As part of a study of southern corn leaf blight infection 30

As part of a study of southern corn leaf blight infection, 30 plants were randomly assigned to one of 6 light levels,from0.4to2.0unitsoflight(samenumberofplantsperlightlevel). Thettedregressionequationwas %leaf area infected = 11.3 - 5.25*light. Diagnostics, including residual plots and a lack-of-t test, indicate no problems with linearity or other assumptions. Other potentially useful information includes: Sample linear correlation coefcient = -0.7 Sample Variance of %leaf area infected=20.25 Sample Mean of all light values = 1.2 Sample Variance of all light values = 0.36
(a) Constructa99%condenceintervalfortheslopeandinterpretthisintervalinthecontextoftheproblem.

(b) Theresearcherneedstoknowifplantsgrowingin1.8unitsoflighthave,onaverage,morethan1.6%of the leaf area infected. Give your answer using an appropriate statistical inference procedure.

(c) The researcher needs to estimate the mean leaf damage with a standard error less than 2%. Is that possible with this model and data? Explain why or why not.

Solution