1.4 RESEARCH DESIGN AND METHODS
1.4.3 Instruments
1.4.3.1 Sleep
The SSCSI, an 11-item parent-report questionnaire, was used for assessing bedtime and nighttime sleep problems, daytime function, and sleep duration problems (Table 10). The instrument was developed from questions in the SSC medical history that were asked at a clinic visit or by telephone interview of parents within three months of study initiation. The SSCSI is comprised of three subscales: Nighttime Problems,
Daytime Problems, and Sleep Duration Problems. The scoring rubric consists of individual items that are
scored as 0 for “no” and 1 for “yes”. A continuous composite score is made by summing the ten dichotomous items and adding one point for sleep duration that was outside the National Sleep Foundation recommendations for the child’s age (Hirshkowitz et al., 2015). Composite scores range from 0 to 11, with higher scores indicating worse sleep. Scores for the Nighttime Problems and Daytime
Problems subscales range from 0 to 4, and scores for the Sleep Duration Problems subscale range from 0
to 2 (reviewed in Johansson, Feeley, Chasens, in press).
1.4.3.2
IQ measures
Intelligence is operationalized by the IQ measures as characteristics of cognitive abilities (verbal receptive and expressive abilities, nonverbal reasoning, and spatial abilities) and functional skills (processing speed, working memory, motor skills). Four norm-referenced instruments were used to quantify IQ in the SSC sample depending on chronologic and mental age (Table 11). Norm-referenced measures are those standardized by the developers of the measure on a group of individuals chosen to represent the general characteristics (ex. age, sex, ethnicity/race, socioeconomic status, geographical region) of those who will be measured with the instrument. Norm-referenced measures are mapped onto a scale so that scores are a rank that indicate the test-taker’s performance relative to the normalization sample. Instruments are then standardized so that, although raw scores are comparable between instruments, standardized scores are on a common metric across instruments. This is done by mapping
raw scores from the individual test onto a common scale with a total score mean = 100 and standard deviation = 15. Thus, each scale produces a common standardized verbal IQ (VIQ), nonverbal IQ (NVIQ), and full-scale IQ (FSIQ) score that can be used to quantify and compare IQ across instruments. The established cut-off for intellectual disability (ID), described by the Autism and Developmental Disabilities Monitoring Network as, a condition in which “a person has difficulty learning at an expected level and functioning in daily life” (Centers for Disease Control and Prevention, 2016), is an IQ test score ≤ 70.
There are limitations to the standardization and measurement process for intellegence measures. Each measure uses different items to quantify IQ, although they are generally in common domains. IQ is often underestimated in individualas with ASD, frequently because of difficulty completing a standardized test (Bolte, Dziobek, & Poustka, 2009). Discrepencies also often occur between nonverbal and verbal IQ scores in individuals with ASD. A primary reason for this is because of verbal and/or language deficits inherent in the disorder. Thus, tests that rely on language skills may not accurately reflect the child’s actual intelligence and functional abilities (Joseph, Tager-Flusberg, & Lord, 2002). For this reason, NVIQ scores were utilized in the present study, and scores were categorized into five groups. Nonverbal IQs above 70 were divided by the population mean (mean=100). Those below 70 were divided by population standard deviation (SD=15) down to NVIQ=40. Based on recommendations from the American Association of Intellectual and Developmental Disabilities (Tasse, Luckasson, & Nygren, 2013), NVIQ scores below 40 were condensed into one category.
The Differential Ability Scales-II (DAS-II) is a standardized battery of tests measuring a range of cognitive abilities in children with mental age 2 years 6 months through 17-years-old. The DAS-II consists of 20 independent subtests within 3 item clusters measure verbal comprehension, nonverbal reasoning skills, and spatial abilities yielding norm-referenced ability scores, t-scores and confidence intervals, cluster scores, and percentile ranks. Six core subtests contribute to an overall General Conceptual Ability score.(Elliott, 2012; PsychCorp, 2016) The DAS-II has two versions used for younger children and school-aged children. The DAS-II Early Years is divided into a lower level for children with
mental age 2 years 6 months to 3 years 5 months with 4 core subtests, and an upper level for children with mental age 3 years 6 months to 6 years 11 months, and has 6 core subtests. The DAS-II School Age is used for children with mental age 7 years to 17 years 11 months, and has 6 core subtests. Additional subtests are available for each of the three levels to measure things like school readiness, working memory, and processing speed. The SSC primarily used the DAS-II (85.0%), with almost equal proportions using the Early Years (50.9%) and School-Age (49.1%) versions (Elliott, 2012; PsychCorp, 2016).
