Notes

• Reaction time:

  • is a proxy for studying mental processes
  • is a face valid component of human performance
  • remains variable even when overall performance is extremely high
• Simple reaction time example: light comes on and you press a button
• Simple RT processes:

  detect stimulus -> press button (perform action)
• Choice reaction time example: matching the color of the light to a button of the same color
• Choice RT processes:

  detect stimulus -> discriminate color -> choose response -> perform action
• choice reaction time is slower than simple reaction time
• go / no-go reaction time example: only press a button when the light is a certain color
• go / no-go reaction time processes:

  detect stimulus -> discriminate stimulus -> perform action (press button)
• types of reaction time:

  • simple reaction time
  • choice reaction time
  • go / no-go reaction time
• Donder's Subtractive Method:

  • CRT - SRT = discrimination + selection
  • CRT - GRT = discrimination only
• factors influencing simple rt:

  stimulus modality (ex. visual vs. audio)

  stimulus intensity (ex. brightness of a light)

  stimulus characteristics (ex. words vs. non words)
• temporal uncertainty - reaction time inscreases as stimulus onset asynchrony / warning interval increases (between blocks)
• expectancy - reaction time decreases as warning intervals increases with blocks (to a point)
• Hick's (Hick-Hyman) Law - reaction time increases with the number of choices
• information theory - only one bit of information is required to reduce the uncertainty of the world by half
• Reaction times can be used to determine whether manipulations affect different stages of processing or different processes
• stages of reaction time:

  1. sensory stage
  2. comparison stage
  3. response selection stage
• RT data issues in mental chronometry:

  what does time of the incorrect trial represent?

  not normally distributed / absolute floor effect

  usually contains outliers

  RT is variable within and between people
• speed / accuracy trade off - when increased performance in one dimension (speed / accuracy) negatively impacts performance in the other (accuracy / speed)
• Explanation of the speed / accuracy trade off: decision tasks require an increased amount of accumulated evidence and a threshold
• When response times are faster, the respondent is using lesser evidence. With lesser evidence, there is a greater probability of error.
• Accuracy can be reported as:

  1. percent correct (accuracy)
  2. error rate
  3. d prime
• In reaction time, a lower, quicker time is better
• Macro trade off - taskwise effects whereby difficulty or condition drives the trade off
• Micro trade off - responsewise effects whereby faster responses are less likely to be correct
• basic signal detection assumptions:

  1. signal can be one of two states (present/absent)
  2. receiver responds according to perception of those two states
• signal detection responses:

  1. hit
  2. false alarm
  3. miss
  4. correct rejection
• In SDT, sensitivity is the separation between 2 distributions
• bias - an indication of a shift in beta (criterion) to favor one response or another
• Applications of SDT:

  • healthcare (medical imaging, DSM diagnosis, referrals)
  • weather forecasting
  • traffic safety
• Fuzzy SDT:

  outcome is based on fussy mapping of state in which hits and false alarms are probability functions based on two values r (certainty of judgement) and s (probability of signal)
• Vigilance refers sustained performance over time
• Conditions for vigilance:

  low frequency of events (signals)

  high frequency of non-events (noise)

  high temporal uncertainty

  watch period of at least 30 minutes
• vigilance decrement is a decline in performance on a sustained task over time, typically seen in hit (detection) rate and response time
• vigilance artifact - hit rate effects are consistent where false alarm effects are variable
• vigilance explanations:

  arousal/boredom

  resource depletion

  expectancy/task adaption
• Vigilance enhancement (person):

  caffeine

  music (rock music, music that is unpredictable)

  transcranial direct current stimulation

  video games
• Vigilance enhancement (environment):

  training

  false signal insertion

  breaks
• Mental models are a mental representation (abstract in part or in whole) that captures sub components and includes operational relationships of components
• Mental models are maps to the real world, though they're not necessarily completely accurate