Children with autistic disorder (AD) display atypical eye contact and struggle with the social imitation of eye contact. Impaired social imitation may be indicative of disruptions in motor learning processes. The application of specific motor learning principles, such as external feedback, may suggest which variables will result in positive change in eye contact. The study aimed to determine the effects of knowledge of performance (KP) and knowledge of results (KR) as types of feedback on the frequency and duration of elicited and spontaneous eye contact in children with AD. A two-phase multiple-probe, multi-treatment (cross-over), single-participant design with a withdrawal component was used. Mixed treatment effects were obtained. Overall effects suggest that KR results in the greatest positive change over a short period of time regarding frequency and duration for both elicited and spontaneous eye contact. This type of feedback seems to be the most effective for spontaneous eye contact. The provision of KP, after elicited and spontaneous eye contact, produced positive effects for duration only. The current Phase 1 evidence suggests that KR (which is goal-directed with fewer additional instructions) may be more beneficial to children with AD. These findings are in accordance with the limb motor learning literature and may therefore support preliminary evidence for disrupted motor learning during eye contact imitation in children with AD.
Children with autistic disorder (AD) experience difficulty in using eye contact to initiate and sustain social interaction (Daou, Vener & Poulson
Imitation in infants and young children serves as a basis for learning and social functions when establishing appropriate social functions within communication interactions (Ingersoll et al.
Children with AD do not understand the significance of eye contact from a social communication perspective (Bayram & Esgin
Motor learning is a set of internal procedures resulting in permanent alterations in movement and response skills through practice (Kim, LaPointe & Stierwalt
Motor learning and operant conditioning underlie Applied Behaviour Analysis (ABA)–motivated interventions (Bhat et al.
Children with AD display atypical eye contact (Daou et al.
Augmented external feedback is an example of such a motor learning principle (Magill & Anderson
The main objective of the study was to determine the effect of different types of feedback when establishing eye contact in a child with AD. A secondary objective was to compare the effect of KP to KR feedback as a motor learning principle, when establishing eye contact in a child with AD.
Treatment was conducted in rooms at the speech-language pathology clinic of a South African university. A clinician was assigned to each of the three participants. Digital video cameras were set up in the corner of each room to record the probes. Equipment used to elicit eye contact included a ball, an iPad, balloons, and soft toys. (The complete elicitation and treatment protocol is available in
The choice of this test as control behaviour was based on its relative stability in cognitive maturation for the duration of the treatment phases (Goodenough
An English male participant aged 5 years 5 months was included in the study. He had previously been diagnosed with AD according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (American Psychiatric Association 2000) by paediatric neurologists. The communication profile of the participant (adapted from Pepper & Weitzman
The participant was recruited via convenience sampling from a speech-language pathology clinic. Hearing screening at the clinic indicated that the participant has normal hearing. He had not received any previous treatment for social or communication difficulties. Individual levels of communication functioning for the participant were not regarded as specific inclusion or exclusion criteria as the objective was to investigate eye contact alone.
The study was approved by a departmental research ethics committee at a South African university. No adverse techniques were applied during intervention and participants were given the option of discontinuing a session or the trial at any time. Participant assent was obtained by means of a child assent form. Parents gave written informed consent for the child to participate. All identifying information of the participant was kept confidential.
A two-phase, multiple-probe, multi-treatment, single-participant design with a withdrawal component was used. Treatments were counterbalanced across phases to control the order effect (Hegde
The speech-language abilities of the participant were assessed and recorded for the purpose of monitoring general progress. These results were not the focus of the present study and are therefore not included here. No treatment was conducted prior to the treatment phases.
The participant received 12 treatment sessions, 6 sessions for each of the 2 phases. Treatment sessions were conducted twice weekly. The first treatment session of each week lasted an hour. Eye contact was addressed throughout this session of play-based communication treatment, through random repetitive elicitation via the elicitation protocol (
Probing procedures The Draw-a-Person Test (Goodenough) was probed as control behaviour prior to and post-treatment as individual control behaviour. The probing protocol was designed according to considerations suggested by the What Works Clearinghouse (Kratochwill et al.
Because of the restricted time frame of the study and the variability in behaviour of children with AD, only two pre-treatment baseline probes for eye contact frequency and duration were obtained prior to the onset of the study during play-based language stimulation sessions. The same eye contact elicitation protocol as for treatment was followed during these pre-treatment probes (
Spontaneous eye contact was measured throughout the 30-min probe.
