An automobile association wants to demonstrate that maintaining tire pressure results in better fuel economy. The association selected a random sample of its members, and asked them to record their fuel consumption in litres per 100 km for all of their driving for one month. It then asked the members to check and adjust tire pressures every 3 to 4 days, for another month, and record their fuel consumption again. The average difference in number of litres per 100 km was 0.4, with the standard deviation of the differences being 1.4. The order of subtraction was [(fuel consumption without checking tires) - (fuel consumption with checking tires)]. The sample size was 20.The histogram of differences appears to be normally distributed. Do these data support the association's claim that maintaining tire pressure improves fuel economy? Use a 4% significance level. What else might explain the results in this case?