Quantity of draw

Estimation ML/Bayesian estimation
1

Dear Professor,

Does the quantity of draw have an impact on the model’s credibility and quality?

For example, does the article with 100,000 (one hundred thousand) draws have more credibility than the article with 3,000,000 (three million) draws?

Note that, both articles have consistency and convergence of the Metropolis-Hastings sampling algorithm.

Thank you so much.

2

No, if the chain has converged then the number of draws should not play a big role. Typically people will not use all the draws for reporting statistics, but only a number of subdraws.

3

Thank you so much, Professor.