Mullen Scales of Early Learning

Age Range: 2–10 years

Sample Size: 118

Topics Addressed:

Concurrent validity between MSEL and DAS-II

Outcome:Farmer, Golden, & Thurm (2015)

Scores on the DAS-II and MSEL were highly correlated, suggesting that they measure a similar construct. However, curve estimation revealed large mean differences in scores that varied as a function of the child's cognitive ability level. DAS-II scores were consistently higher than MSEL scores. For research studies, interchanging MSEL and DAS-II scores in research studies without regard to the discrepancy in scores may produce misleading results in both cross-sectional and longitudinal studies of children with and without ASD.

Conclusion: test selection is important for practitioners, and choosing between the DAS-II or MSEL may have a greater impact than a decision between two other tests, as the DAS-II is likely to produce higher scores, so initial assessments that use the MSEL and later the DAS-II may result in the false impression in considerable improvement over time.

Age Range: Average age 3.38 years (SD=1.14)

Sample Size: 399

Topics Addressed:

Convergent and divergent validity

Outcome:Swineford, Guthrie, & Thurm (2015)

Each of the MSEL domain scores loaded onto a single, latent factor in the entire sample and the factor form held across the ASD and non-spectrum groups. This suggests that the same overall development is measured in both children with ASD and children without ASD and supports construct validity of the MSEL in both groups. Convergent validity was also supported, as indicated by most measures of language and adaptive behavior loaded onto the same factor as the MSEL scores, and suggests that these tests all index the same underlying construct. However, this model was poorly fitting in the ASD group, indicating that loadings for the ASD group should be interpreted with caution. Divergent validity was demonstrated by the loading of MSEL domain scores almost exclusively on the Developmental Functioning factor when examined in the entire sample and by group.

Conclusion: Validity of scores obtained on the MSEL across this developmental span was supported.

Age Range: 2–4 years

Sample Size: 55

Topics Addressed:

Concurrent validity between multiple measures of expressive and receptive language, including the MSEL

Outcome:Nordahl-Hansen, Kaale, & Ulvund (2014)

The results suggested very high agreement between the measures, and this was found across raters (i.e., whether parents or teachers filled out a Communicative Development Inventory). This study showed that there are several valid measures available for measuring expressive and receptive language.

Conclusion: both direct testing and parental reporting can be used for assessing language properties in children with ASD. Teachers, in addition to parents, can provide valid ratings of children’s language abilities.

Age Range: 2–4 years

Sample Size: 94

Topics Addressed:

Criterion validity and sensitivity

Outcome:Burns, King, & Spencer (2013)

Each clinical group autism spectrum disorders (ASD; n = 19), cerebral palsy (CP; n = 14), and epilepsy (EPI; n = 14). demonstrated statistically significant delays across domains relative to the respective matched control group (p < .001). Children failed to demonstrate a “signature” profile for a diagnosis of ASD, CP, or EPI. The clinical sensitivity of the MSEL and the need for obtaining specific intervention services for children diagnosed with these conditions was presented. Finally, these results are discussed within the context of the clinical sensitivity of the MSEL in working with these clinical populations. (Adapted from abstract).

Age Range: 2:7-5:7 years

Sample Size: 72

Topics Addressed:

Convergent validity of the MSEL and the DAS

Outcome:Bishop, Guthrie, Coffing, & Lord (2011)

Results showed good convergent validity with respect to nonverbal IQ (NVIQ), verbal IQ (VIQ), and NVIQ-VIQ profiles. NVIQ scores on the MSEL and DAS were statistically similar (r = .74, p <.01), as were VIQ scores (r = .83, p < .01). For both VIQ and NVIQ, less than one third of children received scores on the two tests that were in different cognitive ability ranges. There were, however, a handful with large discrepancies in terms of NVIQ-VIQ profiles. This may suggest the need for researchers to obtain more than one measure of IQ for each participant; it cannot be assumed that different cognitive tests yield comparable scores.
Conclusion: use of MSEL age-equivalents to generate NVIQ and VIQ scores is supported, though establishing convergent validity of cognitive tests is needed before IQs derived from different tests can be conceptualized as a uniform construct.

Age Range: 16-43 months

Sample Size: 42

Topics Addressed:

Assess differences in testing behaviors and how these were related to MSEL scores

Outcome:Akshoomoff (2006)

Children with ASD required less time to complete the MSEL compared to typically developing peers but spent proportionally more time exhibiting off-task behaviors and less time engaged with the instrument. Scores obtained on the MSEL were positively correlated with level of engagement and negatively correlated with off-task behavior.

Conclusion: the MSEL provides a format conducive for evaluating motor and visuomotor skills, language comprehension, and communication skills in young children with ASD. Testing conditions and behaviors are likely to affect obtained scores for young children with ASD, and re-evaluating cognitive skills to measure response to treatment and getting a better sense of the child’s level of functioning is advised.