Effects of Restricted Interests on the Social Behaviors of Children with Autism Spectrum Disorders

1
Effects of Restricted Interests on the Social
Behaviors of Children with Autism Spectrum
Disorders

• Brian A. Boyd, Ph.D.
• Maureen A. Conroy, Ph.D.
• Peter J. Alter, M.Ed.
• University of Florida
• October 15, 2005

2
What are Restricted Interests?

• All encompassing topics or objects individuals
  with autism pursue with great intensity and
  focus
• Also called
• Circumscribed interests
• Obsessions
• Compulsions
• Special interests
• Narrow interests

3
What are RI (contd)?

• Considered to be a higher form of motoric,
  repetitive behaviors (e.g., hand-flapping)

Lower End Stereotypy
Higher End Restricted Interests, Routines
4
Restricted vs. Preferred?

• How do you distinguish a highly preferred item
  from a restricted item?
• Share 5 characteristics
• Idiosyncratic
• Difficult to redirect child
• Child is intensely focused on interest
• Endure over a long period of time
• Accumulation of mass amounts of information

5
Do RIs change over time?

• Sally Ozonoff asked a group of pre-teens and
  adolescents with Aspergers and HFA about their
  RI
• On average, they had 3-4 RI by that point in
  their lives
• Sometimes they had more than one RI at the same
  time
• Only repetitive behavior found to increase in
  severity over time

6
Do all types of kids on the spectrum have RIs?

• Mmmm!!!!!!!!!
• What we strongly think?
• RIs are more prevalent for students who have HFA
  and are older
• RIs maybe more common for children with
  Aspergers syndrome vs. HFA
• Evidence is really mixed
• RIs may differentiate kids with autism from other
  developmental disorders
• What we know? We know more about lower forms of
  repetitive behaviors
• Children with more severe autism engage in more
  stereotypy
• Children with autism and MR engage in more
  stereotypy self-injury
• HFA also exhibit lower level repetitive behaviors
• To sum up
• We dont know a lot about the relationship
  between functioning level and repetitive
  behavior in general, and in regards to RI, in
  particular

7
Why do students with autism have RIs?

• Mmmmmmmmmm
• Mmmmmmmmmm.
• Lots of Theories
• Arousal theory
• Executive Dysfunction theory
• Perceptual Reinforcement hypothesis

8
Do typically developing (TD) kids have RIs?

• TD kids do engage in repetitive behaviors
  (ritualistic and compulsive)
• Behaviors seem to peak between the ages of 2-5
  (may be the same for kids with autism)
• In TD kids, we see them more during fear-inducing
  situations (e.g., new kid in classroom)
• TD kids engage in repetitive behavior to regulate
  or establish control over their environment

9
What are some common RIs for students with autism?

• Most Common
• Gadgets/devices/electronics
• Power rangers/other action figures
• Dinosaurs
• Video games
• Somewhat Common
• Fantasy/science fiction
• Natural disasters
• Disney characters
• Letters/numbers
• Least Common
• Mythology
• Trains (other than Thomas)
• License plates

10
Should we discourage RIs?

• Repetitive behaviors are hard to extinguish
• It is hard to identify what is maintaining the
  behavior
• It is hard to replace the behavior if it is
  maintained by internal sensory consequences
• If RI is inappropriate for school, then make sure
  student knows when it is ok to talk about or
  engage with their interest

11
Why should we encourage RIs?

• Research studies have shown that using their RI
  gradually decreases the amount of time children
  engage with that item
• Provide children structured and expected
  opportunities to engage with their RI
• Teaches kids a more functional and appropriate
  way to engage with their interest

12
Can RIs be used to encourage social behavior?

• Research suggests it can be used to increase the
  amount of time children engage in peer-related
  social interactions

13
How do RIs encourage social behavior?

• Students with autism appear to be motivated to
  discuss or interact with RIs, either internally
  or externally
• Often they are motivated to engage in 1-sided
  conversations about them,
• OR
• Play with the RI by themselves

