Simulation Of Non Response In The Second Visit Data Set

simulation of nonresponse in the second visit data set in order to arrive with a data set that has nonresponse observations with the assumption that the nonresponse observations are missing at random (mar), the generation of random numbers were introduced. these random numbers were then matched to each observation of the second visit data. the second visit observations were then sorted ascendingly by their assigned random number. for each trial, imputation classes under the nonresponse variable, and nonresponse rate in the simulation study, the first group of observations in the data set was chosen as the observations that would be set to nonresponse. after that trial, the same data set that was sorted ascendingly were used and the second group of observations in that data set were used to be performed by the simulation process and so on until the last group of with the same number of nonresponse observations indicated by the varying nonresponse rates. if there are not enough observations or the last group of observations are finished then the process will start again from the top where the second visit data are to be matched with a new set of random numbers. the same process was also done to the imputation procedure that has no imputation classes. this technique was done to be able to generate nonresponse observations faster at the same time at equally distribute the chance that the observations will be set to nonresponse for each trial. the nonresponse rates for each imputation class were equally distributed primarily due to an imputation class which has less than the number of nonresponse observations for that nonresponse rate.