PANOPEA: the adaptive figure-matrix test

A short adaptive online test of fluid reasoning with figure matrices, from the PANOPEA research edition PA2.0 (November 2026), for ages 16 to 80. Every problem is a matrix of nine panels, three by three, with the last panel missing, and eight answers to choose from; every row follows the same rules. Its eight families are eight kinds of rule: steady change (a figure turns, grows, gains or moves by the same step), three of each (every row holds the same three values of a feature), overlay (the third panel puts the first two together, or takes the second from the first), cancellation (what the first two panels share vanishes), linked features (one feature follows another: a fill its shape, a direction the number of figures), only some move (only the marked figure follows the rule, and the mark may pass from figure to figure), hidden partners (figures trade places or move round the panel, and each keeps its own rule) and rules on rules (a step that itself changes: it doubles along the row, or the number of figures sets it). Every problem is drawn by code from a formal grammar, and an independent solver, a separate program that sees only the panels, proved that every rule of the grammar that fits a problem points to its one answer and to no other, for all 4,368 problems in the bank.

It opens with two practice problems (one where figures change by steady steps, one with three of each along every row), drawn for your attempt. Answer each and lock it: you see at once whether you were right, the right answer and the reasoning. Practice is never scored. Then the scored problems come one at a time, each chosen from your answers so far: two at the middle levels to begin, then each one the problem that tells most about your level, never more than five from one family, for 24 scored problems; from level 8 up, a level opens only once you have solved the one below it consistently. Choose one of the eight answer panels, numbered 1 to 8, and lock it; a locked answer cannot be changed, and you cannot go back. Every scored problem must be answered: a guess is better than none. Nothing tells you whether a scored answer is right, and your result appears the moment the last one is locked.

Every problem has one of twelve levels, computed from how it is built. Its eight answers are every combination of three deciding facts (say, which figure sits on the left, where the spokes point, and the drop's direction), each either right or as the most natural mistake about its rule gives it, so all three must be found. A fact counts for what it takes to find it: a turn of 45° is easier than one of 135°, putting two sets together easier than canceling them, a figure that stays in place easier than one that must be followed as it moves, and a steady step easier than one that doubles along the row or is set by a count. Above level 3, none of the three is a value that simply stays the same along the row. The rest of the panels adds what must be seen through: the other rules at work, and the figures (one figure, a figure inside another, two side by side, three in a row, four round the panel, up to six small figures, a lattice of up to three kinds of figure, or spokes, with or without a figure at their hub). The levels have model difficulties from about IQ 59 at level 1 to IQ 166 at level 12, and their clocks grow with them, from 45 seconds to two minutes. These difficulties are the model's assumptions, from the construction of the problems: they become measured difficulties only once real first attempts are analyzed. The bank PA2.0 (5 November 2026) rebuilt every level of the first bank, PA1.0, whose problems proved easier than their levels assumed; a certificate names the edition it was taken in.

From the problems you were given and your answers to them, by the provisional model PA-M01: each problem has an assumed difficulty from its level, the same discrimination for every family, and a guessing chance of 1 in 8 (a three-parameter model). The model's estimate of your ability is shown as an IQ (mean 100, SD 15) with its 95% interval. It uses a Jeffreys prior, which holds no score back toward the average: in the simulation of the bank, the estimate stayed within about a point of the true level from IQ 70 to 150, where a short test's usual prior would pull a high score down by several points. A problem left unanswered when its clock runs out scores no point. Two people with the same number of right answers can receive different problems and different results.

A high score must rest on consistent success at the levels below it. From level 8 up, PANOPEA climbs one level at a time, as THETIS does above IQ 180: a level opens only once the one below it is solved consistently, which takes at least three right answers at that level, from at least two families, with at least two in three of its problems right; very consistently, for 160+, at least four right, with at least three in four of its problems right. So a score of 135 or more needs level 8 (problems of IQ 127) solved consistently, 145 or more level 9 (IQ 137), 150 or more level 10 (IQ 147) and 160+ level 11 (IQ 156) very consistently. Until a level is solved that way, no problem above it is given, and a score is held just below the next step, however well the answers above it went; the result and the certificate say so.

PANOPEA is a research edition: no human calibration has been done yet (N = 0), so a score is a provisional position on the model's scale, not a percentile of the population, and no percentile is reported. Twenty-four problems make a short test, so its interval is wider than a long test's: PANOPEA answers how good your reasoning is in about half an hour, and for a precise measure in the high range, take THETIS. The scale is reported up to IQ 160+; an estimate below IQ 60, the level of the easiest problems, is reported as below IQ 60. Beside the IQ, as THETIS does, your result gives the rule complexity: the level, from 1 to 12, of the problems you solve half the time beyond chance. It describes what you solved, so it depends on no population norm.

