Bayesian Approach to Electoral Processes
Iberedem Aniefiok Iwok, N. P. Akpan · International journal of statistics and applications · 2016
This work focused on Bayesian inference of Nigerian 2015 presidential polls. Prior beliefs about the proportion ����, prior means and variances of those who favour the two most popular political parties were obtained prior to conducting the election. Based on this information, beta priors ���� (���� ,���� ) distribution that satisfy these beliefs about ���� were obtained. As Bayesian tradition demands; election was later conducted and data obtained. From here, the posterior distribution that summarizes the beliefs about ���� was calculated. The posterior means were also calculated and were found to be Bayesian credible. Thus, prior belief of Nigerians was the basis of which the election was won and lost.