Biological Motion Processing in Autistic Spectrum Conditions: Perceptual and Social Factors

1
Biological Motion Processing in Autistic Spectrum
Conditions Perceptual and Social Factors Lawrie
McKay1, Jennifer Mackie1, Judith Piggot2, David R
Simmons1, Frank E Pollick11Department of
Psychology, University of Glasgow2Cardiff
University
Background
Experiment 3 Discrimination of Affect from
Biological Motion Displays

• It has previously been observed that people with
  Autistic Spectrum Conditions (ASCs) have
  difficulty processing biological motion relative
  to control participants. (Blake et. al. 2003,
  Moore et. al 1997).
• Is this difficulty due to the perceptual problem
  of integrating motion information across space
  and time, or the cognitive problem of
  interpreting the meaning of a veridically
  perceived motion stimulus?
• We devised a battery of tests which required
  different degrees of perceptual and social
  processing.
• Our ASC group consisted of 4 adult males between
  the ages of 18 and 25 with Autistic Spectrum
  Diagnoses according to the Autistic Diagnosis
  Interview (ADI). Our control group consisted of
  five neurotypical adult males between the ages of
  18 and 25.

• Our third experiment set out to test whether the
  ASC group was able to use social information in a
  similar manner to instrumental information.
• Participants were asked to determine whether
  PLDs of throwing actions were angry, happy,
  neutral or sad. ASC condition n 2.

• Noise tolerance thresholds increased at about the
  same rate for both groups however thresholds are
  lower for the ASC group than for the control
  group(see Figure 3).
• Furthermore at low numbers of signals points the
  ASC group appeared almost completely unable to
  detect the biological motion signals.

Experiment 1 Detection of Biological Motion in
Noise
Experiment 2 Discrimination of Instrumental
Actions from Action Blends

• Paradigm similar to that used by Neri et. al.
  (1998) to determine noise tolerance thresholds
  for detection of biological motion in noise.
• Signal points were taken from point light
  displays (PLDs) generated from a single gait
  cycle of human walkers and noise masks consisted
  of a varying number of scrambled signal points
  (Hiris et. al. 2005).
• The number of walker signal points varied between
  3 and 15(See Figures 1 2).

As can be seen in Figure 5, the proportions
correct for the ASC group are smaller than for
the control group, suggesting a difficulty with
identifying the affect.

• Experiment 2 set out to examine the ASC groups
  sensitivity in action recognition.
• Participants were shown PLDs comprising action
  blends created from two actions, a knock and a
  lift (Pollick et. al. 2005).
• The task was to decide which of the two displayed
  actions was being performed.

Preliminary Conclusions

• For Social Judgements the ASC group performed
  more poorly than controls whilst purely
  instrumental judgements were comparable to
  controls.
• However, the ASC group also showed deficits in
  processing biological motion in noise (Exp 1),
  particularly at low numbers of signal points,
  possibly suggesting increased internal noise in
  the visual systems of the ASC group (Dakin
  Frith, 2005).

As can be seen from Figure 4, there was similar
sensitivity (slope) to the stimuli for both
groups, though the ASC group has a slight bias
towards reporting knock.
Acknowledgements We would like to acknowledge the
Wellcome Trust for their support of Jennifer
Mackie in her vacation scholarship.