AGRUPACIÓN A PIE
ESCUELA DE ESPECIALIDADES “ANTONIO DE ESCAÑO”
All participants underwent series of assessments that can each be categorised into one of
the following: background history questionnaires, descriptive behavioural measures, and
primary outcome measures.
3.2.3.1.Background History Measures
Parents were asked to complete a packet of questionnaires in order to provide detailed
information regarding their child’s development as well as familial history.
5 The study design has been discussed in Chapter 2 in detail. A brief summary is provided here for the purpose of completeness of the chapter.
The Early Developmental History Questionnaire (Golding, 2009) is a non-standardised
clinical support tool aimed at collecting early development information. The questionnaire
targets six broad domains: pregnancy, birth and first months, language development, social
development, self-help skills and motor development, play and use of imagination, and other
behaviours. Each domain focuses on development from birth up to 36 months of age. This
questionnaire was given to parents before the T1 assessments in order to collect information
regarding birth and development during the first year of life. This was primarily conducted to
assess exclusion criteria of prematurity at birth and major birth complications6.
Family History Questionnaires were completed before the T1 assessment and included
four questionnaires: mother, father, sibling, and participant history (see Appendix C). Each
questionnaire included questions regarding the individual’s development, schooling details,
details of any major health complications, along with details of any family history of a mental
health and/or developmental disorder. These questionnaires were developed with the purpose
of informal data collection of clinical research information for a study of infant and child
siblings of individuals with autism at the University of California, San Diego.
3.2.3.2.Autism Screening Measures
Two Level 1 autism screening measures were used: Social Communication
Questionnaire (SCQ; Rutter, Bailey, & Lord, 2003) for children 4 years and above, and the
Quantitative Checklist for Autism in Toddlers (QCHAT; Allison et al., 2008) for children 18
to 24 months. Both screening questionnaires assess the presence of social and communication
skills and any restricted, repetitive behaviours or interests. A score of 15 on the SCQ and 50
6 Two parents reported birth trauma but upon enquiry one parent reported that they considered forceps delivery as birth trauma even though there was no trauma to the child due to forceps use. The other parent reported that
on the QCHAT are the recommended cut-off criteria for Autism Spectrum Disorders
symptomatology.
In the study, screening measures were administered based on the age of the participant.
However, the SCQ has been found to have high sensitivity (93%) for children from 2 years to
6 years of age (Allen, Silove, Williams, & Hutchins, 2007). Thus, it was administered to all
children over 2 years. All parents completed the screening questionnaires and all children in
the final sample scored high on both the SCQ and the QCHAT (Table 3.1).
3.2.3.3.Descriptive Measures
3.2.3.3.1. Mullen Scales of Early Learning
Mullen Scales of Early Learning (MSEL; Mullen, 1995) is a standardised
developmental assessment battery for children from birth to 68 months of age with five
subdomains: Gross Motor, Fine Motor, Visual Reception, Receptive Language and
Expressive Language. Each scale comprises of interactive tasks. Some tasks involve parental
input and assistance. Items are presented in a hierarchical order of difficulty with basal
criterion of passing three consecutive items and ceiling criterion of three consecutive zeros.
Raw performance scores are converted to T-scale scores for each subscale. Raw scores for the
four cognitive subscales (visual reception, fine motor, receptive and expressive language) can
be summed and converted to an Early Learning Composite Standard Score that offers a
measure of overall cognitive functioning.
For the purposes of this study, the four subscales: fine motor, visual reception, receptive
and expressive language were administered, as cognitive and language functioning were of
primary interest. However, T-scale scores could not be obtained for all children due to
domains was computed. The MSEL is a stable measure of verbal and non-verbal abilities in
ASD and has high convergent validity with other measures of cognitive assessment such as
the DAS (Bishop, Guthrie, Coffing, & Lord, 2011). It was, therefore, used to calculate Verbal
and Non-Verbal Mental Age, respectively. Mean age equivalents for fine motor and visual
reception scales were calculated to obtain a Non-Verbal Mental Age (NVMA) and mean of
age equivalents for receptive and expressive language were used to obtain a Verbal Mental
Age (VMA) for each child (see participant characteristics Table 3.1).
3.2.3.3.2. Autism Diagnostic Observation Schedule – Generic
The Autism Diagnostic Observation Schedule – Generic (ADOS-G; Lord et al., 2000),
is a semi-structured play-based observational, standardised assessment that measures
symptoms associated with Autism Spectrum Disorder. Four different modules have been
developed and one module is given per participant and is chosen based on verbal ability and
chronological age of the individual. Each module assesses the individual on four primary
domains of reciprocal social interaction, language and communication, stereotyped
behaviours and restricted interests, and play/imagination.