14
How could RIs better be used to encourage social
interactions?

• Antecedent-based uses of RI
• Consequence-based uses of RI

15
Overview of RI Research Findings

• Interventions that utilized the RI of children
  with autism have increased their prosocial
  behaviors (e.g., on-task)
• Only 2 studies have addressed the social behavior
  of children with autism (Baker et al., 1998
  Baker, 2000)
• Majority of studies used the RI as a
  consequence-based intervention
• The child is given access to the RI after the
  occurrence of a targeted behavior (Charlop et
  al., 1990 Charlop-Christy Haymes, 1996 1998)
• Studies also have used the RI as an
  antecedent-based intervention
• The RI is used as an antecedent to set the
  occasion for the childs appropriate behavior
  (Adams, 1998)
• Primary problem associated with the RI literature
• Paucity of studies examining the effects of RI
• Lack of systematic identification of the RI

16
Research Questions

• What is the effect of the presence of a
  restricted interest item in comparison to a less
  preferred item on the social behaviors of young
  children with autism spectrum disorders (ASD)?
• What is the generalization effect of the presence
  of a restricted interest item and other tangible
  stimuli on the social behaviors of young children
  with ASD?

17
Method

• Inclusion Criteria
• Preschool or elementary-aged children diagnosed
  with an independently-obtained autism spectrum
  disorder
• Asperger syndrome, Autism, Pervasive
  Developmental Disorder-Not Otherwise Specificed
  (PDD-NOS)
• Use of at least 2-3 word utterances to express
  basic wants and needs
• Ability to initiate to peers social bids using
  gestural or verbal communication
• Display of low levels of appropriate or high
  levels of inappropriate social behaviors
• Display of high levels of engagement with or
  discussion of a RI tangible item

18
Study Participants
19
Phases of Study

• Assessment Phase
• Descriptive assessment
• Preference assessment
• Structural Analysis (SA) Phase
• Concurrent operant condition
• Free operant condition
• Generalization Phase
• Across settings
• Across other tangibles

20
Assessment Phase

• Descriptive Assessment
• Interview parents and teachers using Social
  Skills Interview (Asmus et al., 2004) to identify
  their current levels of social behavior and RI
• For an item to be initially identified as the RI,
  two informants must independently identify that
  item
• Direct observation of participant using Social
  Skills Screening (Conroy et al., 2004)
• Data is taken on the percentage of intervals
  during a 10-min observation period the TC and
  classroom peers engage in social behavior across
  3-5 classroom activities
• Identified activities within classroom context
  where participant exhibited highest lowest of
  social behavior to obtain a baseline measure

21
Assessment Phase

• Multiple Stimulus Preference Assessment (adapted
  from Roane et al., 1998)
• Repeated presentation of 7 identified tangible
  items found during descriptive assessment that
  the child engaged or played with, including the
  hypothesized RI tangible item
• Data was taken on the number of seconds the TC
  physically touched each item
• For an item to be verified as the RI, the TC must
  touch it for the longest duration of time for 2
  out of 3 (67) experimental sessions
• Identified a less preferred (LP) item by asking
  the TC

22
Assessment IOA

• Calculated using a point by point agreement ratio
• Descriptive Assessment IOA collected for 25-85
  of sessions
• JasonMean 96 (range 93-98)
• AllenMean 99 (range 96-100)
• JinMean 88 (range 75-100)
• Preference Assessment IOA collected for 33 of
  sessions
• Jason, Allen, JinMean 100

23
Assessment Results
Descriptive Assessment
Preference Assessment
24
SA PhaseExperimental Procedures

• SA phase (Concurrent Free Operant Conditions)
• All sessions were 5-minutes in length
• Each session was videotaped and coded using
  Tap-IT software for Dell PDAs (Tapp, 2003)
• Peer was instructed prior to each session to NOT
  initiate to TC
• Peer was instructed to always respond to TC
  initiations
• If peer failed to respond, therapist verbally
  reminded him/her
• All peers were typically developing classmates of
  the TC

