The relationship between speed of information processing and psychometric intelligence was investigated by giving 60 university students two choice RT tests (CRT) and Raven's Advanced Progressive Matrices. The modified Hick paradigm evaded shortcomings of the classical procedure (order effects, visual attention effects, response-bias effects, reaction-time/movement-time strategies). CRT-LOW (low S-R compatibility) was developed as a more complex information-processing task involving permanently changing rules to map signals onto responses. The comparison of the results of the modified Hick paradigm with those obtained typically for the classical Hick paradigm shows higher (negative) RT-IQ correlations for the former. It is concluded that controlling for the sources of artifact produced higher RT-IQ correlations. About 20% of the sample did not conform to Hick's law; excluding these subjects raised nearly all RT-IQ correlations. CRT-LOW mainly produced somewhat higher correlations with IQ (compared to the modified Hick paradigm) with a better fit to Hick's law.