More importantly, in the Lights device that produces much
lower choice reaction times, and which involves obviously
simpler processing demands, the key associations with abstract
reasoning and the Digit Symbol text suggest that the relatively
large effect estimates for intelligence–choice reaction time
associations obtained using the Numbers device (Deary et al.,
2001) are not a result of the Numbers device's requiring complex
processing prior to pressing the response button. Instead, it
appears that the individual differences captured by the Numbers
device are substantially retained in the simpler Lights reaction
time device. Jensen (especially in Jensen, 1987; also in Jensen,
1998; and as discussed in Deary, 2003) stressed that it is
important to pay attention to the changes in parameters—SD as
well as means—between different tests of information processing.
Thus, the increases in SD and mean from the Lights to the
Numbers device indicate that there are extra processing stages in
the Numbers task. However, the individual differences in these
stages do not appear to account for much of the age or higher
cognitive ability variation that is shared with choice reaction
time variance. We also draw readers' attention to our previous
paper, using this same subject sample, that found strong
associations between the Numbers reaction time task and a
computerised task that has similar stimulus–response contingencies
to those used in the Lights task in the present report
(Deary et al., 2011).