Revealing hidden covariation detection: evidence for implicit abstraction at study.

Four experiments in the brain scans paradigm (P. Lewicki, T. Hill, & I. Sasaki, 1989) investigated hidden covariation detection (HCD). In Experiment 1 HCD was found in an implicit- but not in an explicit-instruction group. In Experiment 2 HCD was impaired by nonholistic perception of stimuli but not by divided attention. In Experiment 3 HCD was eliminated by interspersing stimuli that deviated from the critical covariation. In Experiment 4 a transfer procedure was used. HCD was found with dissimilar test stimuli that preserved the covariation but was almost eliminated with similar stimuli that were neutral as to the covariation. Awareness was assessed both by objective and subjective tests in all experiments. Results suggest that HCD is an effect of implicit rule abstraction and that similarity processing plays only a minor role. HCD might be suppressed by intentional search strategies that induce inappropriate aggregation of stimulus information.

List-context effects in evaluative priming.

Evaluative priming effects are often found in the evaluative decision task, in which persons judge the affective connotation (positive vs. negative) of a target word. The present experiments examined list-context effects to test whether evaluative and semantic priming follow the same laws. In Experiment 1, evaluative priming was found at prime-target stimulus onset asynchronies (SOAs) of 0 ms and 100 ms, but not at SOAs of--100, 200, 600, and 1,200 ms. Experiment 2 manipulated SOA (0, 200, and 1,200 ms) and the proportion (25%, 50%, and 75%) of the prime-target pairs that were evaluatively related. Contrary to the typical finding that increases in the proportion of related prime-target pairs lead to increased priming at long but not short SOAs, an effect of consistency proportion was found at SOAs of 0 ms (for reaction times) and 200 ms (in the accuracy data), but not at the 1,200-ms SOA. The pattern of results is discussed in relation to possible explanatory mechanisms of evaluative priming.

[Mental images and mental models]. / Mentale Bilder und Mentale Modelle.

Two experiments are conducted to assess the involvement of the visouspatial sketchpad of working memeory in propositional and spatial reasoning, using a dual-task methodology. While a concurrent tracking task was found to interfere with spatial reasoning, no such interference was found with propositional reasoning. The results are discussed with respect to the mental-models theory of reasoning.

[Cognitive load and listener orientation in monologue instruction]. / Kognitive Belastung und Hörerorientierung beim monologischen Instruieren.

It was experimentally tested whether high mental workload adversely affects the listener-adaptation of speakers giving an instruction. Forty-five subjects were randomly assigned to one of three groups with either low or high mental workloads and gave instructions on the assembly of a small machine model. Speakers in the low workload condition had the model available while instructing, whereas speakers in the high workload condition had to recall the assembly without external aids. In a third group, speakers were given a secondary task while instructing. Only in the low load condition were instructions apparently adapted to the listener. High load and dual task conditions, on the other hand, yielded messages that were equivalent as to their contents, regardless of the different communicative tasks. Additional assessment of speech characteristics demonstrated the experimental manipulation of mental workload to be effective. Results suggest that decreased listener-adaptation reflects a dissociation of the communicative demands of listener-adapted speech and the cognitive resources to meet these demands.