Masking. CogLab: Metacontrast Masking

Masking CogLab: Metacontrast Masking Masking How does mask interfere with perception of target? Mask Form: Pattern mask Occupies same space as targe...
20 downloads 0 Views 2MB Size
Masking CogLab: Metacontrast Masking

Masking How does mask interfere with perception of target? Mask Form: Pattern mask Occupies same space as target

Metacontrast mask Similar contours of target but doesn’t overlap space

4 dot mask Non-similar shape to target

Pattern masking SOA = stimulus onset asynchrony - 300 (mask before target) +300 (target before mask

Temporal dimension Forward masking Backward masking

SOA White dot: low intensity mask Grey dot: med intensity mask Black dot: high intensity mask

Metacontrast masking Backward masking Target (30ms) then mask (30ms)

Greatest at SOA of 50100ms

CogLab: Metacontrast masking Spring ‘12 data (N = 13) Expected result

Report location of “thin” rectangle (1 of 4 locations) 30ms target; 30ms mask; IV: SOA; DV: % correct Expected result: SOA 50-100ms worst acc (“U-shaped”)

Theories of masking Breitmeyer (1984) Pattern mask Finding: worst @ SOA=0ms; ok @ +/- 100ms Theory/conclusion: Integration effect (camouflage/noise)

Weisstein, et al. (1975) Metacontrast mask Finding: U-shaped effect (worst @ SOA=~80ms) Theory: Fast signal = on/off; slow signal = object processing (e.g. id shape) Fast signal of mask turning on interrupts slow object id signal of target

Conclusion: Early visual interruption process

Enns & DiLollo (1997) 4-dot mask …

Two-channel theory “Interruption masking” Onset of each stimulus (target and mask) initiates activity in two channels One signal fast-acting, but short-lived Signals stimulus onset and offset

Other slow-acting, but longer lasting Signals info regarding stimulus shape/color

Metacontrast masking effect Fast-acting signal in response to mask inhibit slow-acting signal generated by earlier target

Will a 4-dot mask have same effect?

Enns & DiLollo (1997) Task: report if left or right side of diamond is missing corner 30ms target; 30ms mask IV: Type of mask IV: # of locations (1 or 3) IV: SOA (+/- 300ms) IV: target/mask share location (same/diff)

Enns & DiLollo (1997) Exp 1 3 location trials (only 1 target) Target/mask in same location IV: central vs parafoveal locations

Enns & DiLollo (1997): Proximity effx Exp 2 IV: contour proximity IV: location (central vs. parafoveal) For central location: Metacontrast mask – only near has effect

For parafoveal location: 4-dot mask effect regardless of proximity

Enns & DiLollo (1997) Exp 3: set-size manipulation (# targets)

Only central location data presented

Enns & DiLollo (1997) Question: Can 4-dot mask create masking effect? What is role of attention in perception?

Method: Detect target (IV: SOA, location, set-size)

Results/conclusions: 4-dot mask can act like metacontrast mask depending on: Location, number of locations, and set-size (# targets)

Conclusion Unattended locations (require spread attention) - have low resolution -> vulnerable to masking effect by 4-dots Object substitution theory: Mask doesn’t just interrupt processing , it is new focus of object recognition Effect of attention mechanism

What variables can be controlled? Shape Mask; Target; Target compared to mask Duration Target; Mask SOA: onset Forward; Simultaneous; Backward Offset Simultaneous; Trailing mask Spatial overlap/ proximity to target Pattern mask; Metacontrast mask; 4-dot mask Location of objects Central; Parafoveal Display size (# of objects) Other: luminance, contrast, practice, stim/mask importance

Aging and Visual Masking Atchley & Hoffman (2004) Black squares: Younger Ss Open circles: Older Ss Solid lines: “Contour mask” (i.e. Metacontrast mask) Dotted lines: “Object mask” or 4-dot mask

Di Lollo’s current research

4-dot mask Object-substitution mask

Masking if: Attention distributed among many similar objects (display size) Requires trailing mask (remains on after target – aka backward mask)

No masking if: Only 1 target Many objects, but target is distinct Simultaneous offset Forward onset

•Task: Report target highlighted by 4 dots •Trailing mask •IV: set-size •Masking if: •Multiple objects •Target not distinct •No spatial cue before target

Enns & Di Lollo’s Object-substitution model of masking Comparison of previous vs. current pattern Searches for match between perceptual code and sensory code If mask continues: there is “mismatch” If many targets, will loose target signal

Masking dependent on Set size Mask duration after target

Influence of spatial attention in masking Mask is new “focus of attention”

Enns (2004)

Theories of masking Breitmeyer (1984) Pattern mask: monotonic effect (strongest SOA=0ms) Integration effect

Weisstein, et al. (1975) Metacontrast mask: U-shaped effect (strongest SOA=80ms) Early visual interruption process Inhibitory signals generated by mask

Enns & Di Lollo (1997) Object-substitution mask: effect depends on condition Attentional process Masking due to object substitution

Shelley-Tremblay & Mack (1999) Was target presented?

Shelley-Tremblay & Mack (1999) Examine targets that capture attention Exp1: Happy face Exp 2: Own name

Examine mask that captures attention Exp 3: Own name

Shelley-Tremblay & Mack (1999) Exp 1

Shelley-Tremblay & Mack (1999) Exp 2 Target was a 4 letter word:

Mask was 2 white horizontal bars:

When Ss detected target, none able to identify it!

Shelley-Tremblay & Mack (1999) Exp 3 Target: “CORK” Mask was:

Shelley-Tremblay & Mack (1999) Question: Can changing the meaningfulness of targets or masks change probability of target detection?

Method: Use happy face or name as target or mask

Results/Conclusions: Targets that capture attention can hold onto attentional processes lowering effectiveness of masks Meaningful masks can divert attention away from target lowering detection rate of target

Attention is necessary for conscious perception!

Why study masking? Why do we care about the topic? Fundamental tool in: Cognitive psychology Limit the amount of information processing

Vision research Understand perceptual process Time to process visual information

Attention research Interaction of attention and perception processes

Masking: Conclusions Questions: What is time course of conscious visual processing? What factors influence object perception?

Methods: Identify/recognize target with mask

Results: Masks reduce perception of target But results depend on many factors

Conclusions: Masks can camouflage target, interrupt processing, or serve as new focus of object recognition process Attention is factor in object perception