Speed is part of reasoning on figure matrices, and a clock keeps every attempt comparable: the same time for the same level, for everyone. Each scored problem's clock is its level's, from 45 seconds at level 1 to two minutes at level 12, and the server keeps it, so reloading the page does not reset it. When the clock runs out, an answer you have chosen is locked; if you have chosen none, the problem counts as unanswered and the next one follows. Between problems there is no clock: the next problem appears only when you ask for it, so you can take a break. Practice has no clock.

Each scored problem has its own clock, from 45 seconds at the first level to two minutes at the hardest, kept by the server: when it runs out, an answer you have chosen is locked, and if you have chosen none, the problem counts as unanswered. Between problems there is no clock: the next one appears only when you ask for it, so you can take a break, and your private link brings you back on any device. The clocks of a whole attempt add up to about 28 minutes at the middle levels and about 35 at the higher ones; most answers come before their clock runs out, so most attempts should take 25 to 30 minutes. From the moment you begin, your attempt stays open for 90 days; one still open then is scored with the answers locked by then, provided it has all 24 scored answers; otherwise it ends without an IQ.

Yes, whenever you like once your previous attempt is finished, and every attempt is on problems you have never been shown. Every problem counts as seen the moment it reaches your screen, together with every problem made from it by renaming its shapes and fills or turning the whole matrix, under this email or another registered with the same name and year of birth; no later attempt meets it again. Practice gets new problems too, while they last. Every attempt has its own fee and its own certificate. Register every attempt with the same email. Only first attempts go into the research data on the test's norms.

The US$19.99 fee, paid through PayPal before the attempt opens, covers one attempt: the practice problems with their explanations, the 24 scored problems, your result with its PDF certificate, and a copy by email. Each retake has its own fee.

Yes, from 5 November 2026. IQ 147 or higher meets the criterion of the Nereid Society and IQ 160+, the top of PANOPEA's scale, that of the Mesopelagic Society (with a one-time membership fee). Just below the criteria, IQ 135 to 146 qualifies for associate membership of the Nereid Society, with a one-time admission fee. The criteria of the Aphotic, Hadal and Exocosm Societies lie above PANOPEA's scale, which ends at IQ 160+. Your fee includes the review of one society application. Your best result counts, as for THETIS, and it counts as PANOPEA's ladder holds it: a high score must rest on consistent success at the levels below it (see “How is my score worked out?”). PANOPEA is a research edition, scored with provisional model norms until enough first attempts are in, and the author reviews every result before admission.

To build PANOPEA's norms from real people. They are re-estimated from first attempts only, anchored on examinees whose IQ is known from a supervised test, such as a Wechsler or Stanford-Binet scale given by a psychologist, Raven's or Cattell's test given under supervision, or a high-IQ society's supervised admission test. Giving yours is optional, and it never changes your own result: it only helps place the problems on the IQ scale for everyone. Only attempts with a clean integrity record are used, and a new calibration is applied only when, checked on people it was not fitted on, it comes closer to their known IQs than the scale in force. If you allow it, the author may write to you to confirm the score.

The answer keys and the bank never leave the server, and your result never shows which scored problems you solved. Only the problem on screen ever reaches your browser: the next one is chosen on the server after you lock your answer, from the answers you gave, so there is no list of problems to look up in advance, and a retake is always on problems you have never been shown. Each problem is drawn on the screen with a faint watermark of your name and certificate ID. While a problem is on screen, the test page records leaving the test window, screenshot shortcuts (Print Screen, Windows+Shift+S, Command+Shift+3, 4 or 5), printing (Ctrl+P or Command+P), copy, right-click and drag attempts, and AI assistant add-ons that insert themselves into the page; the server keeps its own times (when each problem was shown and when its answer was locked) and notes right answers given moments after a long absence, and many right answers in a few seconds. As on THETIS, it checks the answer pattern against the difficulty of the problems you were given (a pattern unlikely for one person at your level is noted, as is a result at the top of the scale), and practice problems are recorded too. The record is sealed when you finish, and its fingerprint is printed on your certificate. The author's integrity report of each attempt lists all of it, problem by problem. None of this changes a score, and unusual results are reviewed, never judged automatically. No online test can rule out every form of help, which is why you declare that you will work alone.

Your private link (shown when your test is ready, and kept in this browser) reopens your test on any device, at the problem you have reached. If you lose it, use “Recover my test” with the PayPal Transaction ID from your receipt and the email you registered with. Every locked answer is on the server.

Yes: the six example problems above, with their answers and reasoning. They were made for this page and are not problems of the test.