An airline wants to select a computer software package for its reservation system. Four software packages (1, 2, 3, and 4) are commercially available. An experiment is set up in which each package is used to make reservations for 5 randomly selected weeks and data on the number of passengers that are bumped over a month are collected. (A total of 20 weeks was included in the experiment.) The variability of the number of passengers that are bumped is found to be roughly the same for the 4 packages. The distribution on the number of passengers that are bumped has been found out to be right-skewed for package 1 and 4, left-skewed for package 2 and normal for package 3. Which of the following tests will be the most appropriate to find out if the average number of passengers being bumped over a month is the same across the 4 packages?
A) paired t test
B) pooled-variance t test
C) Kruskal-Wallis rank test
D) one-way ANOVA F test for differences among more than two means

Question

Quizplus · Accepted Answer

Since we are comparing the means of more than two groups, a one-way ANOVA F test for differences among more than two means is the most appropriate test. The data collected for each week are independent, and the variability is roughly the same across all four groups. Therefore, we can assume that the assumptions for ANOVA are met. The other options, such as paired t test and pooled-variance t test are not appropriate since we are dealing with four independent groups. The Kruskal-Wallis rank test is a non-parametric test and assumes that the distributions across groups are identical, which is not the case in this scenario where the distributions are different for different packages.

An Airline Wants to Select a Computer Software Package for Its