25
Concurrent Operant Procedures

• Concurrentpurpose was to provide further
  validation of the identified RI and to provide
  preliminary evidence of its effects on the
  display of participant social behavior
• Procedural controlcounterbalanced the peer
  holding the item the order the choices were
  given to the TC by the therapist

26
Concurrent Operant IOA

• Calculated using MOOSES software program (Tapp,
  2002)
• Each observer had to code the same behavior
  within a /- 5-second window of time otherwise
  it was counted as an error
• IOA collected for 33 of sessions
• Percentage of intervals TC choose RI vs. LP
  Jason, Allen, JinMean 100

27
Concurrent Operant Results Choice of Tangible
Items
of 30-second intervals
Jason
Allen Jin
RI
LP
Sessions
28
Concurrent Operant ResultsSocial Interactions
Jason
of time
RI
LP
29
Free Operant Procedures

• Free operantpurpose was to provide a more
  naturalistic play situation to evaluate the
  effect of the RI in comparison to the LP
• Procedural controlrandomly alternated the RI vs
  LP sessions randomly selected 1 peer from
  concurrent to participate

30
Free Operant IOA

• Calculated using MOOSES software program (Tapp,
  2002)
• Each observer had to code the same behavior
  within a /- 5-second window of time otherwise
  it was counted as an error
• IOA collected for 26-36 of sessions
• Duration of social interactions
• JasonMean 100 (range 100)
• AllenMean 98 (range 97-100)
• JinMean 98 (range 95-100)
• Rate of initiations
• Jason AllenMean 100
• JinMean 83 (range
  67-100)

31
Free Operant ConditionSocial Interactions
Jason
Allen Jin
of time
RI
LP
Sessions
32
Free Operant ResultsMean Latency to First TC
Initiation
Jason
Allen Jin
Number of seconds
TC did not initiate
33
Free Operant ResultsTreatment Integrity
Probability of Contingent Peer Responses
Mean Rate of Peer Initiations
indicates the range
34
Indirect Comparison of Descriptive Experimental
Outcomes
35
Generalization Procedures

• Generalizationpurpose was to determine the
  extent to which the observed free operant effects
  generalized to another SETTING other TANGIBLE
  ITEMS
• Procedures same as Free Operant except carried
  out in participants regular classrooms
• Experimenter randomly introduced two classroom
  toys to target child-peer dyad
• Peer from Free Operant participated

36
Generalization IOA

• Calculated using MOOSES software program (Tapp,
  2002)
• Each observer had to code the same behavior
  within a /- 5-second window of time otherwise
  it was counted as an error
• IOA collected for 33-50 of sessions
• Duration of social interactions
• JasonMean 99 (range 99-100)
• AllenMean 94 (range 83-100)
• JinMean 92 (range 98-100)

37
Generalization ResultsSocial Interactions
Jason
Allen Jin
Therapist Prompts
Therapist Prompts
Therapist Prompts
Books
Trains
Peop l e
Dinosaurs
Bugs
Pegs
38
Generalization ResultsTreatment Integrity
Probability of Contingent Peer Responses
Mean Rate of Peer Initiations
indicates the range
39
Teacher Social Validity

• At the conclusion of the generalization sessions
  for each participant their teacher completed a
  Likert-type rating scale
• 1. How comfortable were they with the study?
• M6 (R 6)
• 2. How disruptive was the study?
• M1 (R 1)
• 3. Willingness to allow another child to
  participate?
• M6 (R6)
• 4. How useful was the information obtained from
  the study?
• M5 (R 4-6)
• Overall, how would they rate the intervention?
• M5 (R 4-6)

Not at all (1)
Very (6)

40
Expert Social Validity

• 1 expert in field of ASD, naïve to the
  researchers expectations, viewed randomly
  selected 5-min video clips of children in RI and
  LP sessions
• 1. How appropriate was childs play?
• LP M2 RI M5
• 2. How inappropriate was childs play?
• LP M5 RI M2
• 3. How often did the child play with the peer?
• LP M1 RI M5
• 4. Overall, do you think the child enjoyed
  playing with the
• peer?
• LP M1 RI M5

