I have always assumed IQ tests are just matrix puzzles and pattern recognition but I recently saw a breakdown of a proper cognitive assessment and it seemed way more varied than that. What are the actual task types that show up on legitimate IQ tests and why were those specific tasks chosen? Is there a theory behind what gets included or did psychologists just throw in whatever seemed smart?
The selection is very much theory driven. Modern assessments are built on frameworks like CHC which identifies distinct broad cognitive abilities through decades of factor analytic research. A full battery covers verbal comprehension, fluid reasoning, working memory, and processing speed at minimum, with each task type included because it contributes unique variance to the overall g estimate. RIOT IQ extends this across 15 subtests and 6 domains including reaction time which most traditional batteries underrepresent. The variety is not padding, it is what makes the g estimate reliable.
The matrix puzzle stereotype exists because fluid reasoning tasks are the most g loaded items on most batteries 
But a test that only used matrix puzzles would be missing huge portions of the cognitive architecture. The variety is not padding, it is what makes the g estimate reliable.