I've added some detail to my question
Here is how I'd like each trial in my dual task dual RT experiment to go
Black background, white letters
Fixation +
Spacebar press
400 ms delay
presentation of randomized H O or S at centre of screen make speeded
response to idenity of letter using j, k, l respectively record RT letter is
erased after key press SOA from onset of letter of 50, 111, 245, 542, or
1200 ms Presentation of randomized low or high pitched tone for 250 ms Make
speeded response to pitch of the tone using letter a or s respectively
Record RT Must respond to the letter before responding to the tone.
Note that the SOA is from the onset of the letter, not the response to the letter and thus there must be two events, the SOA and the tone presentation, before the letter response and there must be one event following the presentation of the tone but before the tone response (the letter response). This experiment is demonstrating that tone responses will be delayed by SOA.