You work in a product design company and are creating a new alarm clock for children. You want to find the best sound for the alarm clock that wakes children up the fastest. You must design a multifactorial experiment to test different sounds on sleeping children. Your team has identified three variables of interest to test, with two levels for each variable. The variables are:
- alarm intensity (quiet vs. loud)
- pitch (low vs. high),
- sound type (musical vs. buzzer).
a) How many tests will you need to run to test all the variables, and why?

b) Write down a brief procedure you would follow to measure how fast a child wakes up after an alarm, during a single experimental test.

c) Your teammate suggests that you repeat each test on three different children. Is this a good idea, and why or why not?

d) Data is sometimes not perfect, and may contain a lot of variabilities; this is called “noise” (or “noisy data”). What are the three possible sources of noise (or variability) in your data?

e) What are three different methods you could use to control or reduce the noise in your data?