Confidence and accuracy of near-threshold discrimination responses.

This article reports four subliminal perception experiments using the relationship between confidence and accuracy to assess awareness. Subjects discriminated among stimuli and indicated their confidence in each discrimination response. Subjects were classified as being aware of the stimuli if their confidence judgments predicted accuracy and as being unaware if they did not. In the first experiment, confidence predicted accuracy even at stimulus durations so brief that subjects claimed to be performing at chance. This finding indicates that subjects's claims that they are "just guessing" should not be accepted as sufficient evidence that they are completely unaware of the stimuli. Experiments 2-4 tested directly for subliminal perception by comparing the minimum exposure duration needed for better than chance discrimination performance against the minimum needed for confidence to predict accuracy. The latter durations were slightly but significantly longer, suggesting that under certain circumstances people can make perceptual discriminations even though the information that was used to make those discriminations is not consciously available.

Is dual-task slowing instruction dependent?

When 2 tasks must be performed concurrently, each requiring a choice of response, dual-task slowing is typically found. However, E. H. Schumacher et al. (1997) reported that dual-task slowing can be eliminated when equal priority is assigned to each task. Experiment 1 largely confirmed this with the same tasks as Schumacher et al. (tasks using stimulus-response combinations of visual-manual and auditory-vocal pairings). Experiment 2 retained the equal-priority instructions but switched the task pairings (to visual-vocal and auditory-manual); substantial dual-task slowing occurred. Experiment 3 used the same two response sets but only a single stimulus; slowing was again obtained despite equal priority instructions. Equalizing task priority was not sufficient to eliminate interference; relatively unusual cases in which dual-task interference is eliminated seem to depend on task-specific features.

Processing bottlenecks in dual-task performance: structural limitation or strategic postponement?

Recent evidence indicates that a central bottleneck causes much of the slowing that occurs when two tasks are performed at the same time. This bottleneck might reflect a structural limitation inherent in the cognitive architecture. Alternatively, the bottleneck might reflect strategic (i.e., voluntary) postponement, induced by instructions to emphasize one task over the other. To distinguish structural limitations from strategic postponement, we examine a new paradigm in which subjects are told to place equal emphasis on both tasks and to emit both responses at about the same time. An experiment using this paradigm demonstrated patterns of interference that cannot easily be attributed to strategic postponement, preparation effects, or conflicts in response production. The data conform closely to the predictions of structural central bottleneck models.

Perception and production of brief durations: beat-based versus interval-based timing.

Two experiments examined whether timing of short intervals is beat- or interval-based. In Experiment 1, subjects heard a sequence of standard tones followed by 2 test tones; they compared the interval between test tones to the interval between the standards. If optimal precision required beat-based timing, performance should be best in blocks in which the interval between standard and test reliably matched the standard interval. No such effect was observed. In Experiment 2, subjects heard 2 test tones and reproduced the intertone interval by producing 2 keypress responses. Entrainment to the beat was apparent: First-response latency clustered around the standard interval and was positively correlated with the produced interval. However, responses occurring on or near the beat showed no better temporal fidelity than off-beat responses. One plausible interpretation of these findings is that the brain always times brief intervals with an interval timer; however, this timer can be used in a cyclic fashion to trigger rhythmic responses.

Attention and performance.

Recent progress in the study of attention and performance is discussed, focusing on the nature of attentional control and the effects of practice. Generally speaking, the effects of mental set are proving more pervasive than was previously suspected, whereas automaticity is proving less robust. Stimulus attributes (e.g. onsets, transients) thought to have a "wired-in" ability to capture attention automatically have been shown to capture attention only as a consequence of voluntarily adopted task sets. Recent research suggests that practice does not have as dramatic effects as is commonly believed. While it may turn out that some mental operations are automatized in the strongest sense, this may be uncommon. Recent work on task switching is also described; optimal engagement in a task set is proving to be intimately tied to learning operations triggered by the actual performance of a new task, not merely the anticipation of such performance.

Spontaneous allocation of visual attention: dominant role of uniqueness.

Common sense suggests that unique or abruptly changing (transient) elements in a visual scene often draw attention involuntarily. Visual search studies paint a seemingly different picture, however Unique items usually draw attention involuntarily only when observers seek a unique target. One type of transient--abrupt onsets--draws attention involuntarily, but only when the observer is seeking an onset (Folk, Remington, & Johnston, 1992). One way of reconciling common-sense with these findings is to suppose that when people view a scene with no specific goal or task, they adopt a default set, which might prioritize novelty and/or transients. In two experiments, 336 subjects saw a single display of six items for 900 msec, expecting to have to describe it (Experiment 1) or make an aesthetic judgment about it (Experiment 2). One item in the display was either uniquely flashing (surrounded by static items) or uniquely static (surrounded by flashing items). In both studies, the unique item, even if static, was more often reported than the non-unique item, with flashing items enjoying an additional advantage.

