One standard measurement of IQ is the Stanford-Binet exam. It's been around since 1916 with extensive concern given to consistency in the 5 updates it has recei

One standard measurement of IQ is the Stanford-Binet exam. It's been around since 1916 with extensive concern given to consistency in the 5 updates it has received over the century. And that consistency is determined through thousands of trials with demographically representative samples to avoid predictable issues like a self selection bias in testing. But like you mention the tricky part is that IQ is measured across an average in a generation. The mean is normalized to 100 with a standard deviation of about 15 points. An IQ of 115 does not mean you 'scored' 115 in a vacuum, it means you performed better than 84% of people. The idea that IQ would substantially increase or decrease for an entire population over very short periods of time (thus changing the meaning, in a vacuum, of e.g. 115 IQ over short periods of time time) was surprising, and that's precisely what the Flynn Effect [1] is. However that change is precisely measurable and enables researchers to produce comparable values between generations. [1] - https://en.wikipedia.org/wiki/Flynn_effect https://en.wikipedia.org/wiki/Flynn_effect