Sample statistics are: complexity: mean=70; SD=12 fault density: mean=1.5; SD=0.5 corr coeff = 0.6 1) Estimate the fault density of a module of complexity 82. Since you are estimating (that is predicting) fault density for fixed complexity this is a regression problem and does not require normal theory. Note that the run is 12 and so: rise = 12(0.6(0.5/12)) = 0.3 Hence fault density of 1.8 can be expected. 2) For modules of complexity 70 what proportion had fault density less than 1.5. Since (70, 1.5) is on the regression line then 1.5 must be the mean of modules of complexity 70 and so the proportion is 50%. 3) For modules of complexity 82 what proportion had fault density less than 2. We already know that the mean of these programs is 1.8 and so we need to compute the SD. SD(y|x) = 0.5*sqrt(1-0.6^2) = 0.5*0.8 = 0.4 Compute z: z = (2 - 1.8)/0.4 = 0.5 From the standard normal table, 38.3% of the area under the curve is between -0.5 and 0.5. The desired proportion is therefore 50 + 19.15 = 69.15%. 4) What proportion of all modules had a fault density less than 2. Since you are interested in "all" modules, this does not require regression. The z score is: z = (2 - 1.5)/0.5 = 1 From the standard normal table, 68.27% of the area under the curve is between -1 and 1. The desired proportion is therefore 50 + 34.14 = 84.14%.