4 6 0 3 2 29 3 8 1 17 1 8 3 13 0 18 2 12 4 16 0 15 1 13 3 7 4 10 2 28 2 14 4 11 1 13 0 22 3 7 3 7 2 26 4 24 1 14 0 20 T N T N T N T N T N T = treatment: 0 = control 1,2,3,4 = 4 soil fumigants N = number of wire worms per 9 x 9 x 5 inch plot Plots are arranged in a 5 x 5 array. Each treatment occurs in each row, and in each column. This is called a Latin Square Data from Snedecor and Cochran 1980 p 273. This is a repeat (one year later) of the experiment reported as 'Wworm1.dat' MTB > read 'Wworm2.dat' c1-c10; SUBC> nobs = 5. MTB > stack c1 c3 c5 c7 c9 into c11 MTB > stack c2 c4 c6 c8 c10 into c12 MTB > name c11 'trtmnt' c12 'count' MTB > set into c13 DATA> (1 2 3 4 5)5 DATA> end MTB > set into c14 DATA> 5(1 2 3 4 5) DATA> end MTB > name c13 'col' c14 'rows' MTB > plot 'count' 'trtmnt' - 30+ * - * count - * - * - * 20+ * - * * - * * - 3 * * - * * 10+ * - * 4 - * - * - 0+ --+---------+---------+---------+---------+---------+----trtmnt 0.00 0.80 1.60 2.40 3.20 4.00 MTB > describe 'count'; SUBC> by 'rows'. rows N MEAN MEDIAN TRMEAN STDEV SEMEAN count 1 5 12.60 8.00 12.60 10.55 4.72 2 5 13.40 13.00 13.40 3.85 1.72 3 5 14.60 13.00 14.60 8.08 3.61 4 5 13.40 13.00 13.40 5.50 2.46 5 5 18.20 20.00 18.20 7.76 3.47 MTB > describe 'count'; SUBC> by 'col'. col N MEAN MEDIAN TRMEAN STDEV SEMEAN count 1 5 10.00 8.00 10.00 4.18 1.87 2 5 13.20 13.00 13.20 8.26 3.69 3 5 18.20 18.00 18.20 8.70 3.89 4 5 13.20 12.00 13.20 5.40 2.42 5 5 17.60 17.00 17.60 7.57 3.39