restaurant would like to increase its sale of dessert items. They decide to test if bringing out actual dessert items to show customers would entice themto order something. Over a period of one month, the manager on duty tosses a coin whenever a customer finishes their main course. If the coin comesup heads, the server brings them a dessert menu; if it comes up tails, the server brings them a tray showing the actual dessert items. The manager thenrecords whether the customer ordered dessert or not. (a) Suppose the manager finds that customers who were shown the actual dessert items had a significantly higher chance of ordering dessert. Is it validto claim that showing the dessert items caused the increase in dessert orders? 1. Yes or no ? because

2.

a. the dessert items were randomly selected. b. what to show the customer was randomly assigned. c. diff erences in income and other lurking variables are likely the underlying cause.

3.

The manager would also like to determine if customers were satisfied with their dining experience. After their meal, some customers were randomlychosen to fill out a short customer satisfaction survey. Why were the customers randomly chosen as opposed to the server identifying the best customersto participate? Randomization reduces non-response bias. Randomization enables the manager to determine the cause of the diff erence between satisfi ed vs. dissatisfi ed customers. Randomization enables the survey results to be generalized to other customers who didn't fi ll out the survey.