The manager of a large office building needs to buy a large shipment of light bulbs. After reviewing specifications and prices from a number of suppliers, the choice is narrowed to two brands whose specifications with respect to price and quality appear identical. He purchases 40 bulbs of each brand and subjects them to an accelerated life test, recording hours to burnout, as shown in Table 5.16.
(a) The manager intends to buy the bulbs with a longer mean life. Do the data provide sufficient evidence to make a choice?
(b) To save labor expense, the owners have decided that all bulbs will be replaced when 10% have burned out. Is the decision in part (a) still valid? Is an alternate test possibly more useful? (Suggest the test only; do not perform.)
