A consumer testing service compared the effectiveness of four different brands of drain cleaner. The experiment consisted of using each product on 50 different clogged sinks and measuring the amount of time that elapsed until each drain became unclogged. The recorded times were measured in minutes.

a. Which techniques should be considered as possible procedures to apply to determine whether differences exist? What are the required conditions? How do you decide?

b. If a statistical analysis has shown that the times are not normally distributed, can the service conclude that differences exist between the speeds at which the four brands perform?