Set the clock.
Solve the next problem.

The four practice examples are optional and untimed. Choose an answer, use the main button to check it, then continue after reading the explanation. Scored questions give no correctness feedback before submission.

The 25-minute timer begins only when you deliberately start the scored questions. It continues across reloads, navigation and background tabs. There is no pause. At the deadline, your selected responses are submitted automatically; unanswered questions remain unanswered.

Choose an answer, then Next. You can skip, mark questions for review and use the question map to jump back. All answer cards are clickable. Keyboard shortcuts: 1–4 select; Enter continues. You may submit early after reviewing.

Use a quiet setting without calculators, searches or other people’s help. Paper is allowed. The instructions are in English and the visual items require seeing diagrams. If you prefer no countdown, use the separate untimed cognitive activity; its conditions differ.

Progress and the latest completed result are saved locally when storage works. This separate practice is outside My results, score history and ordinary private backups. Download its private result to keep a copy; it contains your responses. There is no automatic answer upload.

Original questions. The form has eight tasks per area. It combines 24 previously authored Ownfacet reasoning tasks with eight newly authored visual matrices. It copies no Mensa or commercial test questions, graphics or scoring tables.

What is scored. Each correct answer earns one point. Incorrect and unanswered responses earn zero. There is no negative marking, partial credit, speed bonus or IQ conversion. The result separates correct, incorrect and unanswered questions, shows time used and provides item explanations after completion.

What has not been established. This form and its 25-minute limit are development choices. They have not been empirically calibrated, independently validated or normed against an appropriate population. These area counts are not standardized ability scales. Adding timing does not establish IQ accuracy.

IQ points. A defensible IQ estimate needs suitable norms or an externally validated score-linking model for the exact form and intended audience. We have not verified that model for these questions. The existing ICAR source preview and published OMIB matrix model remain separate; their results are not combined into IQ.

Retakes and conditions. These are fixed questions. Familiarity, practice, language, vision, device interaction and interruptions can affect performance. Twenty-four questions overlap with the original untimed activity, so using it first can affect this attempt. This practice is unsuitable for diagnosis, selection, placement or membership qualification.