Shodhyantri Engineering Services
Research · Analysis · Editing · Publication Support
We help you choose a calculation method, set realistic assumptions, compute the sample with Cochran, G*Power or Krejcie-Morgan and write a clear justification.
Tell us about your sample size calculation need. We reply within one working day.
Sample size calculation is one of the first questions a research committee asks, and one of the easiest to get wrong. Saying that 300 respondents were selected because it seemed adequate rarely satisfies an RDC. Committees want to see a method, a formula or tool, the assumptions behind it and how non-response was handled. A short, correct justification can save a lot of back-and-forth.
We help you pick the right approach for your design. Survey studies estimating proportions often use Cochran's formula with a finite population correction. Studies testing differences or relationships are better served by power analysis in G*Power. The Krejcie and Morgan table is a quick reference for known finite populations. We compute the number with you and help you write the paragraph for your methodology chapter.
Cochran's formula estimates the sample needed to measure a proportion with a chosen confidence level and margin of error. With 95 per cent confidence, a 5 per cent margin and maximum variability of p equal to 0.5, the base figure is about 384. For a known, smaller population, a finite population correction reduces this number.
Power analysis, commonly done in the free G*Power software from Heinrich Heine University Dusseldorf, calculates the sample needed to detect an expected effect size at a chosen significance level and power, often 0.80. It suits t-tests, ANOVA, correlation, regression and chi-square. The Krejcie and Morgan table lists required samples for given population sizes using the same assumptions as Cochran's approach.
We help you choose between precision-based formulas and power analysis, since descriptive surveys and hypothesis-testing studies need different approaches.
Every value used, from margin of error to expected effect size, is written down with a reason or a citation from earlier studies.
We balance statistical need with field reality, such as access to respondents and time, and adjust for expected non-response.
For stratified or cluster samples across districts, blocks or institutions, we help with proportional allocation and design effect considerations.
For power analysis we provide the input and output settings, which committees and reviewers often like to see in an appendix.
You receive clear notes on what your justification paragraph must state, making it easy to write the methodology section in your words.
Ten stages, each with a clear output. Nothing moves to the next stage until you have reviewed the last one.
— You review every stage first, then decide —
The calculation method is checked against your design and main statistical test before numbers are used.
Each assumption is reviewed for a supporting reason, earlier study or standard methodological citation.
Formula steps and rounding are rechecked, with sample sizes always rounded up rather than down.
The final number is compared with needs of planned analyses such as regression, SEM or subgroups.
Target numbers are compared with realistic access to respondents within your timeline and budget.
Your justification paragraph is checked to state method, values, result and adjustment in order.
Every stage is shared with notes explaining what was changed and why. You and your supervisor check it, and the next stage starts only after your go-ahead.
| Common problem | How Shodhyantri helps |
|---|---|
| Using Cochran's formula for a study whose main aim is testing group differences. | Use power analysis based on the expected effect size for the main test, such as an independent t-test or ANOVA. |
| Citing the Krejcie and Morgan table when the population size is actually unknown. | Use the table only for a known finite population. For unknown or very large populations, use Cochran's base formula instead. |
| Choosing a large effect size in G*Power just to get a smaller sample. | Base the effect size on earlier studies or a meta-analysis, or use a conservative small-to-medium value with a stated reason. |
| Forgetting non-response, ending with fewer usable questionnaires than the calculated sample. | Estimate a realistic response rate and inflate the number of people approached, so usable responses still meet the target. |
| Calculating one total sample but planning many subgroup comparisons with tiny groups. | Check that each subgroup has enough cases for its comparison, or limit subgroup analysis to well-powered groups. |
| Reporting the final number without stating any assumptions or the method used. | State the method, confidence level, margin of error or effect size, alpha, power and adjustments in one clear paragraph. |
The test you plan to run on your main hypothesis determines the right calculation. Settle this before choosing any formula.
If no earlier estimate of a proportion exists, p equal to 0.5 gives the largest and safest sample in Cochran's formula.
Sample sizes are always rounded up to the next whole number. Rounding down slightly reduces the precision or power you claimed.
Cite Cochran's work, Krejcie and Morgan's 1970 paper, or Faul and colleagues for G*Power, so your method is clearly referenced.
SEM and factor analysis have their own sample considerations. Confirm your number also suits these if they are part of your plan.
In stratified samples, allocate by population share unless you have a reason for equal groups, and explain that choice.
Take screenshots of the test family, inputs and outputs. They make an excellent appendix and help answer committee questions.
Some returned questionnaires will be incomplete or careless. Collect a margin above target to protect your final usable sample.
Your work is handled by people who read your field's literature, not by a single generalist.
A PhD-qualified specialist from your field who checks the research logic, gap and contribution.
Handles statistics, models and qualitative coding, and documents every analytical decision.
Improves structure, flow, clarity and academic tone without changing your meaning.
Builds and validates MATLAB, Python, ANSYS or COMSOL work where your study needs it.
Checks grammar, consistency, numbering, citations and the small errors examiners notice.
Applies your university or journal template, reference style and front matter exactly.
It is the result of Cochran's formula with 95 per cent confidence, a 5 per cent margin of error and p equal to 0.5. It applies to large or unknown populations when estimating a proportion. For smaller known populations or hypothesis tests, the right number may differ.
They rest on similar assumptions. The Krejcie and Morgan table is a convenient lookup for known finite populations, while Cochran's formula with finite population correction gives the same type of result with visible working. Many committees prefer seeing the formula steps.
Use G*Power when your main aim is testing a hypothesis, such as comparing groups, testing correlations or running regression. It calculates the sample needed to detect a stated effect size with chosen alpha and power. It is free and widely accepted by examiners.
Ideally, use an effect size reported in earlier studies on similar variables. If none exist, a medium effect by Cohen's conventions is often used, but a smaller assumed effect is more cautious. We help you find and justify a value.
We cannot promise any committee decision. We help you produce a calculation that follows recognised methods with clearly stated assumptions, which is what committees look for. If your supervisor prefers a particular method, we work within that preference.
When the population cannot be counted, use Cochran's formula without the finite population correction, or power analysis if you are testing hypotheses. State clearly in your methodology that the population size is unknown and explain which approach you followed.
Larger samples improve precision but cost time and money, and very large samples can make trivial differences statistically significant. The goal is an adequate sample for your main analysis, collected with good quality, rather than the largest possible number.
We provide research guidance, statistical analysis, editing, formatting and training support. All research content remains the scholar's own work, in line with UGC and university academic integrity norms. We do not guarantee publication, acceptance, indexing or academic outcomes: journals decide through peer review and universities award degrees.
Share your topic, draft or data. We will tell you honestly what it needs, and what it does not, before you pay anything.