Inhibitory abilities in infants. Evidence from eyetracking

1
Inhibitory abilities in infants.Evidence from
eye-tracking
Ágnes M. Kovács Jacques Mehler SISSA, Trieste
Budapest, 2005
2
We infer what infants know from their behavior

• However, the data is very contradictory
• Object permanence (OP)
• at 2.5 months infants demonstrate OP on looking
  tasks (Spelke et al, 1992)
• at 7 months do not demonstrate OP on reaching
  tasks (A not B) (Diamond, 1985)

3
The A not B tasks

• 1. Reaching tasks

• Object hidden in loc A for several times (A, A,
  A)
• Once hidden in loc B
• Delay of 2 seconds
• 7-mo-olds reach to A
• (perseveration to A!)
• By 12 months infants make an A not B error at a
  10 sec delay (Diamond, 1985)

A
B
A
A
4

• 2. Looking tasks (VoE)

• Object hidden in A for several times (A, A, A)
• Once hidden in B
• Delay of 10 seconds
• Object is retrieved from A
• 8-mo-olds show surprise (look more), but look
  less if it is retrieved from B
• (no perseveration!)
• (Baillargeon, 1991)

B
A
A
A
5
Looking data is in conflict with the manual
reaching dataDo 7-months-old infants HAVE or
HAVE NOT OP?

• Explaining the conflict in data
• Performance accounts
• 1. Deficit in inhibiting a habitual response
  (OP)
• 2. Deficit in means to end abilities
  (OP)
• Competence account
• 3. Graded object representation (OP - )
• (Munakata et al 1997)

6
Aim of the present study

• Try to decide between the 3 major theories by
  studying anticipatory eye movements in an A not B
  task
• The 3 theories have different predictions for eye
  movements, because
• eye movements are
• 1. Automatic (motor response)
• but
• 2. Do not require means to end skills

7
Study 1 A not B with eye movements

• Predictions of the theories
• 1. Inhibition account A not B effect
  
• (after several fixations to A they have
  to inhibit the habit to look to A and look to B
  instead. Less anticipations on B trials)
• 2. Means-end account no A not B effect -
• (no means to end demands)
• 3. Graded representation no A not B effect -
• (as the trials go on, the objects
  representation gets stronger, anticipations
  should grow. Similar anticipation pattern on the
  A and the B trials)

8
TOBII eye tracker

• Non-intrusive Eye-Tracking
• Eye-Tracking is almost invisible to the user
• Infant addon
• 0,5 degrees accuracy
• head- movement compensation and drift reduction ?
• binocular
• 50 Hz sample rate

9
Attention getter
Attention getter 1 (400 ms)
10
Attention getter
Attention getter 2 (400 ms)
11
No object Measuring anticipatory eye movements
(1000ms)
12
Object in location A (1000 ms)
A
B
13
Delay/ black screen - 4000ms

• 5 repetitions of the sequence

14
Attention getter
Attention getter 1 (400 ms)
15
Attention getter
Attention getter 2 (400 ms)
16
No object Measuring anticipatory eye movements
(1000ms)
17
Object in location B (1000 ms)
KUKU kuku kuku!!!
18
Delay/ black screen - 4000ms

• 5 repetitions of the sequence

19
Method

• Procedure
• Attention getter -400/400ms
• Empty squares -1000ms
• Object in location A - 1000ms
• Delay/ black screen - 4000ms

• 5 repetitions of A then 5 repetitions of B
• counterbalanced for side
• Subjects 9 7.5-month-olds
• Data analysis
• Criteria for accepting blocks
• Eye gaze data on minimum 4 trials in the block
  in the AB and BA conditions

20
(No Transcript)
21
Preliminary results
AB vs BA p 0.03
22
Increase in anticipations - NODecrease by the
5th trial- YES (4th vs. 5th trial p 0.06)
Is the effect due to variability in looking?

• Decrease in anticipations
• NO

23
Preliminary conclusions Study 1

• The A not B effect in eye movements supports the
  inhibitory account 7 month-old infants possess
  an OP but can not inhibit very well a prepotent
  motor response (oculo-motor in our case)
• The results does not confirm the predictions of
  the means-end and graded representation account
• FURTHER QUESTION
• Are infants poor only in inhibiting habitual
  motor schemas?
• ?or also in inhibiting prepotent MENTAL
  schemas???

24
THANK YOU