More research has been conducted using the CBCL than any of the other instruments included in this domain. The PDDBI is the only instrument in this domain that was developed specifically for measuring problem and pro-social behaviors among persons with ASD. PDDBI test developers found that the measure is both reliable and valid and is useful in providing information not typically available in most instruments used to assess children with pervasive developmental disorders (Cohen, et al., 2003). Recent studies with the PDDBI have supported the PDDBI as a measure of ASD symptoms (Schutte, Devlin, Richardson, Hill, & Hewitson, 2019) and one that provides profiles that may predict both autism severity and adaptive behavior trajectories (Cohen & Flory, 2019). However, some research has raised concerns regarding the validity of some PDDBI scores (McMorris & Perry, 2014).
Though there is no “typical” pattern of emotional and behavioral functioning in ASD, some assessment score patterns have been found. For example, scores on the BASC-2’s Behavior Symptom Index (BSI) were significantly higher for children and adolescents with autism, with the exception of Inattention; and also significantly lower for the Adaptive Skills Composite and all its subscales, with the exception of Adaptability (Goldin, Matson, Konst, & Adams, 2014). Ellison et al. (2016) found that several BASC-2 subscales are consistently clinically elevated for individuals with ASD. These include Atypicality, Withdrawal, and all Adaptive Skills subscales. On the CBCL, significant elevations among those with ASD consistently emerged for the Social, Thought, and Attention Problems subscales, and the Withdrawn/Depressed scale was elevated to at least the borderline clinical range for half of the ASD symptoms (Mazefsky, Anderson, Conner, & Minshew, 2010).
Using information from multiple raters on emotional/behavioral assessment measures is critical to obtaining the most comprehensive functioning of individuals with ASD. With regard to parent-teacher concordance of BASC-2 scores for children and adolescents with ASD, parent ratings were generally more elevated than teachers’, though they were only significantly different on the Adaptive Skills Composite (Lane, Paynter, & Sharman, 2013). Similar results were found by Ellison et al. (2016), who underscored the additive value of teacher BASC-2 scores.
Surprisingly few studies have been conducted specifically to evaluate the utility and validity of using broadband emotional/behavioral assessments, which are not normed on ASD populations, to identify the presence of affective problems in autism. Hoffman and colleagues (2016) determined in their research that the CBCL is useful for evaluating co-occurring emotional/behavior problems in ASD, and a similar conclusion was drawn based on findings from a study conducted with the CBCL/1 ½-5, the version for preschool children (Pandolfi, Magyar, & Dill, 2009). However, though while the CBCL’s sensitivity is good for detecting co-occurring emotional/behavioral problems in autism, its specificity is low (Pandolfi, Magyar, & Dill, 2012). Later, however, Magyar and Pandolfi (2017) demonstrated that the CBCL’s DSM-Oriented Affective and Anxiety Problem Scales can be used to screen for depression and anxiety in youths with ASD. Though these findings are promising, clinicians must avoid overinterpreting scores or relying solely on broadband measures when assessing for emotional/behavioral problems in persons with ASD. Instead, use of these measures in conjunction with other data is urged (Hoffman et al., 2016; Pandolfi et al., 2012).
Several studies suggest that broadband measures included in this section also may identify patterns that suggest the need for an autism-focused evaluation. For example, Volker et al. (2010) concluded that the BASC-2’s Developmental Social Disorders (DSD) scale was highly effective in differentiating between children with high functioning ASD and typically-developing peers. Similarly, with younger children (24-63 months), Bradstreet and colleagues (2017) found that the DSD scale on the BASC-2 Parent Rating Scale- Preschool had adequate sensitivity and specificity values when distinguishing those with ASD from those with no diagnoses but not when differentiating between those with ASD and with other diagnoses. In a study specifically examining the Teacher Rating Scales of the BASC, Hass and colleagues (2012) found that teacher ratings discriminate between students with an educational classification of autism and their nondisabled peers, though teacher scores tended to be in the “at-risk” as opposed to the “clinically significant” range.
Much research regarding the use of emotional/behavioral assessments as indicators of possible ASD have focused on the CBCL. Specifically, Sikora and colleagues (2008) found CBCL/1 ½-5 (downward extension of the CBCL/6-18) subscales to have better sensitivity and specificity in identifying young children with ASD than the Gilliam Autism Rating Scale (GARS). However, a recent study conducted by Havdahl and colleagues warned that the use of CBCL profiles for ASD-specific screening would likely result in many misclassifications, as acceptable specificity could only be achieved for school-age children with below clinical levels of emotional/behavioral problems. Research conducted by Hoffman et al. (2016) concluded that although the CBCL is not a suitable screening instrument for the identification of ASD, high scores on the syndrome scales Social problems, Withdrawn, Thought problems, and Attention problems might be an indication for further and differential diagnostic procedures. Rescorla and colleagues’ (2019) recently completed study with the CBCL/1 ½- 5 confirmed other previous research with this version of the CBCL that the Pervasive Developmental Problems scale and the Withdrawn scale differentiated well between children diagnosed with ASD and those not diagnosed. Concerns about adequate sensitivity and specificity of the PDDBI for ASD diagnostic screening have been raised by Reel and colleagues (2012).
It is helpful to understand that there are also weaknesses in emotional/behavioral assessment tools. Hosp and colleagues (2003) examined the structure of items on commonly used behavior rating scales and found that most scales included negative-action questions and lack-of-action questions, neither of which is useful for assessing positive behaviors and for addressing observable, measurable behaviors.