibly to response locations. Three experiments used vertical stimulus and response arrangements that allowed finger movements to be similar for different hand placements. Experiment 1 provided evidence for spatial coding on the basis of the salient top-bottom feature of the arrangements. Although spatial coding predominated, evidence for coding on the basis of a distinction between hands also was obtained. Experiment 2 showed that hand coding is not used when the hand distinction fails to correspond with the salient feature of the stimulus display. Experiment 3 indicated that when the hand distinction does correspond with the salient stimulus feature, hand coding occurs only when the hands are turned inward. The experiments demonstrate that compatibly assigned spatial-location stimulus and response sets are coded in terms of their salient features, much as are the stimulus and response sets when the assignments are less compatible.Models of human information processing customarily distinguish between at least three basic stages. These stages are referred to by a variety of names (e.g., perception, decision, and response, Miller, 1982; stimulus identification, response selection, and response programming, Schmidt, 1982; perception, translation, and effector control, Welford, 1976). Despite the differences in terminology, in all cited examples the first stage is concerned with encoding stimuli, the second with translating stimuli to responses, and the third with preparing and executing responses (motor programming). Much research has focused on clarifying the properties of these respective stages. The typical procedure is to determine the stage at which a particular phenomenon occurs and then to evaluate the properties of that stage by examining the nature of the phenomenon in detail (e.g.,

No abstract