Not at all (1)
Very (6)

41
Conclusions

• Children with ASD spent more time socially
  interacting when play situation incorporated
  their RI vs LP tangible item
• RI acted as a setting event or establishing
  operation
• Structural analysis served as a viable tool to
  examine the antecedent effects of RI on the
  social behaviors of the participants
• Three-step process provided an effective method
  to validate the RI of participants
• Teacher parent interviews
• Multiple stimulus preference assessment
• Concurrent operant conditions
• Generalization data were more variable
• Therapist prompts had to be introduced to
  facilitate generalization
• Expert Social Validity data reflect clear
  differences in behavior of participants in RI vs
  LP sessions

42
Potential Explanations of Results

• Jasons Undifferentiated Concurrent Operant Data
• Stimulus Control Topography (SCT) Coherence
  Theoryposits that there is not always
  concordance between contingencies arranged by
  experimenter the properties of a stimulus that
  eventually brings the participants behavior
  under stimulus control
• Executive Dysfunctionposits that underlying
  neurological impairments affects the ability of
  individuals with ASD to inhibit prepotent
  behavioral responses engage in alternative
  behaviors
• Variability in Generalization Data
• SCT Coherence Theorycompeting stimuli in
  generalization environment signal occasion for
  other available reinforcers
• Use of Brief Reversal Design
• Difficulty isolating Antecedents occasioning
  Consequences maintaining social behavior in
  natural environment

43
Limitations Future Research

• Limitations
• Generalization concerns
• All children had some social skills prior to
  study
• All children were verbal and on higher end of
  autism spectrum
• SAs were not conducted in the childs natural
  environment
• Small sample size
• Future Research
• Identify the function(s) RI serve for children
  with ASD
• Determine the specific stimulus conditions needed
  for RI to serve as a setting event for social
  behavior in the regular classroom
• Address aforementioned generalization concerns

44
Practical Implications

• Research suggests a starting to place to tap
  social-motivation
• Practitioners can embed RI into structured
  cooperative games or tasks to promote child
  interaction and/or engagement
• Start sharing groups around childs RI

45
Moving Beyond RI Is it Possible?

• Eventually kids may move on because of
  development, they outgrow their interests
• 1 strategy to help kids move on
• Start a sharing group
• Different kids share their interests, including
  the child with autism
• Each child has to learn about the other persons
  interest
• The child with autism learns that other people
  have interests different than his or her own
• The child learns about other peoples interests
• The child learns to share talking and listening
  time around their interest

46
Using RI to Encourage Social Interactions

• 6 steps
• Determine their RI
• Establish a measurable social goal
• Take into account childs play skills
• Developmental level of play
• Stage of play

• Developmental level
• Sensorimotor
• Relational
• Constructive
• Dramatic
• Games with rules
• Stage of Play
• Unoccupied
• Onlooker
• Solitary/independent
• Parallel
• Associative
• Cooperative

47
Example of Direct (Preference) Assessment
48
Antecedent-based uses of RI (contd)

• 3. Embed interest into cooperative games or play
  activities
• Take into account childs skill level and
  developmental abilities
• Can the child match pictures?
• Can s/he sort objects?
• Can s/he read?
• 4. Teach child how to play the game
• Depends on childs social developmental
  abilities
• May need to teach skill in a 11 setting and
  eventually integrate peers into activity
• May be able to immediately introduce activity
  into a small group activity (teacher may need to
  monitor)

49
Example of an Embedded RI Activity
50
Antecedent-based uses of RI (contd)

• Provide child structured and expected
  opportunities to engage with RI game
• Make sure game has a clear ending
• Set limits for when child is allowed to engage
  with game
• Provide choice board
• Daily picture or written schedule
• Monitor the childs progress
• Keep in mind that social relatedness is difficult
  for all kids (HFA or LFA) on the autism spectrum

51
Example of a Choice Board
52
Example of a Daily Schedule

• Written Schedule
• Center time
• Circle time
• Game Time ?
• Small group, reading time
• Game Time ?

53
Any Questions