Attention lec3

• What is automatic processing?


  A fast, parallel (numerous processes can happen at once) type of processing that requires little effort and has no capacity demands (so performance doens't worsen with task difficulty).
• Where does automatic processing come from?

  Through extensive practice, but not through conscious attention/effort. It is difficult to control and you are rarely conscious of it happening, but once you start you cannot stop.

  
• What are controlled processing?

  Slower, effortful processing that is dependent on capacity demands. It is dependent on focused attention and control processes within the brain, though it can be changed quickly

  
• Schneider and Shiffrin (1977) search task

  Participants are presented with items and then asked to search for them in a number of frames with distractors. It is set up to examine the difference between automatic and controlled processing. It consists of two main conditions: variable mapping and consistent mapping.

  
• Schneider and Shiffrin (1977): consistent mapping condition

  The targets were consistently different from distractors across the whole experiment, ie number versus consonants. This condition was hypothesized to trigger automatic processing

  
• Schneider and Shiffrin (1977): variable mapping condition.

  Items that were targets in one set of frames could be distractors in another. This condition was hypothesized to require controlled processing.

  
• What are the other independent variables in Schneider and Shiffrin's task?

  Memory set size (1,2, or 4), frames size (number of items on each frame: 1,2,4) and length of presentation of each frame.Each level of the IVs increases the demands of the task.
• Consistent Mapping Results


  Increasing the capacity demands did not result in any changes to performance

  
• Variable Mapping Results


  Performance was affected by both memory load (number of items to remember) and perceptual load (number of items to search). Increasing capacity demands reduced performance.

  
• Shiffrin and Schneider (1977): can automatic responses be learned through training? Consonants were split into 2 groups: B-L (group 1) and Q-Z (group 2). Used same paradigm as before (consistent mapping using group 1 for target and group 2 for distractors, whereas for variable mapping distractors and targets could come from either group)


  After thousands of trials with consistent mapping, performance became independent of item in memory set and frames set. In other words, the capacity demands stopped affecting reaction time. However, after the groups were switched, performance became very poor - showing its harder to unlearn automatic processes.

  
• Posner's Spatial Cueing Task


  Participants see a fixation point in the middle of the screen. Then, a cue appears in one of two ways. In central cueing, an arrow points to the left or the right. In the peripheral cueing, the cue is where the target will appear. These cues can be valid or invalid.

  
• What elements can be varied in Posner's Spatial Cueing tasks?

  Cue type (peripheral versus central), validity (valid or invalid), usefulness of the cues and cue target delay
• Posner's Task: Informative versus Uninformative cues


  For informative cues, the percentage of valid cues is high in relation to invalid cues. For uninformative cues, the percentage of valid cues is lower in relation to invalid cues.

  
• Posner's Task: Cue Target Delay (CTOA)


  The time between the cue and the presentation of the target. These delays can be 0 to 500ms.

  
• What is the dependent variable of Posner's cueing paradigm?

  The Validity factor/effect, which is the difference between the invalid cues' RT and the valid cues' RT. The Validity factor tells us if the cues are being used.

  
• What does it mean if Invalid RT minus Valid RT is positive?

  The invalid trials are slower than the valid trials, showing that participants are focused on where the target will appear and they are following cues.
• What does it mean if Invalid RT minus Valid RT is zero?

  There is no real difference between RTs from the type of trial, showing that participants are not focused on where the target will appear and they are not following the cues.
• What does it mean if Invalid RT minus Valid RT is negative?

  The invalid trials are faster than the valid trials, showing that participants are focused away from where the target will appear (which is strange)
• What does data look like for informative peripheral cues, and what does this suggest?

  A validity effect appears very quickly - within the first hundred milliseconds. This effect is consistent across cue target delays. This suggests that peripheral cueing is at least initially an automatic response.

  
• What does data look like for informative central cues, and what does this suggest?

  The validity effect response is slower acting and remains consistent across cue target delays. This suggests that central cueing is not an automatic response because arrows have to be processed and interpreted.

  
• What does data look like for uninformative central cues, and what does this suggest?

  No validity affect appears because people are not using the cues. This suggests that participants realize cues are unhelpful, suggesting that central cueing is a controlled response

  
• What does data look like for uninformative peripheral cues, and what does this suggest?

  Even though the cues are not helpful, people still use the cues. This suggests that peripheral cueing is an automatic response, because the cue captures attention despite being unhelpful

  
• How do we infer automaticity from the Posner cue task?


  If the validity effect is there, then participants are following the cue. However in the uninformative conditions, participants shouldn't be using the cues. They don't use the cues in the central cueing condition, showing they have controlled their attentional focus. However, they do follow the peripheral cues with short cue target delays, showing that participants cannot stop the attentional focus, suggesting it is automatic.
• Posner's Attention Model


  Attention is a spotlight that is focused on the fixation point. On cuing, the spotlight disengages from the fixation point and moves to the cued spot. If the spotlight is on a valid cue when the target appears, there is a faster reaction time. However if the spotlight is on an invalid cue, the process of disengaging and moving focus has to repeat, slowing the response.

  
• Central cueing summary


  Slower than peripheral cueing, under conscious control, can be ignored when they aren't helpful, seems to need interpreting, are symbolic cues, and are acontrolled process.
• Peripheral cueing summary


  Faster than central cueing, not under conscious control, cues still cue even when they aren't helpful, seems to be driven by the stimulus,automatic process.
• Endogenous attention processes


  Top-down conscious allocation of attention, effortful and interpretive, controlled or modulated by goals and context, voluntary
• Exogenous attention processes


  Bottom-up automatic allocation of attention, controlled by external events, reflexive and involuntary, attention is captured.
• What happens to the validity effect of uninformative peripherial cues after cue target delays of 200ms?


  Inhibition of return: attention drifts away whilst waiting for the target and doesn't automatically come back.

  
• What are some possible explanations of inhibition of return?

  There are many possible explanations, but one is that attention moves around naturally

  
• Frisesen & Kingstone 1998: are eye gaze cues automatic? Participants told to ignore central eye gaze cues. An equal number of valid, invalid, and neutral cues were presented. A validity effect would suggest an automatic response.


  There was a definite validity effect with participants quicker to cued (valid) targets than uncued (invalid) targets. Effects present very early at 100 ms and persistent effects over Cue target delay

  
• Driver et al (1999): used real faces as central eye gaze cues. Cues were uninformative and cue target delay was manipulated. If validity effects happen in shorter cue target delay, this would be evidence of automaticity.


  Although a validity effect presents, it is only really at 700 ms. This might be because the task is a judgement task, not a straight response, adding an extra layer of complexity
• Driver et al (1999): simplified the task, first presenting the face so people could get used to it and then presenting the cue, followed the target. Cues were again uninformative.


  Better effects than previous study, with validity effects at 300ms.

  
• Driver et al (1999): after the first edition of the simplified task, they made the cues contraindicative with 80% being invalid.


  They still found a validity effect at 300ms, but an opposite effect at 700ms. This suggests automatic cueing early on, but perhaps controlled attention later on.

  
• What do eye-gaze studies show about central cues?

  Certain central cues, such as eye gaze, can provoke automatic-like repsonses