The role of iconic memory in change-detection tasks.

In three experiments, subjects attempted to detect the change of a single item in a visually presented array of items. Subjects' ability to detect a change was greatly reduced if a blank interstimulus interval (ISI) was inserted between the original array and an array in which one item had changed ('change blindness'). However, change detection improved when the location of the change was cued during the blank ISI. This suggests that people represent more information of a scene than change blindness might suggest. We test two possible hypotheses why, in the absence of a cue, this representation fails to produce good change detection. The first claims that the intervening events employed to create change blindness result in multiple neural transients which co-occur with the to-be-detected change. Poor detection rates occur because a serial search of all the transient locations is required to detect the change, during which time the representation of the original scene fades. The second claims that the occurrence of the second frame overwrites the representation of the first frame, unless that information is insulated against overwriting by attention. The results support the second hypothesis. We conclude that people may have a fairly rich visual representation of a scene while the scene is present, but fail to detect changes because they lack the ability to simultaneously represent two complete visual representations.

How persuasive is a good fit? A comment on theory testing.

Quantitative theories with free parameters often gain credence when they closely fit data. This is a mistake. A good fit reveals nothing about the flexibility of the theory (how much it cannot fit), the variability of the data (how firmly the data rule out what the theory cannot fit), or the likelihood of other outcomes (perhaps the theory could have fit any plausible result), and a reader needs all 3 pieces of information to decide how much the fit should increase belief in the theory. The use of good fits as evidence is not supported by philosophers of science nor by the history of psychology; there seem to be no examples of a theory supported mainly by good fits that has led to demonstrable progress. A better way to test a theory with free parameters is to determine how the theory constrains possible outcomes (i.e., what it predicts), assess how firmly actual outcomes agree with those constraints, and determine if plausible alternative outcomes would have been inconsistent with the theory, allowing for the variability of the data.

Evidence for split attentional foci.

A partial report procedure was used to test the ability of observers to split attention over noncontiguous locations. Observers reported the identity of 2 targets that appeared within a 5 x 5 stimulus array, and cues (validity = 80%) informed them of the 2 most likely target locations. On invalid trials, 1 of the targets appeared directly in between the cued locations. Experiments 1, 1a, and 2 showed a strong accuracy advantage at cued locations compared with intervening ones. This effect was larger when the cues were arranged horizontally rather than vertically. Experiment 3 suggests that this effect of cue orientation reflects an advantage for processing targets that appear in different hemifields. Experiments 4 and 4a suggest that the primary mechanism supporting the flexible deployment of spatial attention is the suppression of interference from stimuli at unattended locations.

Visual selection mediated by location: feature-based selection of noncontiguous locations.

Experiments using two different methods and three types of stimuli tested whether stimuli at non-adjacent locations could be selected simultaneously. In one set of experiments, subjects attended to red digits presented in multiple frames with green digits. Accuracy was no better when red digits appeared successively than when pairs of red digits occurred simultaneously, implying allocation of attention to the two locations simultaneously. Different tasks involving oriented grating stimuli produced the same result. The final experiment demonstrated split attention with an array of spatial probes. When the probe at one of two target locations was correctly reported, the probe at the other target location was more often reported correctly than were any of the probes at distractor locations, including those between the targets. Together, these experiments provide strong converging evidence that when two targets are easily discriminated from distractors by a basic property, spatial attention can be split across both locations.

Do images involuntarily trigger search? A test of Pillsbury's hypothesis.

Pillsbury (1908) suggested that deciding to search for something in a scene consists of nothing more than forming a visual image of the target. If so, imaging should trigger search even when it would be more advantageous not to search. Subjects were cued to form an image of a specified object (e.g., tiger) and to press a key when they had done so. This initiated the presentation of a sequence of pictures, with a single target digit interspersed; the subject's task was to report this digit. The sequence contained a picture of the same type of object that the subject had just imaged (e.g., a tiger), either before or after the target digit. If this picture was detected involuntarily, an attentional blink should have impaired digit detection when the picture preceded the digit. This was confirmed in two experiments, even when instructions specifically encouraged subjects to discard the image and to avoid searching for it. The results support Pillsbury's hypothesis.

When two memories can and cannot be retrieved concurrently.

We examined whether two memories can be retrieved concurrently from long-term memory. In Experiment 1, the subjects recalled words, either from two categories--alternating between the two--or from just one category. In Experiment 2, the subjects recalled two words belonging to either the same category or different categories, and the category prompts for these two responses appeared either simultaneously or successively. The results of both studies are consistent with the view that two items from different categories must be retrieved serially, whereas two items from the same category can be retrieved in parallel.

Does perceptual analysis continue during selection and production of a speeded response?

