The Perfect Doughnut Study You have been employed as a statistician for a multinational confectionery company to help them develop the perfect doughnut. You conduct some market research and discover that there are a variety of factors that make for a good doughnut, including; the diameter (in millimetres); the sugar content of the dough (in grams); and whether or not it is iced, jam-filled, or has sprinkles/sugar dusted on top. Given the large budget, you are able to commission a variety of tester donuts created to your specifications (N=213), each of which you measure diligently, check the recipe (for sugar content) and categorise by topping/filling. Rather than have a single taste-tester eat every doughnut, making for confounders in time of testing, you instead dip into your pool of product testers. Each of the 213 participants tastes a different doughnut, and rates how much they liked it on a scale of 1 to 10. You collect demographic information from each tester to ensure that you can control for aspects that might affect perception of taste, such as age and gender. Question: When you run an analysis to try and identify the optimal (most 'liked') type of doughnut, you are presented with