A proposed null hypothesis states that there is no difference in the water consumption of households in two districts. The sampl
e mean difference is 8.5 liters, and the standard deviation of the distribution of the difference in sample means is 4.5 liters. The null hypothesis must be accepted if we choose the confidence interval.
In order to be able to consider this a valid null hypothesis is to accept the fact there is no significant difference between the two subject population, this statement is true for inferential statistics.. In this case the water consumption in the two district has a confidence interval should be chosen with respect to what is the closest to 100%.
