Michael Stiso's List: MS project

• • Initial system for allowing fire-fighter commanders to simulate possible outcomes of decisions.

• • Seems NOT to be map-based.

  • Provides a number emergency responder profiles, use cases, and types of interactions.

• "These results suggest that dynamic discontinuities defined as changes in luminance over time (e.g., onsets, offsets, motion) are capable of capturing attention involuntarily." - Michael Stiso on 2009-05-05
• "Theeuwes (1991b, 1992) claimed that in visual tasks in which the defining attribute is a singleton, selection occurs in purely stimulus-driven fashion."
  
  Our task does not consist of singleton stimuli. How will that affect selectability of color, onset, etc.? - Michael Stiso on 2009-05-05
• "The results indicate that Folk et aI.' s (1992) hypothesis that involuntary capture of attention is dependent on the control settings does not hold: Even when subjects have a clear attentional set for a static discontinuity, a dynamic discontinuity interferes, and vice versa.
  The results agree with Theeuwes's (l991b, 1992) hypothesis
  that visual selection depends on the relative saliency of the singletons present in the visual field. In conditions in which there is only one singleton present (i.e., the nodistractor condition), the mean RT represents the relative saliency of the singleton; that is, it depicts the time it takes for attention to be captured by the singleton. As evident in Figure 2, the RT analysis indicates that the mean RTs averaged over display size for the no-distractor conditions color and onset were not different (550 ms for color and 535 ms for onset), suggesting that they did not differ in their
  saliency. In line with the Theeuwes hypothesis, comparable
  interferences for the color and onset conditions were expected
  and observed." - Michael Stiso on 2009-05-05
• "According to Theeuwes's model (1993), in an unfocused state of attention, the preattentive process simply calculates differences in features within dimensions, resulting in patterns of activation at different locations, followed by an automatic shift of spatial attention to the location with the highest difference signal. In this model, selection operates irrespective of the task demands. The automatic shifts of attention can be considered the result of relatively inflexible, "hardwired" mechanisms that are triggered by the presence of specific stimulus properties."
  
  Can clusters of level 3 dots be considered to have a higher difference signal than other locations on the display? - Michael Stiso on 2009-05-05

• • "Several theories have been proposed that shed light on the cognitive processes involved in multiple target tracking (TT). One such, described by Pylyshyn and Storm (1988) and supported by a wealth of subsequent research (e.g. Trick & Pylyshyn, 1989, 1993; Pylyshyn, 1994; Pylyshyn et al., 1994) proposes that the ability to track moving objects is dependent upon a limited number of preattentive ‘indexes’ that may be attracted to objects as they appear in the visual field and, thereafter, continue to point to them as they move, facilitating the allocation of attention to the objects’ locations."

• no access - Michael Stiso on 2009-04-30

• • "A possible explanation for the discrepancy between the optimizing tracking and optimizing memory conditions is suggested by a current model of visual attention. According to the zoom lens model (Eriksen, 1990), the size of the attentional focus can vary with task demands. When more processing resources are required for more rapid or finer discriminations, resources would zoom in on the desired stimulus. The resolving power increases with a corresponding restriction in the field of view. A wider field of view is associated with low resolving power for details and loss in processing speed or efficiency. When subjects were optimizing the tracking task, the center of the visual field would follow the tracking cursor and the memory stimuli would at times fall on the periphery of the visual field and processed with reduced efficiency. When subjects were optimizing the memory task, the memory stimulus would be at the center of the visual field, but although efficiency for fine discriminations decreased in peripheral vision, efficiency of processing continuously moving objects is relatively unaffected (Leibowitz et al., 1982; Shapiro, 1990), thus protecting the tracking performance. Subjects' com- ments also supported such an explanation. Importantly, although the peripheral vision hypothesis may be able to explain why performance of the concurrent task was affected when tracking was optimized and not when the memory task was optimized, it cannot account for the graded performance changes across priority levels without further assumptions."
    
    If so, then participants focusing on our number task may have better performance on both tasks than participants who focus on dot task.
• It was hypothesized that the tracking task was protected by peripheral vision when subjects focused on the optimized memory task.

• • Using the PDF model, increasing the difficulty of a task should have one of three results: 1) The subject devotes more resources to maintain performance on that primary task, reducing the resources devoted to, and thus hurting performing on, a concurrent task; 2) resources devoted to the concurrent task remain fixed, leaving fewer for the primary task and hurting performance on it relative to a single-task condition; 3) some combination of 1 and 2.
    
    Cites previous research supporting the above view, but says that the "difficulty-performance trade-off" will not hold for highly data-limited tasks, for which performance is relatively resource-independent. It will also not hold when the two tasks are different enough to draw from separate resource pools, as in the multiple-resources theory.

  • Data-limited tasks tend to be easy, highly skilled, or well-practiced.
  • 8 more sticky notes...

  • • Resource-limited tasks are ones in which performance improves with the allocation of greater resources to it.
      
      Data-limited tasks are those in which performance remains constant with increasing resource investment.
      
      Which kind are our two tasks?

    • See Norman & Bobrow (1975) for the performance operating characteristic (POC graph), which cross plots performance on two tasks.

    • Previous studies asking subjects to vary the effort they devote to primary and secondary tasks have shown corresponding performance trade-offs on one task vs. the other, supporting the idea that resources can be withdrawn from one task and used to improve performance on the other.

