Links Between Imitation and Mental Development A Cross-Lagged Twin Study

Links Between Imitation and Mental Development: A Cross-Lagged Twin Study Introduction Figure 1. Birdhouse imitation task Figure 2. Puppet imitation task • Despite extensive research on imitation in childhood, little is known about the stability and change in imitation across age. • Single headed arrows represent standardized path estimates for the effects of latent factors on phenotypes, cross-age stability, and cross-lagged effects. • Moreover, the longitudinal relation between imitation and mental development has not been examined in typically-developing children. • Research demonstrating the importance of working memory and language in relation to childhood imitation (Riggins et al., 2013, Sloate & Voyat, 1983) suggests its potential relevance. • Double headed arrows linking two latent factors indicate genetic, shared environmental, or nonshared environmental correlations between the two constructs within an age. Figures 3. Rattle imitation task Goals of current study were to explore the: • Concurrent and longitudinal relations between imitation (IM) and mental development (MD) at ages 2 and 3. • Underlying genetic and environmental contributions to these associations as well as to stability and change in imitation and MD across age. Results Phenotypic Correlations Sample • At age 3, there were novel genetic and nonshared environmental effects on IM and MD. Half of the new genetic factors for IM and MD overlapped. • 314 twin pairs (MZ=145; DZ=169) participating in the Boston University Twin Project at ages 2 and 3 years. • Zygosity was determined via DNA analyses using DNA obtained from cheek swab samples. Mental Development • As seen in Table 1, IM was associated with MD at both ages. • MD was assessed at each age with the Mental Development Index scores derived from the Bayley Scales of Infant Development-II. • The age-to-age correlation for imitation was modest, whereas MD was highly stable. • Assessed via performance on three multi-step, observer-rated imitation tasks derived from prior research (see Figures 1 - 3). • Scoring allotted one point for each step correctly replicated, and one point for correct sequence. • A composite elicited imitation score was then derived by summing the total number of points received. • There were significant genetic, shared and nonshared environmental influences on IM and MD at both ages. • At age 2, the association between IM and MD was due to overlapping genetic, shared and nonshared environmental influences. Methods Elicited Imitation Figure 4. Biometric Cross-lagged Model (shown for one twin). Dashed paths indicate parameters that could be dropped from the best-fitting model for parsimony. Solid paths indicate parameters retained in the best-fitting reduced model. Model-Fitting Analyses • A cross-lagged biometric model was fit to the data (Figure 4) to examine the relation between IM and MD from 2 to 3 years. • Latent factors appear in circles (i.e., A for genetic factors, C for shared environments, and E for nonshared environments), whereas observed (measured) variables appear in rectangles. • The latent factors for age 3 represent residual variances specific to that age. • Table 2 decomposes the genetic and environmental influences on IM and MD at age 3 into effects that are transmitted from age 2 and effects that are new at age 3. • Nearly all of the genetic and nonshared environmental influences on IM at age 3 were agespecific (98.7% and 99.5%, respectively). • The only effects on IM that were transmitted from age 2 to age 3 are those that come from MD (i.e. cross-lagged effects). • For MD, only stability effects were transmitted across age. These stability effects accounted for 30% of the genetic variance, 21% of the nonshared environmental variance, and 100% of the shared environmental variance at age 3. Conclusions • Although IM is genetically influenced at both ages, the genetic effects that influence IM at age 3 are largely independent from those that operate at age 2. • MD shows more stability that IM, yet most of the genetic and nonshared environmental variance on MD at age 3 is new and reflects developmental change. Shared environmental variance on MD at age 3 is, however, entirely transmitted from age 2. • MD at age 2 predicts IM at age 3, but not vice versa. • Overall, despite the fact that IM and MD are correlated at both ages and have overlapping genetic effects, the impact of MD on imitation is modest.