Social network robustness testing via random data removal (Gasterosteidae)

Knowledge graph centered on Social network robustness testing via random data removal (Gasterosteidae) with 20 nodes and 70 connections. Top connected: Social network analysis of animal behaviour: a pro, Marmota flaviventris, Atriplex canescens, prairie dogs, Social network analysis (Sciuridae).

Method synopsis

A computational approach to test the stability of social network parameters by systematically removing increasing percentages of observational data and recalculating network measures to determine sensitivity to sampling effort.

Synthesized from method descriptions across 1 paper using this protocol.

Procedure as described in the canonical source

Steps below were extracted from the paper that introduces this protocol — Social network analysis of animal behaviour: a promising tool for the study of sociality (2008), Animal Behaviour. Implementations in other papers (listed below) may differ.

1. Test parameter robustness
   Randomly remove 10%, 20%, 30%, up to 90% of observations and recalculate network measures to test stability of parameter estimates under reduced sampling
   Quantities: 10 removal levels from 10% to 90%, replicated for each subsetDuration: Not specifiedConditions: Computational simulation
   Equipment: statistical software
2. Statistical analysis of robustness
   Use analysis of variance (ANOVA) to determine what percentage of data removal resulted in significant changes to network measures
   Quantities: ANOVA tests for each network parameterDuration: Not specifiedConditions: Statistical analysis
   Equipment: ANOVA software

Typical Equipment

• statistical software

Output Measurements

• network parameter stability
• sampling robustness estimates

Papers Using This Protocol (1)

Species Studied With This Protocol