    • Spreading limited resources across two tasks will likely result in one or both receiving too little, resulting in a performance hit on either (Gopher, 1986). That is the common finding in dual-task studies, and it is consistent with the resources concept. Also consistent with phenomena such as attention switching and confusion, though.

    • PDR model: P = R/D
      
      As objective task difficulty (D) increases, performance (D) decreases; as resources (R) increase, so does performance.

    • Figure 1.1 (p.5) shows the performance-resource function from Norman and Bobrow (1975) for primary and secondary tasks. As task difficulty increases, more resources must be devoted to maintain the same level of performance.

    • In one sense, "resources" are the "mental effort that is invested to improve performance". (p.4)

    • Resource concept: Humans have limited processing resources to allocate to task performance. Two tasks are more demanding than one, potentially leading to a situation in which not enough resources are available for either, leading to performance decrements in one or both.

• no access - Michael Stiso on 2009-04-30

• In 1988, Pylyshyn and Storm [11] convincingly overturned this single focus view by demonstrating continuous tracking of multiple targets (also see Box 1). Their experiments used arrays of identical items that moved within a rectangular area, bouncing off the borders and each other. At the beginning of a trial, some of the items were marked briefly as the targets and they then reverted to being identical to the other items (Figure 1). All items then moved randomly within the target area for the next 7–15 s. To measure performance at the end of the trial, a single item was probed and observers reported whether or not it was a target item. Most subjects were able to keep track of as many as 4 or 5 targets over several seconds.

  • Could perhaps use this method to investigate whether color vs. blinking vs. size helps users to track multiple items.
• A recent study shows that air traffic controllers have higher tracking capacity than undergraduates [4].

  • Better tracking for randomly moving targets, or for targets that move along somewhat predictable paths?
• A recent finding of hemifield independence [13] to be discussed near the end of this review has ruled out both grouping and switching models, at least in their simplest form.
• Pylyshyn's early results suggested that the limit for tracking independent targets was 4
• In a typical attention task (e.g. [9]), subjects are given a cue indicating a region of interest and a target is then briefly presented at that location (or not, on some small proportion of trials). By contrast, in a tracking task, there are no brief events; items typically remain constant in every respect but their location and the task is to monitor that changing location, making it more akin to the real-world allocation of attention. Multiple-object tracking is also an inherently active task like many real-world activities, as opposed to the sustained but fundamentally passive vigilance required in many monitoring tasks, characteristics that more resemble the demands of tasks like police stake-outs. Tracking also allows for a disciplined control of load by varying the number of targets whereas most classic attention paradigms vary load by restricting the time available for processing (e.g. via masking or brief displays). Compared with standard attention tasks, tracking tells us how attention connects to a target and how that connection is maintained as the target changes location and in some tasks, identity. We discover not the benefits of attention but the logistics.

3 more annotations...

• subjects are able to track a subset of up to 5 objects in a field of 10 identical randomly-moving objects in order to distinguish a change in a target from a change in a distractor

• Noticeability, which is sometimes referred to as conspicuity, is often used to describe the extent to which the design of a warning will gain or attract attention against a field of competing visual stimuli. Therefore, to have high noticeability, it is essential that a warning be as salient (e.g., stand out, be prominent, or conspicuous) as possible to capture the attention of individuals who might be focused on some other task.
• The salience of a visual warning can be enhanced using (1) large, bold print, (2) high contrast, (3) color, (4) borders, (5) pictorial symbols, and (6) special effects like flashing lights.
• Adding color to a warning can increase its ability to attract attention ( Gill et al., 1987) provided that the warning color is distinguishable from background and surrounding colors.

1 more annotation...

• "Several sources (Anastasi, HiJI, Murphy, Cardosi, Guttman, & Amaldi, 1995; Boff & Lincoln, 1988; Christ, 1975: Gerathewoht, 1951: 1952: 1953: 1954: Gilmore, Gertman, and Blackman, 1989; Military Sranclards, 1989; Van Orden and DiVita, 1993) suggest that blinking targets are more alerting than steady targets, and can aid the user in finding the targets quickly. Those sources recommend target size, color, shape, brightness contrast, frequency of blink and parameters for the ratio of time the blink should be on relative to the time it is off." - Michael Stiso on 2009-05-07
• "There is some debate in the literature concerning the distracting qualities of blinking on a display screen (Van Orden, DiVita & Shim, 1993; Van Orden & DiVita, 1993)." - Michael Stiso on 2009-05-07
• "The small (20%) change in intensity during blink had little attention-getting value, as confirmed by experimental findings of Mertens and Milburn (1996)."
  
  Small changes in intensity between blinks weren't attention grabbing. Perhaps in the MakeSense blink condition, the intensity changes between blinks in lower-level warnings could be reduced to alleviate distraction. - Michael Stiso on 2009-05-07
• "Stable accuracy and response time performance is maintained for amplitudes ranging from 75% to 100%, and a steep decline in accuracy, combined with increasing response times, is evident for the amplitudes ranging from 12.5% to 25%. This figure should aid display designers in their selection of an optimum blink amplitude for attentional purposes based on both of those performance measures." - Michael Stiso on 2009-05-07