LOGO

The role of attentional breadth in perceptual chan

ge detectionProfessor: LiuStudent: Ruby

Objective

• Examined the relationship between perceptual change detection and attention which relating individual differences in attentional breadth to observers’ ability to detect changes in driving scenes.

References

• When changes to scenes at the same time with saccadic eye movements, we have a limited ability to detect them.

(Grimes, 1996; Henderson, 1997; McConkie & Currie, 1996)

• The objects of central interest probably attracted attention through higher level cognitive processes.

(Rensink et al. 1997)

References

• The FFOV represents the spatial area that is needed to perform a specific visual task without occurred eye or head movements. (Ball, Roenker, & Bruni, 1990; Mackworth, 1965, 1976)

• The size of the FFOV decreased with age.

(Ball, Beard, Roenker, Miller, & Griggs, 1988)

References

• Older drivers had poorer driving performances on the FFOV, a skill that would seem to because an ability to detect change in the environment.

(Isler, Parsonson, & Hansson, 1997; Rizzo, Rinach, McGehee, & Dawson, 1997)

Method - participants

• 25 Younger group• 13 women, 12 men.• Age from 18 to 33 years.

• 26 Older group• 18 women, 8 men.• Age from 55 to 80 years.

• Each participant had corrected visual acuity better than 20/40.

• Each participant had a driver’s license for 2 years, and drove over 25 miles per month.

Method - apparatus

• A Micron Millenia MME computer.

• A 12 × 16in. Viewsonic monitor.• 56 cm from the screen.

• A Fresnel lens.• Was used to remove the accommodation cues.• Increased the subjective size of the image regi

on.

Method – perceptual change task

• Each image was displayed for 240 msec and each blank screen for 80 msec.

• 80 photographs were taken from the driver’s view inside a car.

• The objects and their changes were categorized along three parts: eccentricity, meaningfulness, and salience.

• When they detected the change, press the mouse button and describe the change.

Method – perceptual change task

• The first pilot study (meaningfulness and salience)• 14 younger and 10 older participants.• Subjects saw two images of a scene on color p

rinted pages in a notebook. (82 scenes)• They were asked to rate the change according

to a 6 point Likert scale.• Meaningfulness was defined to the importa

nce of the change to driving performance.• Salience was defined to noticeable change

should be the high salient.

Method – perceptual change task

• The second pilot study (meaningfulness and salience)• 6 younger and 6 older participants.• Rate a single object in each of the 82

scenes.• Meaningfulness was defined to the

importance of the object to driving performance.

• Salience was defined to noticeable object should be the high salient.

Method – perceptual change task

• The perceptual change performance, the 80 driving scenes were divided into four categories:• Low meaning/low salience.• Low meaning/high salience.• high meaning/low salience.• high meaning/high salience.

Method – atentional breadth task

• An oblique target appearing in 11 vertical distractors.• Targets and distractors appear randomly at on

e of three eccentricities (10, 20, and 30 deg from fixation) along 8 radial meridians for a total of 24 possible positions.

• After finish the change detection task, they moved the mouse to one of the 24 possible target positions to indicate their response.

Results – change detection performance (RT)

• Main effects were significant for all four factors.• Age: younger adults performed

significantly faster than older adults, F(1,48)=41.02, P<0.001.

• Eccentricity: central changes were detected more quickly than peripheral ones, F(1,48)=35.14, P<0.001.

• Meaningfulness: low = 9, high = 8.2 sec; F(1,48)=9.65, P<0.003.

• Salient: low = 10.9, high = 6.8 sec; F(1,48)=313.93, P<0.001.

Results – change detection performance (RT)

• A significant two-way interaction was between age and salience, F(1,48)=6.53, p<0.014.• This result is not found in the previous

literature.

• The age and eccentricity interaction was no significant.• It may because didn’t control the eye

movements.

Results – change detection performance (RT)

• The three-way interaction between age, meaningfulness, and salience. F(1,48)=7.94, P<0.007.• Increase

meaningfulness had no effect on performance for either age group when changes were highly salient.

• When salience change was low, increasing meaningfulness help the performance of young, but not old.

Results – change detection performance (RT)

• A significant three-way interaction was also found between eccentricity, meaningfulness, and salience. F(1,48)=9.64, P<0.003.• When changes were

both high meaning and salience, the central changes were detected faster than peripheral changes.

• Different meaningfulness did not influence performance when changes were both peripheral and low salience.

Results – change detection performance (accuracy)

• Main effects were significant for:• Age[ F(1,38)=39.8, P<0.001 ]• Eccentricity[ F(1,38)=31.7, P<0.001 ]• Salience [ F(1,48)=64, P<0.001 ]

Results – change detection performance (accuracy)

• Two-way interactions was found for age × salience. [ F(1,48)=19.3, p<0.001 ]

• Also found significant two-way interactions for meaningfulness × salience. [ F(1,48)=4.7, P<0.03 ]

Results – relationship between FFOV and change detection performance

• A larger FFOV correspond to faster detection of object changes. (r=-0.68, p<0.001)

Results – relationship between FFOV and change detection performance

• The size of the FFOV appear to be related to change detection for central changes and perhaps even more strongly for peripheral changes.

• The correlation was -0.54 (p<0.01), for centrally located changes.

• The correlation was -0.66 (p<0.01), for periphery located changes.

Discussion – age, change characteristic, and change detection.

• Salient scene characteristics were more responsible for driving attention to change than meaningful change characteristics, especially for older adults.

• The salient changes to objects are quickly detected, but nonsalient changes are detected by slow, serial processing.

Discussion – age, change characteristic, and change detection.

• Older drivers had more difficulty detecting change under most situations adds a new dimensions to the present literature on change detection.• Older and younger drivers showed

differences in detecting changes, but not unsurprising given other findings for age-related differences on many visual search tasks.

Discussion – attentional breadth and change detection

• A strong correlation between breadth of attention and change detection.• A smaller FFOV corresponded to slower chang

e detection.• The breadth of attention plays an impo

rtant role in change detection.• Reducing the number of attentional samples r

equired to detect a change.