In an article in Public Roads (1983), Bissell, Mason and Woods study the bridge safety (measured in accident rates per 100 million vehicles) and the difference between the width of the bridge and the width of the roadway approach (road plus shoulder). They find that the data collected does not satisfy the regression assumption hence they transform the dependent variable using quartic root. y* = y^.25

The prediction interval for bridge safety when x=-6 is given as [3.2,4] for the transformed regression. What is the prediction interval for y? (Round to 4 decimal points)

[, ]