Sample size calculator for your MD/MS thesis
Pick your study design, enter values from the literature, and get the sample size with full working, standard references and a paragraph you can paste into your synopsis.
1. What is your study?
2. Enter your values
Settings
Source of your values (optional, goes into the paragraph)
Formula and working
For your synopsis
References
Don't know which values to use?
In the free Thesis Manager app, type your thesis title and AI finds the values in real PubMed papers, cites them, and exports the full calculation to Word.
How to choose the right formula
Prevalence or one proportion
For cross-sectional studies that estimate how common something is. You need the expected proportion from an earlier study and the precision you want (often 5% absolute; use half the proportion if it is under 10%).
Estimating a mean
For descriptive studies reporting an average, such as mean serum vitamin D. You need the standard deviation from literature and the precision in the same unit.
Comparing two proportions
For RCTs and comparative studies with a yes/no outcome such as cure or response. You need the expected proportion in each group.
Comparing two means
For RCTs and comparative studies with a measured outcome such as a score or a lab value. You need the difference you want to detect and the standard deviation.
Before–after (paired)
For a single group measured before and after treatment. You need the expected mean change and the standard deviation of that change.
Correlation
For studies of the relationship between two measurements. You need the expected correlation coefficient (r).
Diagnostic accuracy
For studies comparing a test with a gold standard. You need the expected sensitivity, specificity, the prevalence of disease among those tested, and the precision.
Case–control
For studies of risk factors. You need the exposure rate among controls and the odds ratio you want to detect.
Where do I get the values for the calculation?
From a published study with a similar population — ideally Indian — and cite it in your synopsis. A pilot study is the other option. Don't pick values just to make the number small; ethics committees check.
Should I add extra patients for drop-outs?
Yes for any study with follow-up. The calculator divides by (1 − drop-out rate); 10–20% is common for interventional studies.
What alpha and power should I choose?
Most theses use a two-sided α of 0.05 (95% confidence) and 80% power. Use 90% if your guide or ethics committee asks for it.
Why is my result 1–2 lower than G*Power?
These formulas use the normal approximation, as in the standard textbooks. t-test based software can add one or two per group. Either is acceptable if you cite the method.
Results are for planning and should be confirmed with your guide or a biostatistician.