Overview
The Stanford-Binet Intelligence Scales – Fifth Edition (SB-5; Roid, 2003) is an individually-administered, norm-referenced instrument that measures intellectual and cognitive ability, yielding not only a Full Scale IQ but also a Brief IQ, Verbal IQ, Nonverbal IQ, as well as five factors: Fluid Reasoning, Quantitative Reasoning, Visual-Spatial Reasoning, Working Memory, and Knowledge. The SB-5 can be used to assess individuals from 2 through 85 years of age and older. In addition, the Early SB-5 provides a lower-cost version of the test for preschool assessment. The SB-5 is widely used and, because of numerous high-end and low-end items can be useful for measuring individuals with scores in both the giftedness and intellectual disability ranges; more specifically, its Extended IQ scale permits calculation of FSIQ higher than 160 and lower than 40. Testing begins with administration of “routing” tests in Nonverbal Fluid Reasoning and in Verbal Knowledge, which helps examiners determine the appropriate start points for examinees. Half of the SB-5 has nonverbal content requiring little to no verbal responses from the examinee, making this test useful for assessing persons with Limited English Proficiency, deaf and hard of hearing conditions, nonverbal learning disabilities, ADHD, traumatic brain injury and autism spectrum disorder. This test can be hand-scored or scored with the optional SB-5 Optional Scoring and Report system.
Summary
Age: Stanford-Binet Intelligence Scales - Fifth Edition (SB5)/Roid (2003)
Time to Administer: 2 years 0 months to 85 years +
Method of Administration: Individually administered, norm-referenced assessment of cognitive abilities; administered by clinician
Yields standard scores (M = 100, SD = 15) scaled scores (M = 10, SD = 3), percentile ranks, confidence intervals, age equivalents, and change sensitive scores
Subscales: Verbal, Nonverbal, and Full Scale IQs; Index scores for each of the five Nonverbal and Verbal cognitive factors: Fluid Reasoning, Knowledge, Quantitative Reasoning, Visual-Spatial Processing, and Working Memory
Autism Related Research
Twomey, O’Connell, Lillis, Tarpey, & O’Reilly (2018)
Age Range: 3-5 years
Sample Size: 58, with ASD
Topics Addressed:
Utility of SB-5 abbreviated IQ (ABIQ) score in estimating FSIQ for young children with ASD
Outcome:Twomey, O’Connell, Lillis, Tarpey, & O’Reilly (2018)
For preschoolers with ASD, correlations between ABIQ and FSIQ were strong, supporting the utility of the ABIQ for estimating FSIQ in this group. Notably, the SB-5 ABIQ performed better in the estimation of FSIQ than similarly abbreviated versions of the WISC-III. However, clinically significant over-estimation occurred in 17.5% of cases in this study. Considered in conjunction with other larger studies, authors noted that the overestimation of FSIQ by ABIQ is most marked when overall cognitive abilities are particularly low and/or when there are significant differences between scores on the two routing tests comprising the ABIQ.
Overall conclusion: the SB-5 ABIQ is likely to have good utility in estimating FSIQ for ~80% of young children with ASD, but for ~20% may overestimate. Therefore, caution in using the ABIQ for children with very low cognitive ability.
Baum, Shear, Howe, & Bishop (2015)
Age Range: 10–16 years
Sample Size: 40
Topics Addressed:
Convergent validity between SB-5 and WISC-IV
Outcome:Baum, Shear, Howe, & Bishop (2015)
Corresponding intelligence scores were highly correlated
(r = 0.78 to 0.88), but full-scale intelligence quotient (IQ) scores (t(38) = −2.27, p = 0.03,d = −0.16) and verbal IQ scores (t(36) = 2.23, p = 0.03; d = 0.19) differed between the two tests.
Conclusion is that comparison of IQ test scores for persons with ASD (and other special groups) are important, as it cannot be assumed that convergent validity findings in typically developing children and adolescents hold true across all pediatric populations.
Sansone, Schneider, Bickel, Berry-Kravis, Prescott & Hessl (2014)
Age Range: Age Range (in years as given in study) 3–38 years
Sample Size: 205 With ASD and 106 with fragile X syndrome
Topics Addressed:
Problems in estimating IQ for low functioning individuals
Outcome:Sansone, Schneider, Bickel, Berry-Kravis, Prescott & Hessl (2014)
Traditional methods for generating IQ scores in lower-functioning individuals with intellectual disability (ID) are inaccurate due to floor effects, leading to erroneously flat profiles
Using the SB5, the researchers found that the standard method of converting raw scores to standardized scaled, index, and IQ scores was not meaningful in ID children. Rather, calculating deviation scores yielded better distributions of scores (i.e., more accurate) and stronger correlations with adaptive behavior, vocabulary, and nonverbal reasoning.
Though deviation scores and standardized scores were highly correlated, they are not equivalent scores. The use of the deviation z-score method improved sensitivity of the test through addressing flooring effects rectified the problem, and accounted for significant additional variance in criterion validation measures, above and beyond the usual IQ scores. Additionally, individual and group-level cognitive strengths and weaknesses were recovered using deviation scores.
Conclusion was that use of deviation scores can help ameliorate flooring problems and more sensitively capture the cognitive abilities of individuals in the low range of measured intellectual ability.
Grondhuis & Mulick (2013)
Age Range: 3–12 years
Sample Size: 47 with ASD (drawn from retrospective chart review of 1,071 cases)
Topics Addressed:
Comparison of SB-5 and Leiter-R scores
Outcome:Grondhuis & Mulick (2013)
Differences in scores on the SB5 and Leiter-R were compared for a sample of children with autistic disorder and PDD-NOS. Results suggest that they are not equivalent measures of cognitive functioning when administered to individuals with ASD. As a result, depending on the measure used, misclassification of intellectual functioning could occur. Key results are summarized below.
A significant difference in scores were found between the Leiter-R and the SB5 (F(1,40) = 58.99, p < .001). The Leiter-R was, on average, 22.45 points higher than the SB5. The mean score on the Leiter-R was 87.11 while the mean on the SB5 was 64.66.
Diagnostic differences were found. Autistic disorder (Leiter-R mean IQ = 79.42; SB5 mean = 59.12); PDD- NOS (Leiter-R mean = 96.61; SB5 mean 71.52). Differences were not significant (F(1,40) = 3.16, p < . 08). The discrepancy was significantly greater (Leiter-R IQ scores higher than SB5) in preschool children compared to those in grade school (F(1,40) = 4.88, p < . 033.
Scores on the Nonverbal domain of the SB5 (mean = 70.40) were significantly lower than the Leiter-R full- scale composite (mean = 87.12); t(46) = 7.751, p <.001.
Scores on the Verbal domain of the SB5 (mean = 62.00) was significantly lower than the Leiter-R total score (mean = 87.24); t(44) = 10.020, p <.001.
Grondhuis, Lecavalier, Arnold, Handen, Scahill, McDougle, & Aman (2013)
Age Range: 4-14 years
Sample Size: 80 with ASD
Topics Addressed:
Comparison of nonverbal and verbal IQ scores in ASD as measured via different tests
Outcome:Grondhuis, Lecavalier, Arnold, Handen, Scahill, McDougle, & Aman (2013)
The SB-5 and Leiter-R do not provide comparable IQ scores for many children with ASD. Discrepancies vary, with some scoring significantly higher on the SB-5 and others scoring significantly higher on the Leiter-R. Importantly, the test differences altered the IQ category for 62.5% of children in this sample.
Key similarities and differences between this study and the Gronduis and Mulick (2013) study (see above for summary): both found that the Leiter-R and SB-5 resulted in different outcomes, but the Grondhuis and Mulick discrepancy between the two was much greater. Samples differed, as the Grondhuis and Mulick sample was a clinical sample of convenience from a single clinic, whereas the sample in the present study was larger, a research sample (i.e., rigorously confirmed ASD diagnoses), and was more geographically diverse—the current authors point out that the sample in the Grondhuis and Mulick report may have been intrinsically biased (i.e., the administering clinicians may have selected the Leiter-R because they suspected- correctly- that the children had relatively underdeveloped verbal abilities.) Consideration of both together underscores that whether a community or research sample, IQ test results can vary greatly for children with ASD depending on the test choice.
Coolican, Bryson, & Zwaigenbaum (2008).
Age Range: 3–16 years
Sample Size: 63, with ASD
Topics Addressed:
Cognitive profiles in ASD on SB-5 and whether the abbreviated IQ represents the FSIQ
Outcome:Coolican, Bryson, & Zwaigenbaum (2008).
Nonverbal and Verbal IQ scores were significantly different for individuals with ASD with nonverbal scores higher: (t(63) = 4.52, p = .000, d = .321). No significant differences were found based on age ((F(2, 57) = 2.21, p = .143, η2 =.037) or diagnostic subgroup (F(18, 540) = .975, p = .488, η2 = .031). Additional profile patterns were reported in the study.
Validity of the abbreviated battery was examined. Overall, children scored higher on the abbreviated battery than the full-scale IQ (t(62) = -3.32, p = .002, d, = -.137); however, the mean difference was not clinically significant. Some children scored higher on the FSIQ than the ABIQ. The abbreviated IQ accounted for 89.9% of the variance in full-scale IQ.
The authors concluded that caution should be used when using the abbreviated battery because it misrepresented ability in a small number of subjects in the study.