Question

A pizza delivery driver, always trying to increase tips, runs an experiment on his next 90 deliveries. He flips a coin to decide whether or not to call a customer from his mobile phone when he is five minutes away, hoping this slight bump in customer service will lead to a slight bump in tips. After 90 deliveries, he will compare the average tip percentage between the customers he called and those he did not.

g) Is the experiment blind? Can it be double-blind? Explain.

h) Name some confounding variables that might influence the experiment's results.

Answer 1

g) Since both the pizza delivery boy (experimenter) and the customers (participants) do not know which type of treatment is being applied to which participant in advance, this experiment is a double blind experiment. This experiment ensures there is no observer bias in the setup.

h) Some confounding variables that might influence the experiment's results can be:

1) Occasion: Customers might mostly extend tips during festivals so the tip collected could be abnormally high given the occasion.

2) Delivery area: If the pizza delivery boy delivers in a very posh area, he might end up collecting more tips than his co-worker in another delivery area.

3) Time of the day: Deliveries done at midnight can be expected to fetch higher tips given the customers might feel more obliged getting their favorite meal delivered at a time of craving.