... write it under timed conditions. For each question, after reading the stimulus, go ... only allowed myself to consider each AC once (so if ... would move to the next question).
>
> ... and, untimed, breaking down each stimulus and AC into abstract ...