# Alter

Disputed - Joshua D. Angrist - not applicable

## Introduction

DONE Is it possible from the data of cross-cultural research to piece together the processes of human evolution, the alternative developments of social organization and structure, modes of behavior, sexuality, systems of belief, morality, and religion.

DONE Cite work on fractals--Marcus J. Hamilton and use DEf to model with Geo Regression

2009 Pearl p90: "In general, the addition of arcs to a causal diagram can impede, but never assist, the identification of causal effects in nonparametric models. This is because such addition (p91:) reduces the set of *d*-separation conditions carried by the diagram."

Angrist Joshua Jorn-Steffen Pischke. 2009. Mostly Harmless Econometrics - An Empiricist's Companion. Princeton. Princeton University Press.

## Geo Regression and Bayesian Learning

We may generalize from DEf regression a more general conception that has four basic empirical elements in the observations of cases and outcomes of analysis

- exist in space (e.g., a set of coordinates)
- exist in history (e.g., devolving from ancestral predecessors, language groups, etc.)
- may produce probabilistic Bayesian-like predictors - seen by the R library(bnlearn)
- may have alternative probabilistic bnlearn models when examined by library(bootstrap)

The implications of these features are not restricted to cross-cultural studies or samples but may apply to other samples of observed cases. Imputation of missing data will apply if researchers supply a sufficiently range of completely-coded variables for principal component analysis that may be used to predict values of missing values.

New observational samples may be produced for Geo Regression Analysis (GRA) and added to the Co-SSci databases if they conform to these general features and are of general interest and suitable for DEf analysis.