The triangle test is a popular method in sensory evaluation. It is used to determine whether there is perceivable difference between two samples. Based on the total number of triangle test participants, and the amount of those participants which were correct, we are able to assign a level of statistical significance to the result.

In a previous post I provided a triangle test calculator based in Google sheets. This calculator provides an approximate and exact p-value based on the correct proportion of tasters in a sample.

I’ve produced a new calculator in “shiny”, a web framework for small web applications built in R. This calculator also calculates the exact p-value, but additionally it provides a graphical representation of the p-value.

Here, the p-value is the probability of having observed an equal or greater number of correct tasters with the triangle test if the null hypothesis were true (that the population proportion is equal to 1/3). A small p-value indicates evidence against the null hypothesis – that is, that there is actually is a perceivable difference between samples. You can play around with the sample size and the number of correct tasters to visualize changes in this p-value with the new calculator.

Here are the links to the github repository where the code is located.

**Edit**: this calculator has proven so popular that the free hosting I’ve used for it quickly runs through it’s bandwidth each month. I suggest running the code locally using R and shiny, or deploying your own instance of it on shinyapps.io.

## Comments