A Bayesian statistics approach to hospitality research

ABSTRACT

Bayesian statistics approach contraposes inferential statistics by the fact that it introduces experts’ opinion in the quantitative analysis. While this approach has played an increasingly important role in various fields of research, its application to hospitality research has been limited. Bayesian statistics helps resolve the issue of the shortage of observations, which is a frequent problem in certain areas of the hospitality industry. Secondly, the Bayesian approach is particularly well suited when the variables used are already subjective or abstract. Therefore, this study aims to explain how a Bayesian statistics approach contributes to the advancement of hospitality management and demonstrates how this approach can be applied to analyse guests’ online reviews for a hotel.

KEYWORDS:

• Bayesian statistics
• inferential statistics
• hospitality research

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The article is limiting its analysis to the analytical derivation of a posterior distribution. There are cases, however, on which Equation (2) cannot be computed analytically (Johnson et al., 2007). Different methods have been developed to deal with this problematic, including what has become the most popular method, i.e. the Markov chain Monte Carlo (MCMC), which uses ‘a Markov chain to sample from the posterior distribution' (Johnson et al., 2007). Those methods aim to solve more complex statistical models (Carlin & Chib, 1995), which are beyond the scope of this paper.

2 The mean of the Poisson distribution ($\lambda$) is also equal to the variance of the Poisson distribution.

3 For a more detailed solution see Appendix 2 in Donovan and Mickey (2019).