The Effect o f Stratified Sampling with proportional Allocation on Inference about Population Mean

A proper analysis o f survey data requires that sampling design be taken into account, when conclusions are wanted about finite population. However, many computer programs fo r standard statistical analysis implicitly assumes simple random sampling. Accordingly, this study undertakes to examine the effect o f proportional stratified sampling on the common type o f simple statistical analysis. The result o f the study indicates that the usual use o f standard procedures can lead to erroneous variance estimate and therefore invalid conclusions