PakMediNet Discussion Forum : Biostatistics : sample size query
sample size query
Respected Sir,
i have already completed my study on the tumor markers.keeping control to be 60 and tumor size to be 30 but now i realised that sample size explaination is required. sir plz help me how can i defend my sample size.
Posted by: NaumanRaufPosts: 2 :: 18-12-2009 :: | Reply to this Message
Re: sample size query
Dear Sir
from your query i could not understand what you want. the sample size based on certain features like prevalence of disease, power of test, confidence interval, precision, without these we can not calculate the sample size.
Posted by: ibrahim_apPosts: 138 :: 19-12-2009 :: | Reply to this Message
Re: sample size queryYou have not provided enough information. I am assuming that you are measuring tumor markers (probably in the blood) on a continuous scale. You have controls (N=60) and cases (N=30). Lets say that you set alpha at 0.05 (a traditionally chosen value). You also need to provide the difference in means of tumor makers between the cases and controls, commonly called delta. Then you also need to provide within groups standard deviation of the outcome, usually referred to as sigma. Once you have these numbers, you can calculate the power of your study. This will give you some idea as to how far you are from the usual power (generally set at 0.80 or higher). I hope I explained you in a simple way of what is needed to get power of your study. From there you can decide how you would like to justify your sample size.
Posted by: rqayyumPosts: 199 :: 24-12-2009 :: | Reply to this Message