BACKGROUND: Our primary measure of interest is relational complexity, a measure that encapsulates both an individuals's processing capacity and task complexity and is proposed to account for capacity limits found in both reasoning and working memory. At a variance components level, we show that a genetic factor influencing relational complexity accounts for most of the genetic covariation between measures of reasoning and working memory, and further, that there is a strong overlap with IQ. A criterion of our genetic variants of interest was that they must influence variation in all measures (i.e. relational complexity, IQ, reasoning, and working memory), while our genes of interest were the top results for relational complexity and for IQ.

AIM: To assess our most promising genetic variants and genes in independent samples. Two independent groups (the LBC1936 Scottish sample and the NCNG Norwegian sample) have shown some consistency with our findings, but results from a third group (the NTR Netherlands sample) were inconsistent.

HYPOTHESIS: Replicating genetic association and gene-based analyses in multiple independent samples increases the likelihood of avoiding Type 1 (and Type II) error and thereby increases the value of our manuscript.

INDEPENDENT VARIABLES: genotypes for 99 single nucleotide polymorphisms (SNPs) - 11 SNPs are specific genetic variants of interest and the remaining 88 SNPs represent the two genes of interest (list of 99 rs numbers to be provided if application is successful). Note that our genotyping was done using the Illumina HumanHap 610 quad chip (versus HumanHap 550 used by ALSPAC) and it is possible we may need to supply proxies for some SNPs.

DEPENDENT VARIABLES: Matrix Reasoning, WASI IQ (measures available for the other samples: LBC1936 and NCNG both had measures of Matrix Reasoning and IQ, while NTR had a measure of Raven's Progressive Matrices)

(possible) CONFOUNDING VARIABLES: Sex, Age, and Multidimensional Scaling (MDS) components as appropriate to the ALSPAC sample to adjust for population stratification

ANALYSES: Genetic association for 99 SNPs for the measures Matrix Reasoning and WASI IQ, with sex, age, and MDS components included as covariates. (NOTE: for a subset of these SNPs (i.e. 88 of 99) the association results will be used to run a gene-based analysis for 2 genes - this analysis to be conducted at QIMR)