Statistical inference and spatial patterns in correlates of IQ

Abstract

Cross-national comparisons of IQ have become common since the release of a large dataset of international IQ scores. However, these studies have consistently failed to consider the potential lack of independence of these scores based on spatial proximity. To demonstrate the importance of this omission, we present a re-evaluation of several hypotheses put forward to explain variation in mean IQ among nations namely: (i) distance from central Africa, (ii) temperature, (iii) parasites, (iv) nutrition, (v) education, and (vi) GDP. We quantify the strength of spatial autocorrelation (SAC) in the predictors, response variables and the residuals of multiple regression models explaining national mean IQ. We outline a procedure for the control of SAC in such analyses and highlight the differences in the results before and after control for SAC. We find that incorporating additional terms to control for spatial interdependence increases the fit of models with no loss of parsimony. Support is provided for the finding that a national index of parasite burden and national IQ are strongly linked and temperature also features strongly in the models. However, we tentatively recommend a physiological via impacts on host-parasite interactions rather than evolutionary explanation for the effect of temperature. We present this study primarily to highlight the danger of ignoring autocorrelation in spatially extended data, and outline an appropriate approach should a spatially explicit analysis be considered necessary.

Publish (your data) or (let the data) perish! Why not publish your data too?

“…Together with IT-specialists and librarians at the University of Amsterdam, we are currently developing a repository that will eventually include much of the data from over 40 years of our freshmantesting program. Moreover, Elsevier's ScienceDirect offers the possibility to archive data as an online appendix (e.g., Hassal & Sherrat, 2011;Johnson & Bouchard, 2011). To illustrate the possibilities offered by Elsevier, we attach a data set to the current editorial that is potentially interesting for future use.…”

Does intelligence explain the association between generalized trust and economic development?

“…One important statistical issue that has recently been brought to the attention of intelligence researchers is that of spatial autocorrelation, or spatial dependence as it is also known (Gelade, 2008;Hassall & Sherratt, 2011). Interestingly, the issue was originally identified by Sir Francis Galton at a meeting of the Royal Anthropological Society in 1889 (Naroll, 1961).…”

“…In their recent paper, Hassall and Sherratt (2011) addressed the issue of spatial dependence by re-calculating p-values based on the method put forward by Clifford, Richardson, and Hemon (1989). One alternative approach is to calculate pvalues based on so-called clustered standard errors (Primo, Jacobmeier, & Milyo, 2007;Wooldridge, 2010, pp.…”

“…Countries tend to bunch together by continent when plotted along a dimension of national intelligence (Hassall & Sherratt, 2011). This suggests one grouping according to which standard errors might be clustered in a cross-national analysis, namely continents.…”

Consanguinity as a Major Predictor of Levels of Democracy

“…Furthermore, the study is based on correlation only, and so any imputation of causation must be tentative although path analysis indicates a reasonable fit to models in which assumptions about the direction of causation amongst the variables consistent with the theory are tested. A final issue is spatial auto-correlation, which results from the non-independence of data-points owing to their proximity in space (Hassall & Sherratt, 2011). It has been argued that controlling for spatial autocorrelation might actually obviate meaningful ecological relationships (Legendre, 1993), and on these grounds a case can be made for not incorporating it into ecological study designs.…”