A comparison of the Wald conventional confidence interval and Wilson confidence quadratic - type interval is made in relation to how large the sample size should be. Some times statisticians in research projects are asked if the sample they use is sufficient. Naturally, determining the size of a sample is not a simple task. It is. admittedly, itself a skilled task. However, even if it is accepted that the random sample is suitable, the issue is not easy to resolve. Schwertman (1996) refers to a connection between Wilson’s quadratic type confidence limits and Wald's conventional limits. He says: “For large samples which is required for the normal approximation the two philosophies have nearly identical limits'’. The purpose of this note is to provide examples where although n is large enough, the above two philosophies reported fail to be congruent.