I took a cognitive assessment recently and got back not just one number but a whole profile of different scores across different ability areas. That made me wonder whether the single IQ number most people talk about is actually the most useful thing the test produces or whether it is hiding more interesting information underneath it. How does the measurement process actually work and what does it capture beyond the headline score?
The single number is real and useful but it is genuinely just the surface of what a well designed assessment captures 
The measurement process works by sampling performance across multiple cognitive domains, things like fluid reasoning, working memory, processing speed, and verbal ability, and then looking at the shared variance across all of them. That shared variance is g and it gets reported as your full scale score. But the subtest and index level scores underneath it are often more informative for understanding how someone actually thinks. Two people with identical full scale scores can have completely different cognitive architectures, one might have exceptional verbal reasoning and slow processing speed while the other is the opposite. Those differences matter enormously for predicting specific outcomes and designing interventions. RIOT IQ is built around this logic, reporting across 6 CHC domains with 15 subtests so you get the full picture rather than just the headline 
The profile underneath the number is honestly the most interesting part and most people never see it 
A single composite score is convenient but it flattens information that could actually tell you something useful about how your mind works.