This study asked whether perceptual analysis of a stimulus can continue while a response to this stimulus is being generated. In Experiment 1, subjects rapidly named a word that was visually degraded with superimposed pixels. Near the time of response, degradation was removed for a short time and then followed by a mask. Subjects then made a second, unspeeded judgment about the identity of the word. Unspeeded judgments were more accurate, showing that the degradation-free stimulus exposure was processed. In Experiment 2, the task was the same, but the degradation was gradually faded out for an individually adjusted duration. Comparison of unspeeded-response accuracy on trials with and without a speeded response showed that stimulus analysis continued at full efficiency during speeded-response generation. The results support conclusions of Rabbit and Vyas (1981) that perceptual analysis continues during response stages. This form of continuous processing does not necessarily contradict discrete stage models of human information processing, however.

Attentional limits in memory retrieval.

The hypothesis that episodic memory retrieval can occur in parallel with other cognitive processes was tested in 2 experiments. Participants memorized words and then performed speeded cued recall (Experiment 1) or speeded yes-no recognition (Experiment 2) in a dual-task situation. The psychological refractory period design was used: The participant was presented with a single test item at various stimulus onset asynchronies (SOAs; 50-1,200 ms) after a tone was presented in an auditory-manual 2-alternative choice reaction task. Reducing the SOA increased the memory task reaction times. This slowing was additive with the effect of variables slowing retrieval in the memory task. The results indicate that memory retrieval is delayed by central processes in the choice task, arguing that the central bottleneck responsible for dual-task interference encompasses memory retrieval as well as response selection.

Visual selection mediated by location: selecting successive visual objects.

In these experiments, each stimulus consists of a series of frames, each containing a target digit of one color and a distractor digit of another color. The task is to name the highest digit of the target color. Subjects make fewer errors when successive targets appear at the same location than when they appear at different locations, apparently because they select target objects by using a mechanism that is based on location. When successive targets appear at the same location, there is no need to "move" the selection mechanism to a new location, leaving more time to identify the stimuli. These experiments show that location-based selection is used even though selection by color would be more direct. They also demonstrated a method of measuring location-based selection that can be applied to a variety of visual tasks. Further experiments reveal that although location-based selection is used to identify a digit in the presence of a digit distractor, it is not used to identify a digit in the presence of a letter distractor, suggesting that this selection mechanism is not used in this situation to prevent interference among the basic features making up letters and digits, but to inhibit responses associated with the distractors.

Repetition blindness: perception or memory failure?

Repetition blindness (RB) is the failure to report a repeated item when reporting all of the elements from a rapidly displayed sequence of visually presented items. N. Kanwisher (1987) has characterized RB as a perceptual problem. Experiments 1 and 4 manipulated the order of report of the repeated items. Experiments 2-4 employed tasks other than full report that should still have shown RB according to a perceptual hypothesis, but none was found. In Experiment 5, RB was found when one of the repetitions was spoken and the other was presented visually; this RB could be localized to retrieval processes. RB appears to reflect processes and strategies peculiar to full report from rapid displays rather than any fundamental perceptual limitation and might be the same as the Ranschburg effect found in immediate recall tasks.

Spatial attention and vernier acuity.

This study examined whether or not vernier acuity would be improved if the location of a briefly presented vernier stimulus was pre-cued. The vernier target appeared alone, or together with straight lines or ellipses. Effects of spatial pre-cuing were found only when straight line distractors were present. It is suggested that since the straight lines are confusable with the vernier targets, they introduce statistical noise in decision. Precuing the most probable location that contains a target may help by allowing this noise to be excluded.

Sequential operation of disconnected cerebral hemispheres in split-brain patients.

When two concurrent sensorimotor tasks require separate responses, selection of the first response generally delays selection of the second. Dual-task performance was examined in four patients who had undergone surgical transection of the forebrain commissures including the corpus callosum. One light flash was presented to each visual field in succession, and patients made a choice response to each stimulus with the ipsilateral hand, thereby confining the tasks to separate hemispheres. All four showed dual-task interference very similar to that found with normal individuals. Therefore, still-intact subcortical structures must play a critical role in sequencing response selection processes (the 'dual-task bottleneck'), confirming the distinction between the attentional limitations involved in planning actions and those involved in perceptual analysis.

Dual-task interference in simple tasks: data and theory.

People often have trouble performing 2 relatively simple tasks concurrently. The causes of this interference and its implications for the nature of attentional limitations have been controversial for 40 years, but recent experimental findings are beginning to provide some answers. Studies of the psychological refractory period effect indicate a stubborn bottleneck encompassing the process of choosing actions and probably memory retrieval generally, together with certain other cognitive operations. Other limitations associated with task preparation, sensory-perceptual processes, and timing can generate additional and distinct forms of interference. These conclusions challenge widely accepted ideas about attentional resources and probe reaction time methodologies. They also suggest new ways of thinking about continuous dual-task performance, effects of extraneous stimulation (e.g., stop signals), and automaticity. Implications for higher mental processes are discussed.