A Bayesian Correlational Study on Three Coffee Bean Type and Brewing

Abstract

Barako Coffee is a coffee variety unique to the Philippines, and the research gaps showed an absence in the study of Barako Coffee and immersion brew time. Other bean types (Arabica and Robusta) were included due to the rarity of Barako Beans, as they are only sourced from a specific farm in Batangas and Cavite. This study hopes to contribute to the literature using a single experiment Bayesian correlation between brew time and perceived bitterness. The single experiment is done over three months, yielding 100 cups of coffee using three types of coffee beans: Arabica, Robusta, and Barako. For each brew, the variables of bean type, perceived bitterness, dilution of coffee, and brew time duration in minutes were correlated. Correlation analysis was done using Bayesian Analysis (BF10) in favor of the alternative hypothesis, which states a significant correlation exists between these Variables. Statistical Correlation pointed out that only two variables have a significant correlation. The first set of variables is Bean Type and Bitterness, which has a strong positive correlation, with a Pearson rho of 0.670 and BF10 Value of 3.366×10^+1. The second variable is minutes of brew time and dilution, with a Pearson rho of 0.355 and a BF10 of 80.519; a strong evidence of correlation. These two variables were significant at the alpha level of 0.05. The findings validated the commercial application of longer brew time to yield a more robust cup of coffee; however, they also raised the need for further investigation and study on the variables of Brew Time and Dilution. AeroPress Coffee is known for the dilution of coffee brew. Perhaps dilution can also be used to adjust the taste of coffee to personal preference. This research article is a product of the research capability training conducted by the University of the Visayas during the summer of 2024, which focuses on applying Bayesian correlation using Jeffrey’s Amazing Statistical Program