The fallacy of "two years below grade level for age" as a diagnostic criterion for reading disorders

none - a methodological critique with a worked psychometric argument

US school-age children being screened for reading disability using grade-equivalent scores against grade placement

A critique rather than a study; grade D by the rubric. It is held because it is the clearest peer-reviewed statement of the specific arithmetic failure that makes grade equivalents dangerous, and because that failure is a property of the scale rather than of a sample, so it does not inherit the limits of the reading-disability context it was written for.

Comparing a child's grade-equivalent score against their grade placement - the "two years below grade level" rule that dominated learning-disability identification - is not a defensible criterion, and it fails in OPPOSITE directions at different ages. The same gap in grade-equivalent units substantially overstates the severity of a difficulty at upper grade levels and understates it in the early grades, because grade-equivalent units are not equal-interval: the same number of raw-score points buys a much larger grade-equivalent movement where the norm group's growth curve is flat, and a much smaller one where it is steep. The correct comparison is a standardized discrepancy against expected achievement, not a difference in grade-equivalent years.

Not applicable - a scale-property argument with no estimates and no trait attribution.

Not an empirical finding. The non-equal-interval property of grade-equivalent scales is not in dispute and is restated in every measurement textbook and in test publishers' own interpretation guidance.