The Standards for Educational and Psychological Testing are the best-practice reference for serious assessment claims. IQ WARS follows the spirit of those standards by separating competitive game scores from clinical or high-stakes claims.

IRT models connect item difficulty, response data, scoring, validation, and test linking. IQ WARS is not yet a calibrated IRT exam, but the product direction is aligned with accumulating item-level evidence over repeated official runs.

Carpenter, Just, and Shell modeled Raven-style item solving as rule induction, goal management, and visual-relation integration. IQ WARS uses that tradition for daily competitive reasoning, not clinical diagnosis.

Jaeggi and colleagues reported fluid-intelligence gains after adaptive working-memory training. That result is part of the reason brain-training claims became popular.

Melby-Lervag and Hulme found working-memory training reliably improves practiced tasks, but broad, durable transfer to intelligence measures is much less certain.

Simons and colleagues reviewed the field and found strong evidence for trained-task improvement, weaker evidence for near transfer, and little evidence for everyday far transfer.

Modern RPM-style datasets and reviews use visual analogy problems to test whether machine systems learn relations, rules, and abstractions rather than memorized templates.