From Complex Network Perspective: Is Anecdotal Inference in Social Network Effective?

Anecdotes play a significant role in human communication and understanding. Inference on individual experiences and isolated cases is defined as anecdotal inference here. This study employs complex network methods to model the anecdotal inference to four small-scale sampling algorithm. The limitation is verified through theoretical derivation and numerical simulation. Theoretical result is marked by bias between the expected value, positively correlated with high-degree nodes’ attributes, and the latent parameters. Experimental result is shown by comparing results error of the algorithms in scale-free, small-world, random, and CNN network models, and the results validate the theoretical conclusion. Additionally, different network structural properties show varying influence on the error of estimated value.