The Weschler Intelligence Scale for Children-Fourth Edition (WISC-IV) is a standardized intelligence test used in children with mental age 6 years to 16 years 11 months. The WISC-IV measures intellectual ability and cognitive processing with ten core and five supplemental subtests resulting in four index scores: verbal comprehension, perceptual reasoning, working memory, and processing speed. The scale has good reliability in standard samples (Full scale: α=0.88-0.97, subtests: α=0.70-0.90) and special- education samples (full-scale not calculated, subtests: α=0.82-0.93; Williams, Weiss, & Rolfhus, 2003). The scale has been used in ASD samples (Khor, Gray, Reid, & Melvin, 2014; Oliveras-Rentas, Kenworthy, Roberson, Martin, & Wallace, 2013; P. E. Williams et al., 2003).
The Wechsler Abbreviated Scale of Intelligence (WASI) is a standardized intelligence test used in individuals with mental age 6 years to 90 years 11 months. The WASI is a shortened version of the WISC-IV and Wechsler Adult Intelligence Scale, and has the same standardized scoring as the WISC-IV. The WASI has a 2-scale version resulting in VIQ, PIQ, and FSIQ; and a 4-scale version resulting in just FSIQ (Axelrod, 2002; Canivez, Konold, Collins, & Wilson, 2009; Goldstein & Mazefsky, 2013; McCrimmon & Smith, 2012). The WASI has good reliability in a standard sample of children (α=0.81- 0.97) and adults (α=0.84-0.98; The Psychological Corporation, 1999), and excellent reliability in a sample of adults with neurologic or psychiatric disorders (α=0.96-0.98; Axelrod, 2002).
The Mullen Scales of Early Learning (MSEL) is a standardized cognitive test used with children from birth to a mental age of 5 years 8 months. The MSEL produces an Early Learning Composite score which measures cognitive abilities with five subscales: Gross Motor, Visual Reception, Fine Motor,
Receptive Language, and Expressive Language. The Gross Motor subscale is only used for children up to 33 months. Scores are reported as T scores for subscales (mean=50, SD=10) and standardized scores for the full scale (mean=100, SD=15; Bradley-Johnson, 1997). The MSEL has been used with TD children and those with ASD (Akshoomoff, 2006; Bishop, Guthrie, Coffing, & Lord, 2011). The Early Learning Composite has good reliability (α=0.83-0.95; Dumont, Cruse, Alfonso, & Levine, 2000).
1.4.3.3 ASD severity measures
ASD severity was measured with the ADOS, a semi-structured observational instrument used for diagnosis of ASD and classification of severity. The measures that make up the ADOS total score are operationalized as characteristics of social responsiveness, communication, language delay, and spatial ability. Four modules are available with developmentally appropriate tasks for testing individuals with different language abilities: (1) “Preverbal/single words/simple phrases,” (2) “Flexible phrase speech,” (3) “Fluent speech child/adolescent,” and (4) “Fluent speech adolescent/adult (Lord et al., 2000).” Items are scored on a 4-point Likert-type scale from 0 (no abnormality) to 3 (moderate to severe abnormality), and fall into measures of either social affect (SA) or restricted repetitive behavior (RRB). Calibrated severity scores (CSS), standardized for age and language level, are calculated for SA items (CSS-SA) and RRB items (CSS-RRB), and a total ADOS-CSS is the sum of these scores. Children with a total ADOS-CSS score from four to five are classified as having “autism spectrum,” and those with a score of six or more are classified as having “autism” (Gotham, Pickles, & Lord, 2009; Gotham et al., 2008; Hus et al., 2014). The present research team performed standardization of module 4 scores to match prior standardization of modules 1-3 by mapping raw social affect and restricted, repetitive behavior scores onto total, social affect, and restricted repetitive behavior severity scores (V. Hus Bal, personal communication, March 24, 2017).