Covariance matrix estimation is important in statistics, especially when dealing with multiple variables. Different prior distributions were compared in a Bayesian analysis to see which one works best. The inverse Wishart prior can lead to biased results when the true variance is small compared to the prior mean, causing the posterior variance to be larger and the correlation to be closer to zero. Other priors like the scaled inverse Wishart, hierarchical inverse Wishart, and separation strategy perform better in these cases.