The digital recordings of both phases of treatment were used for visual analysis and scored by a panel of three independent and untrained viewers. The number and duration of both spontaneous and successfully elicited eye contact occurrences were scored for each participant for each probe session. These eye contact occurrences were scored for KR and KR feedback conditions, respectively. The panel was guided by specific guidelines for scoring (
High inter-rater reliability, with a set criteria of 80% or higher, was therefore established. The results on the recording sheets were analysed in Microsoft Excel (
Data for the participant are presented in
Mean number (a) and duration (b) of elicited and spontaneous eye contact moments.
Effect sizes for participant.
Condition | Elicited |
Spontaneous |
|||
---|---|---|---|---|---|
Frequency | Duration | Frequency | Duration | ||
Phase 1 | KR | Small + | Large + | Large + | Small - |
Phase 2 | KP | Large + | Large - | Small + | Large - |
KP, knowledge of performance; KR, knowledge of results.
The participant showed no improvement in his ability to draw a person. His age-equivalent scores remained stable at <3y3m, <3y3m and <3y3m. Experimental control was maintained.
Maturation, history of event, or the repeated assessment can therefore be ruled out as contributing to possible treatment effects. The participant always required a family member in the room, which was distracting at times. The data for KR probe session 1 and KP probe session 1 were normalised as these two sessions were ended earlier than planned, upon request of the mother. However, as these two probes were not included in the calculation of the two data sets, it did not influence the calculation of effect sizes.
During Phase 1 the participant was provided with KR. The value was 3.67 for successfully elicited eye contact frequency (SD = 1.53), compared to the value of 3.33 (SD = 1.89). The d-index of 0.20 reflects a small effect. Elicited eye contact duration had a value of 1.62 s (SD = 0.62) and a value of 1.19 s (SD = 0.03), resulting in a d-index of 1.12 (large positive effect). KP was provided during Phase 2. The elicited eye contact frequency value of 3.33 was measured against the value of 2.00 (SDpooled = 1.41).
The d-index of 0.94 signifies a large effect. Elicited eye contact duration produced a value of 1.09 s and a value of 1.33 s (SDpooled = 0.14). The d-index of -1.73 signifies a large negative effect.
In Phase 1 (KR) spontaneous eye contact frequency had a value of 75.33 (SD = 12.66) and a value of 27.83 (SD = 10.61). The resulting d-index of 4.01 indicates a large positive effect. Spontaneous eye contact duration produced a value of 2.38 (SD = 0.24), and the value it was measured against was 2.52 s (SD = 0.83). The d-index was -0.29 (small negative effect). For Phase 2 (KP) spontaneous eye contact frequency produced a value of 69.33 and a value of 61.00 (SDpooled = 0.14). The d-index of 0.24 reflects a small positive effect. The duration of spontaneous eye contact had a value of 1.72 s and a value of 2.19 s (SDpooled = 0.33).
These values resulted in a d-index of -1.40 (large negative effect).
The KR phase reflected small negative (elicited frequency) and large positive (elicited duration and spontaneous frequency) effect sizes. A small decline from baseline conditions was calculated for the duration of the spontaneous eye contact moments. KR was effective in increasing successfully elicited eye contact behaviour. The small decline in spontaneous eye contact duration is challenging to interpret. As there was a large increase in spontaneous eye contact frequency, it is possible that the measures chosen precluded accurate measurements of therapeutic effects.
The participant may have displayed more frequent eye gaze responses for shorter periods, which could have led to increased overall looking time per session. This measure was not within the scope of the current study. If accurate, the current findings would support the longstanding reporting of Kendon and Cook (
The study aimed to determine and compare the effect of KP and KR on spontaneous and elicited eye contact frequency and duration. Mixed treatment effects were obtained for the different phases. Overall effects suggest that KR feedback results in the greatest positive change over a short period of time regarding frequency (elicited and spontaneous) and duration (elicited) of eye contact. Furthermore, providing this type of feedback seems to be the most effective when eye contact has been established spontaneously. This is positive, because spontaneous eye contact is more important for social competence (Clifford & Dissanyake
The current Phase 1 evidence (Beeson & Robey
A number of factors that may influence results of controlled single-participant designs are encountered when working with children with AD: eye gaze, individual levels of arousal, attention and emotions. Most of these factors could be related to methodological constraints. A criticism for using verbal reinforcement for children with AD is that there are many paralinguistic variables that have to be processed and fixating on a specific part of the message may result in failure to receive the intended message (Oostyen
The authors acknowledge the National Research Foundation of South Africa for providing funding for this project (Project Number: 92711).
