Understanding parameter differences between analyses employing nested data subsets

TL;DR

This paper offers an analytical framework to interpret parameter shifts between full datasets and subsets, helping assess data consistency and model fit by distinguishing genuine changes from noise-induced variations.

Contribution

It introduces a method to quantify expected parameter differences due to noise, aiding in evaluating data coherence and model appropriateness.

Findings

Provides a measure for interpreting parameter shifts

Helps distinguish noise from genuine data changes

Assists in assessing model fit and data consistency

Abstract

We provide an analytical argument for understanding the likely nature of parameter shifts between those coming from an analysis of a dataset and from a subset of that dataset, assuming differences are down to noise and any intrinsic variance alone. This gives us a measure against which we can interpret changes seen in parameters and make judgements about the coherency of the data and the suitability of a model in describing those data.