fetchFiniteSD: Finite-population standard deviations.

In StatisticsNZ/demest: Bayesian Demographic Estimation and Forecasting

Finite-population standard deviations.

Description

Extract the finite-population standard deviation of main effects and interactions in hierarchical models.

Usage

fetchFiniteSD(filename, probs = c(0.025, 0.5, 0.975), iterations = NULL)

Arguments

filename	The name of the file where the output from the estimate function is kept.
probs	A vectors of quantiles, for summarizing the standard deviations.
iterations	A vector of positive integers giving the iterations to be extracted.

Details

In the hierarchical models discussed in Model, cell means are moderned using terms formed from cross-classifying dimensions such as age or sex, or from interactions between these dimensions. Adapting ideas from ANOVA, Gelman (2004: 408-409) propose that the finite-population standard deviations of these terms be used as a measure of the terms' relative importance. For instance, is an 'age' term has a greater finite-population standard deviation than a 'region' term in a model, this suggests that age is a more important predictor of outcomes than region.

Value

A list of objects of class "FiniteSD" if object contains several hierarchical models; a single "FiniteSD" object if' object contains a single hierchical model; and NULL if object contains no hierarchical models.

References

Gelman, A. (2004) Analysis of variance: why it is more important than ever (with discussion). Annals of Statistics. 33(1): 1-53.

See Also

estimateModel, estimateCounts, estimateAccount

StatisticsNZ/demest documentation built on Nov. 2, 2023, 7:56 p.m.