Measures of Cognitive Distance and Diversity

September 2014

Abstract: I use a model of human causal learning, Causal Support (Tenenbaum & Griffiths, 2001), to derive a meaningful measure of Cognitive Distance – the degree to which two people differ in their opinions about the workings of the world. Next, I amend this measure to quantify the notion of Cognitive Diversity. Cognitive Diversity measures the degree to which opinions vary within a human collective, such as a political party or a research department of a firm or university. Measures of Cognitive Distance and Diversity are important for theoretical and empirical work which aims to link Cognitive Diversity to collective Wisdom – an organization’s success in recognizing structure in the universe of interest – in order to make robust collective decisions in uncertain environments.

Causal Inference with Spatially Disaggregated Data: Some Potentials and Limits.

(with Marion Dumas and Petr Gočev) May 2012

Abstract: In studies of civil strife, the ecological fallacy seems to befall all large-n studies and thus there has been a big push, by several researchers, in recent years to gather disaggregated, spatially explicit data. An interesting example is the recently assembled geo-reference Ethno-Power Relations dataset (Wimmer et al.), and used by Cederman et al. (2011) to argue that political exclusion and group-level economic inequality lead groups to organize to rebel against the political status quo. While we think there is much potential in this new approach to measuring the determinants of conflict, we find that the resulting data can not be analysed in conventional ways, if the estimation of causal effects is the goal. The reason is that analysis of sub-national entities may bring about other dangers: the violation of the Stable Unit Treatment Value Assumption (SUTVA). To be specific, one ``treated'' group's enemy could hardly be its control. We get around this problem by changing the causal effect of interest and by carefully re-aggregating the lower level data so as to preserve its most salient information. Restricting our analysis to groups that are excluded from power, we find some tentative evidence that such groups are less likely to engage in conflict if they are more spatially integrated with other groups.