solution

In a study of water quality trends, EPA compiled stream biological monitoring data in the Mid-Atlantic region before and after 1997. They are interested in whether there was a shift in biological conditions in streams in the area. The indicator they used is EPT taxa richness (number of taxa belong to three genera of flies commonly known as mayfly, stonefly, and caddisfly). As distributions of count variables are typically skewed, log transformation was used. The log mean and standard deviation of EPT taxa richness before 1997 are 2.2 and 6.9 (n = 355), and are 1.8 and 5.4 after 1997 (n = 280). Is the difference in log mean statistically significant? How do you report the difference in terms of EPT taxa richness?