Pat I used thresholds of 2.32634787404074 2.43237905858500 2.57582930355009 for .01, .0075 and .005 respectively. .01 A 1 1 1 95.0 0.8211 0.7367 0.9076 6 1 6 1 C 1 1 1 95.0 0.0954 0.0129 0.1751 0 0 6 1 E 1 1 1 95.0 0.0835 0.0693 0.0988 6 1 6 1 M 1 2 1 95.0 0.9165 0.9001 0.9308 0 1 0 1 D 1 2 1 95.0 0.5060 0.4654 0.5450 0 0 0 0 .0075 A 1 1 1 95.0 0.7852 0.7042 0.8595 0 1 6 1 C 1 1 1 95.0 0.1340 0.0547 0.2104 6 1 6 1 E 1 1 1 95.0 0.0809 0.0671 0.0923 6 1 6 1 M 1 2 1 95.0 0.9191 0.9034 0.9329 0 0 0 1 D 1 2 1 95.0 0.5266 0.4876 0.5640 0 0 0 0 .005 A 1 1 1 95.0 0.7375 0.6610 0.7790 6 1 6 2 C 1 1 1 95.0 0.1859 0.1179 0.2582 6 1 6 1 E 1 1 1 95.0 0.0767 0.0649 0.0913 6 1 6 1 M 1 2 1 95.0 0.9233 0.9085 0.9364 0 0 0 0 D 1 2 1 95.0 0.5546 0.5177 0.5900 0 0 0 0 So the rarer it is, the more C there is. Quite interesting. Heterogeneity Chi-squareds: Prev ACE(df=22) R(df=22) .01 135.87 141.28 .0075 149.29 153.45 .005 163.43 166.19 psycho> gt ascszbest.mxo | awk '{printf "%f\n", sum+=$7, " "}' 359.003000 psycho> gt ascszrbest.mxo | awk '{printf "%f\n", sum+=$7, " "}' 359.003000 psycho> gt ascszrw.mxo | awk '{printf "%f\n", sum+=$7, " "}' ascszrworst.mx ascszrworst.mxo psycho> gt ascszrworst.mxo | awk '{printf "%f\n", sum+=$7, " "}' 10041.246000 psycho> gt ascszworst.mxo | awk '{printf "%f\n", sum+=$7, " "}' 10041.246000 psycho> gt ascsz.mxo | awk '{printf "%f\n", sum+=$7, " "}' ascsz.05.mx ascszcombb2.mx ascszrcomb.075.mx ascsz.075.mx ascszcombb2.mxo ascszrcomb.1.mx ascsz10.mx ascszcomb.mx ascszrcomb.mx ascsz10.mxo ascszcomb.mxo ascszrcomb.mxo ascsz12.mx ascszcombno357.mx ascszrcombno357.mx ascsz12.mxo ascszcombno357.mxo ascszrcombno611.mx ascsz.1.mx ascszcombold.mx ascszrcombold.mx ascsz2.mx ascszcombold.mxo ascszrcombold.mxo ascsz4.mx ascszcombyoung.mx ascszrcombyoung.mx ascsz5.mx ascszcombyoung.mxo ascszrcombyoung.mxo ascsz8911.mx ascsz.mx ascszr.mx ascszbest.mx ascsz.mxo ascszr.mxo ascszbest.mxo ascszno357.mx ascszrno357.mx ascszcomb.05.mx ascszr.05.mx ascszrno611.mx ascszcomb.075.mx ascszr.075.mx ascszrworst.mx ascszcomb.1.mx ascszr.1.mx ascszrworst.mxo ascszcomb2.mx ascszr2.mx ascsztest.mx ascszcomb2.mxo ascszrbest.mx ascszworst.mx ascszcomb8911.mx ascszrbest.mxo ascszworst.mxo ascszcomb8911.mxo ascszrcomb.05.mx psycho> gt ascszcombold.mxo | awk '{printf "%f\n", sum+=$7, " "}' 7326.095000 psycho> gt ascszcombyoung.mxo | awk '{printf "%f\n", sum+=$7, " "}' 3167.514000 6343.716000 9512.160000 psycho> gt ascszcombryoung.mxo | awk '{printf "%f\n", sum+=$7, " "}' grep: ascszcombryoung.mxo: No such file or directory psycho> gt ascszrcombyoung.mxo | awk '{printf "%f\n", sum+=$7, " "}' 3165.964000 psycho> gt ascszrcombold.mxo | awk '{printf "%f\n", sum+=$7, " "}' 7326.095000 Heterogeneity Chi-squareds (all df=22): Prev ACE R various 124.87 130.85 .01 135.87 141.28 .0075 149.29 153.45 .005 163.43 166.19 Subset Heterogeneity Chi-Squareds Prev ACE R df Older 96.59 96.59 12 Younger 15.34 19.75 8 Less good 112.49 113.46 14 Good 93.92 88.923 6