This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Stat 402A, HW 7 Answers 1) Plankton tows My SAS for parts a-c: data plankton; infile ’plankton.txt’; input tow spp $ abdn; logabdn = log(abdn); proc glm; class tow spp; model abdn = tow spp; estimate ’diff spp 1 - 2’ spp 1 -1 0 0; output out=resids r=resid p=yhat; proc plot; plot resid*yhat; proc sort data=plankton; by tow; proc means noprint data=plankton; by tow; var abdn; output out=means mean=meanabdn; data new; merge plankton means; by tow; proc plot; plot abdn*meanabdn=tow; run; (a) 2 pts. The appropriate model includes spp and tow, both as class variables (i.e. defining groups). The desired test is is the F test for spp. F = 251.7, p < 0.0001. (b) 2 pts. The estimate is -10.4 with a s.e. of 12.3 and a p-value of 0.40. (c) 2 pts. The plots are not enclosed. The residual plot shows strong curvature and unequal variance. The obs vs block means plot clearly shows fanning lines. Neither assumption seems reasonable. (d) 2 pts. The SAS code for parts d and e is identical to the above code, but logabdn is used as the response instead of abdn. You could choose either mean abundance or mean log abundance as the X axis in the plot. That’s really a minor issue that doesn’t affect theabundance as the X axis in the plot....
View Full Document
This note was uploaded on 01/27/2010 for the course STAT 402 taught by Professor Staff during the Spring '08 term at Iowa State.
- Spring '08