• Purdue University

• IE 533

Lady Tasting Coffee: A Case Study in Experimental Design
By Prof.Hong Wan
Part 1:
1. Hypothesis: If the milk was added to the cup before the coffee, then the model could
tell the milk was added first and she would drink the coffee.
2. In this experiment, it is important to offer the model more than two cups because many
different trials can help support the hypothesis and experiment. If there are only two cups
to offer the model (one with milk before coffee and one with coffee before milk), the
hypothesis couldn’t be proven with little information and data. Also, more cups of coffee
will help distinguish if the model is only guessing on flavor with milk first rather than
knowing the flavor for sure.
3. The model should taste 10 cups of coffee (5 milk before coffee, 5 coffee before milk)
so the data and results from the experiment can be accurate and fully support the
hypothesis. Also, the more cups of coffee in the experiment will establish if the model is
truthful or not when deciding which coffee is which.