More posts with this tag

We investigated (1) whether 10-year-old internationally adopted Chinese girls who, on average, showed below-average intellectual functioning two and six months after adoption (Times 1–2, N = 92), showed catch-up in intellectual functioning, school achievement, and executive functioning nine years later (Time 3, N = 87). We tested (2) effects of type of pre-adoption care (institutional versus foster care), and effects of parental sensitivity (factor score for supportive presence and for intrusiveness compiling all three time points), and we examined (3) whether the latter buffered the effects of type of care. The children showed significant recovery and complete catch-up in intellectual functioning at Time 3, and did not lag behind at school compared to non-adopted children. Contrary to the results of Times 1 and 2, type of care was not associated with cognitive development at Time 3. Parental sensitivity did not buffer the link between type of care and cognitive development.

Authors’ own highlights:

• The Chinese adopted girls showed above-average intellectual functioning at age 10.
• Nine years after adoption, the Chinese adopted girls did not lag behind at school.
• Type of pre-adoption care was not associated with the cognitive outcomes at age 10.
• Parental sensitivity did not buffer the link between pre-adoption care and cognition.

So the results of this one closely mirror the findings of the larger Swedish study Odenstad et al 2008.

Was there a catch-up effect? Not necessarily. What tested was used here? Turns out it’s psychomotor ability, not intelligence!:

In infancy (Times 1 and 2) the intellectual abilities of the children were assessed during the home visit with the Dutch version of the second edition of the Bayley Scales of Infant Development (BSID-II-NL; Van der Meulen, Ruiter, Lutje Spelberg, & Smrkovsky, 2004). The BSIDII-NL is a standardized assessment instrument used to measure the mental and psychomotor development of children aged between 1 and 41 months

The time 3 measures were more standard:

At Time 3 two broad cognitive abilities were assessed, namely crystallized intelligence and fluid intelligence. These two cognitive abilities have the highest loadings on general intelligence according to the Cattell-Horn-Carroll (CHC) theory of cognitive abilities (Alfonso, Flanagan, & Radwan, 2005). To assess crystallized intelligence (i.e., culturally acquired knowledge; Cattell, 1971) the information (general knowledge questions) and vocabulary subtest (defining words) of the Dutch Wechsler Intelligence Scale for Children-Third Edition (WISC-IIINL; Kort et al., 2005; Wechsler, 1991) were administered. Raw scores on both subtests were converted to age-corrected standardized scores which were transformed into an IQ score (Sattler, 1992). To get an indication of children’s general reasoning ability (an aspect of fluid intelligence; Cattell, 1971), the analogies (reasoning by analogy) and categories (reasoning by categorization) subtests of the Snijders-Oomen Non-Verbal Intelligence Test (SON-R 6-40; Tellegen & Laros, 2011) were administered. Raw scores on both subtests were transformed to age-corrected standardized scores which were used to estimate total cognitive performance on the SON-R 6-40.

The test norms seem to be fairly recent, so we don’t have to worry too much about Flynn effects causing spurious superiority. This is the most important point made by Thomas (2017)’s review of these kind of studies.

Correlation matrix:

They report a lot of additional interaction models, but these have very limited meaning since the sample size is too small for that.