The authors declare that they have no financial or personal relationships which may have inappropriately influenced them in writing this article.
S.G. designed the study, set up the treatment protocol and compiled the article. M.M., A.V.d.M. and M.S. contributed to the data collection and interpretation of the results. E.S. assisted in the writing of the article.
Treatment design with different phases and probes.
Treatment Phase 1 |
Treatment Phase 2 |
||
---|---|---|---|
Treatment type/condition | Description | Treatment type/condition | Description |
Knowledge of Results | ‘Hey! Good looking at my eyes! Thank you!’ | Knowledge of Performance | ‘Try looking at my eyes for longer’. |
Once during pre-treatment. Once during Phase 1. |
Once during Phase 2. |
KP, knowledge of performance; KR, knowledge of results.
Treatment protocol for knowledge of results and knowledge of performance conditions.
Time of session | Knowledge of results | Knowledge of performance |
---|---|---|
Meet and greet: Model appropriate verbal and non-verbal greeting pragmatics and language. Orientation to the session: Explanation of the temporal layout of the session using a visual time table. Phonological awareness: Book reading and rhyming activities in order to provide exposure to the sound structure of English. According to the hierarchy of Bernthal, Bankson and Flipsen ( |
Meet and greet: Model-appropriate verbal and non-verbal greeting pragmatics and language. Orientation to the session: Explanation of the temporal layout of the session using a visual time table. Phonological awareness: Book reading and rhyming activities in order to provide exposure to the sound structure of English. According to the hierarchy of Bernthal et al. ( |
|
Conduct treatment and maintenance probes as discussed in Turn taking Requesting Protesting Focused stimulation of target words (Roth & Worthington 2011) |
Conduct treatment and maintenance probes as discussed in Turn taking Requesting Protesting Focused stimulation of target words (Roth & Worthington 2011) |
|
Naturalistic Language Intervention: The clinicians will provide language elicitation and stimulation using a common theme with graded levels of difficulty specific to the level of each participant. Literacy Pre-literacy activities:
Reading: Holding book right side up, tracking from left to right page, distinguishing between pictures and print, showing sustained interest during book-reading, recognises own name. Writing: Scribbling, correct writing posture, copying line, circle, cross, block, triangle, diamond, tracing letters, copying letters and words, writing own name. Phoneme–grapheme coupling will be targeted by implementing the phoneme song in combination with visual stimuli and writing task, in accordance with each participant’s level of literacy. These activities will be individualised to each participant. Greeting Model-appropriate verbal and non-verbal greeting pragmatics and language. Treatment of eye contact will occur throughout all three activities. Every time the participant makes spontaneous eye contact, KR feedback will be provided. In addition, a ball will briefly be placed in front of the clinician’s face every 3 min to elicit eye contact. Every instance of eye contact will be followed by: ‘Hey! Good looking at my eyes! Thank you!’ A continuous feedback schedule will be followed (Roth & Worthington 2011). |
Naturalistic Language Intervention: The clinicians will provide language elicitation and stimulation using a common theme with graded levels of difficulty specific to the level of each participant. Literacy Pre-literacy activities:
Reading: Holding book right side up, tracking from left to right page, distinguishing between pictures and print, showing sustained interest during book-reading, recognises own name. Writing: Scribbling, correct writing posture, copying line, circle, cross, block, triangle, diamond, tracing letters, copying letters and words, writing own name. Phoneme–grapheme coupling will be targeted by implementing the phoneme song in combination with visual stimuli and writing task, in accordance with each participant’s level of literacy. These activities will be individualised to each participant. Greeting Model-appropriate verbal and non-verbal greeting pragmatics and language. Treatment of eye contact will occur throughout all three activities. Every time the participant makes spontaneous eye contact, KP feedback will be provided. In addition, a ball will briefly be placed in front of the clinician’s face every 3 min to elicit eye contact. Every instance of eye contact will be followed by: ‘Try looking at my eyes for longer!’ A continuous feedback schedule will be followed (Roth & Worthington 2011). |
KP, knowledge of performance; KR, knowledge of results
Operational definitions for scoring eye contact.
SCORING SHEET FOR VIDEO ANALYSIS |
|
---|---|
Variable | Operational definition |
Eye contact episode duration | Time in seconds during which the child maintained eye contact with the clinician. |
Eye contact frequency | Number of eye contact episodes. |
Spontaneous eye contact | Eye contact initiated by the participant, in the absence of elicitation. |
Elicited eye contact | Eye contact prompted by an object briefly placed at the clinician’s eye level. |