What this site is
zstat.org is a free set of statistics calculators and university admission score references, available in English, Traditional Chinese, Simplified Chinese, Korean and Japanese. It covers twelve tools: seven general statistics tools and five admission-score references for Hong Kong, Taiwan, Japan, Korea and mainland China.
It started from a narrow, practical observation. Japanese hensachi, Korean CSAT standard scores, IQ deviation scores and psychometric T-scores are all the same conversion — a Z-score on a different scale — but every online tool treated them as separate, unrelated calculators. Building one place that showed a value in all of its forms at once, and then extending that to the admission systems that rely on the same mathematics, is what this site is.
Who maintains it
zstat.org is built and maintained by an independent developer, not by an institution, a publisher or an examination body. It has no editorial board and no institutional backing, which is precisely why the methodology is documented in as much detail as it is: you should be able to check the work rather than take it on trust.
Corrections are welcome and acted on. If a number here is wrong, the contact page explains how to report it and what happens next.
How the tools are built and checked
Every calculation runs client-side in your browser using published, citable algorithms rather than lookup tables — the Abramowitz–Stegun approximation for the normal CDF, Acklam's algorithm for its inverse, and the standard incomplete beta and gamma function treatments for the t, F and chi-square distributions. The methodology page lists all of them with their accuracy bounds.
Each tool also discloses the specific choices that affect its results:
- The two-sample t-test uses Student's equal-variance form, not Welch's.
- The 2×2 chi-square applies no Yates' continuity correction.
- Power analysis uses the normal approximation, not the exact non-central t and F distributions.
- Effect size conversion between d and r assumes roughly equal group sizes.
- Quartiles use linear interpolation (R type 7 / Excel
QUARTILE.INC).
None of these is hidden. A calculator that quietly makes a choice on your behalf is worse than one that tells you what it did and lets you decide whether it fits.
How the admission data is handled
The admission reference data is the part most likely to mislead, so it is treated with more caution than the mathematics:
- Placement bands — rough ranges synthesised from public sources, explicitly labelled as unofficial. They answer "am I in roughly the right territory?" and nothing more precise than that.
- Per-programme admission figures — verified individually against the university's own published statistics, with a source link on every row. This list is deliberately short. Where two sources disagreed on a figure, it was dropped rather than reconciled by guesswork.
- Scheduled review — reference data carries a machine-readable freshness marker with a verification date and a review interval, and a quarterly automated check flags entries that are due for re-verification.
- Deliberate omissions — the Gaokao page has no calculator, because no valid cross-province conversion exists. Publishing one would be more harmful than publishing nothing.
How it is funded
The site is free and carries advertising through Google AdSense to cover hosting and domain costs. There is no paid tier, no account system, no newsletter and no data sold to anyone. What the advertising and analytics do collect is set out in the privacy policy.
Advertising placement has no influence on content. There are no sponsored programme listings, no affiliate links to preparatory courses, and no paid placement in any band table or reference list.
What it deliberately does not do
- It does not predict admission. The bands are historical orientation. Actual outcomes depend on programme-specific weightings, interviews, cohort size and year-to-year variation that no reference table captures.
- It does not offer non-parametric tests. Mann–Whitney, Wilcoxon and Kruskal–Wallis are not implemented. The test wizard tells you when you need one and points you to software that has it, rather than pushing you toward a parametric test that would fit badly.
- It does not store your data. There are no accounts and no database of user inputs.
- It is not affiliated with any examination authority or university. Every admission page links to the official source, which is what you should act on.
Languages
All twelve tools are available in five languages, each on its own URL so that search engines and readers get the right version directly. The language selector in the header moves you between them while keeping you on the same tool.