Several weeks ago an article appeared in the press comparing the Fit Bit exercise monitor to the Garmin monitor. In fact, earlie
r reports suggested that the Fit Bit monitor was inaccurate in a number of its readings.You are a statistician who has been asked to compare the Fit Bit tracker to the Garmin tracker to see if there are significant differences between them. The variable you are interested in studying is the number of steps an individual takes during the day when wearing one of these trackers. You are being asked to construct a study for this question. Please write out your answers to the following questions. You have not undertaken the survey as of yet, so all you have to provide is a plan.How large a sample would you consider taking for each tracker? In other words you have to decide n1 and n2.What would your null and alternative hypotheses be?Are you conducting a dependent or independent test?What reading would you record for each member of your sample?Would you have any restriction on the individuals included in your sample? Would you restrict them by age, weight, athleticism? Why?Do you think you would know the standard deviations for each population (tracker)? What should be done about this?Now suppose the question changes and you are asked by a medical society to measure the effectiveness of the Fit Bit tracker. Your question would be does use of the Fit Bit tracker lead its users to lose weight?What test would you suggest conducting here? Would you use two independent samples or a dependent sample? Why?What data would you collect?What would your null and alternative hypotheses be?
