Earlier versions of IQ tests used a more "generous" formula for calculating children's results, ranking those who scored as well as subjects older than them higher on the scale on the assumption that intelligence increases as children get older. The age at which such growth slows/ceases varies with the individual (generally age 16-18) and is obviously inapplicable to adults, who are scored strictly in relation to overall average results.
As the amount of test data grew over time, old tests were refined, and newer tests were created, scoring calibrations have been "tinkered with", so children's results are usually scored in comparison to others close to their own age. It is the scoring metrics, rather than the children's actual results, that have changed.
You know this for a fact? College entrance exam scores have been on the decline. Testing at all grade levels reflect growing percentages of students who are underachieving academically.
It is not just IQ scores.
Earlier versions of IQ tests used a more "generous" formula for calculating children's results, ranking those who scored as well as subjects older than them higher on the scale on the assumption that intelligence increases as children get older. The age at which such growth slows/ceases varies with the individual (generally age 16-18) and is obviously inapplicable to adults, who are scored strictly in relation to overall average results.
As the amount of test data grew over time, old tests were refined, and newer tests were created, scoring calibrations have been "tinkered with", so children's results are usually scored in comparison to others close to their own age. It is the scoring metrics, rather than the children's actual results, that have changed.