In an experiment conducted to see if electrical pricing policies can affect consumer behavior, 10 homeowners in Wisconsin had to pay a premium for power use during the peak hours. They were offered lower off-peak rates. For each home, the July on-peak usage (kilowatt hours) under the pricing experiment was compared to the previous July usage.
Year
Previous Experimental
200.......................... 160
180.......................... 175
240.......................... 210
425.......................... 370
120.......................... 110
333.......................... 298
418.......................... 368
380.......................... 250
340.......................... 305
516.......................... 477
(a) Find a 95% confidence interval for the mean decrease.
(b) Test H0: μD = 0 against H1: μD ≠ 0 at level α = .05.
(c) Comment on the feasibility of randomization of treatments.
(d) Without randomization, in what way could the results in parts (a) and (b) be misleading?
[What if air conditioner use is a prime factor, and the July with experimental pricing was cooler than the previous July?)