Mento does not write your thesis. It makes it defensible.
The hard part of a thesis is not producing words. It is answering the panel when they ask how you chose your sample size, why that statistical test, and how they know your hypothesis came before your results.
Mento keeps the record that answers those questions — and refuses to let you claim more than your design supports.
A sample size you can justify
Most calculators written for Philippine theses implement one formula and apply it to every study. Slovin's formula is valid for estimating a proportion, at 95% confidence, with a known population — outside that it answers a different question. Ours asks what you are estimating first, then shows the formula, the inputs, and the source.
Try the calculator — free, no accountA question your later steps can read
PICO, PICOT, PEO, SPIDER, SPICE and CIMO, recommended from what your study actually has rather than picked from a list. The framework is chosen by asking whether you have a comparison group and whether a timeframe is part of the claim — so you are not invited to invent a control that does not exist.
A design that decides what you may claim
The design is worked out from what you can physically do: assign participants, randomise that assignment, measure more than once. Randomisation is what separates an experiment from a quasi-experiment, and the design you record governs whether your write-up may say one thing affected another or only that they were associated.
A hypothesis with a timestamp
Locking a hypothesis records when it was formed, and the lock cannot be moved or cleared afterwards. That timestamp is the difference between a prediction and an explanation written once the results were in — which is the distinction a panel is really testing when it asks how you arrived at your hypothesis.
The right test, with its reasoning
Test selection follows from your variables' measurement levels, your group count, and whether observations are paired. You see why a test was chosen and what was rejected — because “use Welch's t-test” is an answer you cannot defend. Assumptions are checked on the model's residuals, not the raw data.
Statistics computed by scipy, not by a chatbot
The analysis runs in your browser against scipy — the same library the field validates its own work with. No language model computes any figure or writes any analysis code. Every run records its engine version and a hash of your data, so the result can be re-executed and checked rather than taken on trust.
A record of what you decided, and when
One document assembling your locked hypothesis, your sampling computation with its citation, every analysis you ran including the ones you did not report, and which assumptions were checked on what. It reports its own gaps rather than hiding them — an export that quietly dropped them would launder the thing it exists to evidence.
What it will not do
- It will not decide whether your topic is novel — that needs a literature search, and guessing would be worse than saying so.
- It will not clear your study ethically. That is a committee's job; Mento only helps you budget the time and assemble the packet.
- It will not replace your adviser. It gives you something concrete to bring them, and a read-only link so they can see it.
- It will not write a citation, and it will not invent one — the reference tooling is still being built, so today Mento simply does not touch your bibliography.