Awesome Buneary

Preregistration of analyses

Analysis #1

There is a relationship between coffee acidity and their continent/region. We will use hypothesis testing to test our hypothesis by comparing the mean values of acidity across each region. Since we are not interested in the exact population parameter, we chose not to use a confidence interval. The null hypothesis would be “The mean value of coffee acidity across each continent is the same.” The alternative hypothesis would be “The mean value of coffee acidity across each continent is different.”

Also, we can train and evaluate a machine learning model on our data. For example, we can use a decision tree or random forest to predict coffee region based on other variables such as acidity.

Analysis #2

There is a relationship between coffee aroma, sweetness, and acidity. We will use regression model and hypothesis testing to test our hypothesis. The null hypothesis would be “the relationship between coffee aroma and sweetness is positive, control for acidity” The alternative hypothesis would be “the relationship between coffee aroma and sweetness, is not positive, control for acidity” We will use bootstrapping to simulate the sample.