An administrator trained to research reliability, blind to group assignment of the
participants, administered the ADOS at both intake and follow-up assessment. All children,
except two, received Module 1 during the intake assessment process, as verbal ability of
participants comprised of single words. Two children received Module 2, as they had
developed phrase speech. All children in the final sample (n = 24) met the criteria for ASD
3.2.3.3.3. Vineland Adaptive Behavior Scales – Second Edition
The Vineland Adaptive Behavior Scales (VABS-2; Sparrow, Balla, & Cicchetti, 2005)
is a semi-structured parent interview that assesses communication, daily living skills, social,
and motor skills in individuals aged from birth to 90 years. The VABS is a standardised
assessment tool that provides age equivalents and standard scores for each of eleven
subscales, as well as overall adaptive functioning.
The VABS was conducted either face-to-face or on the phone with a parent. Of the final
sample of 24 participants, parents of 21 children completed the VABS. The first three
children were enrolled in the study during the proof-of-concept stage where VABS was not
included in the assessment battery. The Adaptive Behavior Standard Scores for children in the
Treatment group were comparable to those of children in the Wait-List Control group (Table
3.1)
3.2.3.4.Primary Outcome Measures
3.2.3.4.1. Unstructured Imitation Assessment
The Unstructured Imitation Assessment (UIA) is an adaption of an assessment
developed by McDuffie and colleagues (McDuffie et al., 2007); and has been used in research
on the effects of RIT previously (e.g. Ingersoll, 2010b). The UIA is a play-based assessment
of spontaneous object and bodily-gesture imitation in an unstructured setting. The UIA is
conducted in a socially interactive manner, whereby the examiner engages in free-play with
the child in a room full of two sets of several developmentally appropriate toys. The examiner
then alternates imitating the non-verbal behaviour of the child and modelling actions for the
child to imitate. While modelling actions to the child, the examiner verbally describes each
examiner’s actions. There are two scales derived from the UIA: object and gesture imitation
scales (Appendix E). Each scale is composed of ten models with each model presented to the
child three times irrespective of presence or absence of an imitative response. The child’s
responses are scored on a scale of 0-2, where ‘0’ reflects either no response or an incorrect
response, ‘1’ for partial correct imitation, and ‘2’ for complete correct imitation. The highest
score on each scale is 20 and the total score on the UIA ranges from 0-40.
The UIA was the primary outcome measure to evaluate the effect of RIT on
spontaneous imitation skills of children with autism in the context of an unstructured play
setting. The UIA was administered to participants at both T1 and T2 by an examiner who was
blinded experimentally to the group assignment of the participants. Three blinded independent
observers analysed all UIA videos. In order to calculate inter-rater reliability, an interclass
correlation coefficient (ICC) analysis was employed as scoring used a Likert scale and
weighted Kappa can be used only for two observers while this study had three observers
(Hallgren, 2012). Norman and Streiner (2008) have demonstrated that weighted kappa with
quadratic weights for ordinal scales gives identical values to single measures ICC and
therefore the two can be used interchangeably (Hallgren, 2012). Inter-rater reliability was
assessed for 25% of the videos using a two-way, mixed consistency single-measures ICC and
was found to be .99. This indicates excellent agreement between the independent coders and
minimum measurement of error.
3.2.3.4.2. Structured Imitation Assessment
The Structured Imitation Assessment (SIA) was adapted from the Pre-school Imitation
from which the SIA has been adapted, is one of the only measures of imitation that has been
standardised on both typical and autism populations (Vanvuchelen et al., 2011b). The
drawback of the PIPS is that it has been standardised only on a Dutch population and English
translations of the measure were unavailable. Therefore, an adapted version of the PIPS,
based on the descriptions provided in Vanvuchelen et al. (2011b), was used.
The SIA is a structured assessment measuring elicited or prompted imitation in children
with autism. This 30-item assessment measures four different aspects of imitation: single
bodily imitation, sequential bodily imitation, goal-directed procedural imitation, and non-goal
directed procedural imitation (for definitions see Chapter 1, Section 1.1.2.1.1).
Simultaneously, it measures three different kinds of imitation: action-on-object, gestural and
facial imitation. Scoring ranges from 0-4 where some items were scored on a three-point
Likert scale (0-2), others from 0-3, and a few on a five-point scale from 0-4 (Appendix E).
According to the guidelines, three practice tasks are administered at the beginning of the
assessment. The aim of these practice tasks is to help the child understand the nature of the
assessment and what is expected. Each practice task is administered three times, and every
time the child does not respond or gives an incorrect response the examiner physically and
verbally prompts the child to complete the task. Test items are administered only once, with
no verbal or physical prompt provided to the child upon no response. The examiner presents
the task, gives a brief instruction, “You do it”, and waits for five seconds for child to respond.
No reinforcement is given for imitation.
The SIA was administered by an experimentally blinded examiner at T1 and T2. Three
blinded independent observers analysed all of the SIA videos. For the UIA, a two-way, mixed
25% of the videos. ICC was found to be .89. This indicates excellent agreement (Cicchetti,
1994) between independent coders and minimum measurement of error.