All parameters on Bayesian are considered as random variables. Prior distribution for parameter has to determined. Prior distribution is proposed by researcher as the basic of available information  before data is used. Bayesian is used to find the posterior distribution of the parameter given by the observed data.  However, the problem regarding calculation posterior distribution will arise if the different prior distribution is used. The existing prior information needs to be considered on the structure  of the  prior distribution, while the non-existing prior information needs to attract any prior parameter choice which may impact on the posterior  distribution.  The Bayesian hierarchical Model is defined when a prior distribution is also assigned on the prior parameter associated with likelihood parameters. The prior distribution on model is characterized by two levels of hierarchy: first and second level. The hyperparameter influence on hyperprior (second level) and hyperprior influence the prior distribution (first level ), while prior  distribution is  influence likelihood data. This article discusses more detail about the hyperparameter application on the Polynomial  Regression Model  via  MCMC  and  INLA.