Relational abilities index: A experimental study of the procedure at different trial durations

The relational abilities index (RAI) has been shown to consistently correlate with standardized measures of intellectual aptitude, such as the Wechsler Adult Intelligence Scale. However, the procedure has not been systematically studied and, when used with adults, it has problems in discriminating between medium and high-ability participants. Therefore, the aim of this experiment was to study participants' performance at different levels of analysis and under different trial durations. Sixty-two participants were randomly assigned to one of two groups (30 s or 20 s). We used a version of the RAI involving 55 syllogistic premises (trials), each associated with a YES/NO question. Trials were dived into four blocks, each testing a different relational frame. Results showed that total scores were lower when trials were shortened. However, this reduction was evident only for lower-ability participants. RAI scores in the 20 s group approximated a normal distribution and trial difficulty increased when trial duration decreased. Trial difficulty increased as a function of trial sequence within a block in only half of the blocks. Nodal distance was predictive of trial difficulty. Based on these results, a list of proposed changes to the procedure is provided and discussed.