A graduate student in computer science wants to design an experiment in which he
ID: 3202219 • Letter: A
Question
A graduate student in computer science wants to design an experiment in which he randomizes n = 18 students, labelled 1, 2, ..., 18 to three distraction treatments (see below). He gives the same difficult take-home programming quiz to all students and counts the number of errors they make. bullet T1: no distractions bullet T2: Streaming video of "Bob's Burgers" on Netflix bullet T3: No video, but popular music plays on medium volume Provide a randomized plan for this experiment. Why is randomization important? Would it be ok to just line up the students alphabetically by their last name and assign the first 6 to T1, the next 6 to T2, and the last 6 to T3? Why or why not?Explanation / Answer
(a) The plan can be to randomly divide the students into 3 groups and then each group is assigned one treatment.
(b) Ransomization is important because otherwise personal or observational bias will be included in the testing.
(c) This can not be done because while doing this, randomization does not come into play