I'm taking some measurements using Raman spectroscopy of healthy pig eye samples using a new excitation wavelength and my supervisor said to choose a sample size to allow for spec/sens of 85%. These measurements will be identical just on different, healthy eye tissue to simply show that I obtain similar things each time. Is there a way to estimate/calculate what my sample size should be for this or do I need previous data, numerical limits etc? Any help would be great, thanks!
Georgia, Start by setting the research question(s) and then choose the appropriate statistical test(s) to answer those questions. Afterwards, with the help of a statistical calculator, you need to find the equilibrium between: significance, power, effect size and sample size - here you use the chosen statistical test(s), and the pre-specified values for sensitivity and specificity.
Dear Georgia Harris
Please check these articles, it may help. Good luck.
Article Novel Feature Selection and Voting Classifier Algorithms for...
Article MbGWO-SFS: Modified Binary Grey Wolf Optimizer Based on Stoc...
Georgia, Start by setting the research question(s) and then choose the appropriate statistical test(s) to answer those questions. Afterwards, with the help of a statistical calculator, you need to find the equilibrium between: significance, power, effect size and sample size - here you use the chosen statistical test(s), and the pre-specified values for sensitivity and specificity.
Hi. i think to achieve sensitivity 0/85 you need uses sample size for one proportion formula with p=0/85 and q=1-p.
Georgia Harris , my two cents: I believe you should be using the formula for calculating the sample size for a hypergeometric distribution (above), for n the sample size, N the total number of observations (the population), e the error you are willing to assume and phi the fractile associated with a gaussian distribution with a significance level of alpha/2. In your case, p = 0.85.
I am not 100% sure if this is what you need so I am happy to hear other opinions.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics