Trial-to-trial RT variability outpredicts mean RT for general IQ

Correct: C — The intra-individual standard deviation of reaction times across many trials

Across decades of work summarized in 'The g Factor' (1998), Jensen reported that the within-person variability of reaction times — how much a participant's responses fluctuate across many trials of a simple task — predicts general intelligence somewhat more strongly than mean RT itself. He interpreted RT-SD as a proxy for the temporal stability of neural information transmission: noisier nervous systems show both slower mean RT and larger trial-to-trial fluctuations, and the latter is the more sensitive marker. Subsequent meta-analyses have largely supported the claim, though effect sizes are modest and sit within broader processing-speed accounts of individual differences.