Eye Movement Strategies In Navigational Tasks

1
Eye Movement Strategies In Navigational Tasks

• Austin Ducworth
• Melissa Falzetta
• Lindsay Hyma
• Katie Kimble
• James Michalak

2
Introduction

• Navigation involves keeping track of ones
  current location and planning routes to new
  destinations.
• Maps are the most common navigational tool.
• Coordinating reference frames is a key difficulty
  in navigational tasks

3
Individual Differences
4
Cardinal Direction Tasks

• Prove to be more difficult than other
  navigational tasks
• Prone to show wide individual differences in
  performance

5
Set-Up

• Started w/ Dr. Ds GCI code.
• Removed the pre-defined duration within the
  scenario (.tri) file.
• Checked for a key press event for the operator to
  switch between images.
• Modify the margins so the 640x299 and 350x340
  images were centered within the 640x480 display
  window.

6
Analysis

• Objective We wanted to turn the raw coordinate
  and time data into interpretable output.
• Problem The analysis program included in Dr. Ds
  gci code overlaid the scanpath over the image,
  however, it did not show the eye movements in
  sequence

7
ANALYSIS Approach

• Approach Write a java application to read in por
  data and write the sequence of roi hits for each
  image.

8
ANALYSIS 1st Attempt

• Parse through the por data and determine if the
  coordinate represent a point within pre-defined
  ROIs.
• If a point lies within an ROI, a hit and the
  duration is recorded in an output file.
• PROBLEMS
• Defining ROIs
• Calibration
• Inconclusive Results

9
ANALYSIS 2nd Attempt

• Divide all points within the image into a grid
  of user specified proportions.
• PROBLEMS
• Data Analysis Lengthy
• Not Enough Testing

10
Sample Output
11
ANALYSIS 3rd Attempt

• Analyze the scanpath using the analysis program.
• PROBLEMS
• Sequence must be guestimated
• ADVANTAGES
• Quick Analysis
• Previously Tested Results
• Ass Saving Technology

12
Experiment

• Between subjects design
• 2 strategies
• Heading reference strategy
• 3D First strategy
• 8 participants (4 per condition)
• Excluding 1 participant for failure to follow
  instructions

13
Design

• Initial training
• Strategy training
• 24 practice trials
• Eye movement trials

14
Strategy Training
15
Eye Movement Trials

• Individual map
• Table knock

Answer Screen
16
Analysis

• Each scanpath was superimposed on its respective
  map
• Analyzed eye movements
• Looked for expected patterns to emerge

17
(No Transcript)
18
Analysis Expected Patterns

• 3D First
• First look at region 12 and then to the parking
  lot with the cars (regions 16 or 17) to form an
  angle.
• Then Transpose that angle to the right side of
  the map, mentally rotate the angle to coincide
  with the heading direction of the airplane
  (regions 15 and 20).

Heading Reference First expected to look at
regions corresponding to the map arrow (regions
15 and 20) Then look at regions 2,7.or 12 to
orient themselves on the 3D map. They look to
the left and right of these regions (regions 6,11
and 7) to determine the direction of these
parking lots
19
3D First
Heading Ref.
20
Performance Results

• Heading reference 76.4 correct
• 3-D First 66.7 correct
• Not a notable performance difference between
  strategies

21
Strategy Use Results

• Heading Referencing used strategy on 50 of
  trials.
• 3-D First used strategy on 74 of the trials.
• One sample t-test a significant number of
  participants did not use the strategy

22
Qualitative Observations

• Some subjects did not have eye movements over the
  arrow showing airplane direction.
• 3-D First strategy Some subjects exhibited a
  visual angle formation in their scanpaths while
  others may have been forming an angle mentally
  rather than via eye movements

23
Discussion

• Did not show a significant difference in
  performance depending on strategy used
• Did not use strategy taught a significant
  percentage of time

24
Recommendations

• Use a more objective strategy for data analysis
• Use less training more mistakes may give more
  insight into a superior strategy
• Use more sensitive measures response time

25
Questions??